Designing A Limiter For Headphone Amplifiers.

(A HeadWize Design Paper)

limiter.gif

When used with headphones – especially personal monitors, audio limiters (also called clippers) can help maintain safe listening levels. Without a limiter, large transients in audio signals could generate dangerous SPLs in headphones, although the average volume might be set at an acceptable level. Because the ambient noise on stage can be fairly high, performers are tempted to steadily increase the volume of their personal monitors in order to hear the mix. Limiters can ensure that headphone volume never exceeds a pre-determined threshold.

Because a limiter is actually a specialized form of audio compressor, the circuitry of audio limiters can vary tremendously in complexity from a pair of diodes to multi-stage voltage controlled amplifiers with split frequency bands. Diode-based limiters are instantaneous, simpler in design and have a more accurate loudness response, but suffer from distortion in the clipping region. VCA-type limiters have low distortion, but can exhibit breathing or pumping effects from poor gain control. This article takes especial note of design principles for diode-based limiters and the wide range of clipping characteristics. For more information about limiters and compressors, see Signal Processing Fundamentals.

A diode has infinite resistance until the voltage across it is high enough to forward bias it (typically 0.7V for a silicon diode) at which point, current flows. The diode’s voltage drop remains fairly constant throughout the operating range, and it is this property that is exploited in diode-based limiter circuits. Since audio signals are AC, two diodes together can symmetrically clip both the positive and negative halves of the waveform.

limiter1.gif
Figure 1

The output of a hard limiter has a maximum voltage that is the forward bias voltage of the diodes. Figure 1 shows the most basic of diode limiter circuits, a hard limiter. The design assumes that the input will exceed 1V. The trimmer pot P is adjusted for maximum volume from the headphone amplifier. Because the clipping can be abrupt and drastic, the distortion of a hard limiter in the clipping range is harsh (crackly) and is sometimes used for a fuzzbox-like effect. By scaling the resistors down, the hard limiter (and the soft limiter in figure 2) could drive some headphones directly.

limiter2.gif
Figure 2

The hard limiter can achieve an output with a slightly more dynamic characteristic with the addition of a resistor to form a voltage divider with R1. The soft limiter shown in figure 2 still clips a waveform above the forward bias voltage, but also lets through an amount (Vin – 0.7V)/10 by taking the output above R2. The graph in figure 3 compares the response of both types of limiters. The soft limiter is a type of compressor with a compression ratio determined by R1 and R2. Increasing R2 will give a more dynamic sound at the expense of level control.

limiter3.gif
Figure 3

The harsh characteristic of a hard limiter is partially due to the odd-order distortion harmonics that are generated during clipping. Jack Orman at the Analog Music Zone has analyzed diode clipping and has documented that germanium diodes (such as the 1N34A) have a smoother, tube-like clipping quality because they distort with more even-order harmonics, which are more pleasant to the human ear. Germanium diodes have a forward bias voltage of about 0.3V and are usually deployed 3 per side.

limiter4.gif
Figure 4

A pair of LEDs improves on silicon or germanium diodes as they clip with stronger even-order harmonics. Orman believes that a pair of one red and one green LED will sound better still, because the mismatching will highlight the even-order distortion. The limiter in figure 4 is an active design with the LED clippers in parallel with the feedback resistor of an opamp. The clipping threshold is about 1.9V, which is the forward bias voltage of an LED.

limiter5.gif
Figure 5

Figure 5 shows two limiter configurations based on zener diodes. The circuit in 5a clips when the input exceeds the breakdown voltage of one zener and the forward bias voltage of the other. The second limiter in figure 5b requires only one zener diode and clips when the input is greater than the breakdown voltage of the zener combined with the forward bias voltages of each diode pair. Zeners have the advantage of being easier to match for symmetric clipping than other types of diodes over a wide range of voltages.

limiter6.gif
Figure 6

The limiter circuits shown so far have fixed clipping levels (and are adapted to different headphone amplifiers by way of attenuators). One way to vary clipping levels is to apply a bias voltage to the normally grounded side of the diode array. The circuit in figure 6a permits manual adjustment of the positive and negative portions of the waveform separately. The variable clipper in figure 6b puts the audio signal through an inverting amplifier, the output of which connects to the bottom of the zener array. The limiter clips when the audio signal exceeds |(Vf + Vz) / (1 + (R2 * R5 / R3 * Rx))|, where Vfis the forward bias voltage of a zener and Vz is the breakdown voltage. R5 adjusts the feedback of the inverting amp and thus the clipping level.

limiter7.gif
Figure 7

Broadband limiters work across the entire frequency range. There are times when clipping should be frequency dependent. For example, if the audio signal is laced with noise, then clipping the high frequencies can result in smoother, more intelligible sound. The circuit in figure 7 is one example of a bandwidth restricted limiter. The low frequency clipping level of the diode array is modulated by the output of the low pass filter, so that only the high frequencies are affected.

limiter8.gif
Figure 8

Another way to configure a frequency dependent limiter is pre-emphasize the audio signal before the clipping stage. The analog tape compressor by LXH2 in figure 8 simulates the high frequency clipping distortion in analog tape recorders (useful in digital recording) by applying a 70 microsecond pre-emphasis (high frequencies above 2.4kHz) in the first stage network. After the clipping stage, the signal is de-emphasized in another network with the same corner frequencies. The diode array is a mix of silicon and germanium diodes to achieve the desired clipping characteristic.

limiter9.gif
Figure 9

When the audio signal is clipped, a limiter will generate harmonic and intermodulation distortion above and below the frequency of clipping. A multiband limiter (figure 9) minimizes these distortion components by splitting the audio signal into pre-determined frequency bands, each with their own clipping stage. The output of each clipping stage is then filtered for distortion outside the band before all bands are recombined into a single waveform.

In practice, multiband limiters are difficult to implement. Because the various filters of different orders will phase-shift the signals in each band, the overall response of a multiband limiter may not be flat and can sound unnatural. Several techniques can improve the performance of multiband limiters. First, the bands can be chosen according to psychoacoustic masking curves that help conceal distortion. Second, compressors can be introduced before the clipping stage to control the degree of clipping. Distortion detectors can control the amount of compression. Third, all-pass filters can be sprinkled about to collimate the phase responses of the bands.

limiter10.gif
Figure 10

On average, the bass frequencies in an audio signal usually have the highest amplitudes. When a low frequency signal of sufficient amplitude trips a limiter, high frequencies are also “pinched.” The design in figure 10 is a multiband clipper that specializes in minimizing the harmonic and intermodulation distortions due to low frequency clipping only. The high and low pass filters split the audio signal with the low frequencies going through a variable clipper that is controlled by an intermodulation distortion detector. The second low pass filter after the variable clipper removes any harmonics caused by the clipping before summing the output with the high frequency band.

The IM detector estimates the amount of high frequency clipping in the fixed clipper resulting from low frequency clipping, and correspondingly decreases the threshold in the variable clipper to decrease the low frequency amplitude. The differencer recovers high frequency peaks, which determine the clipping level in the variable limiter. The peak detector enables the IM detector only when there are peak bass frequencies.

limiter11.gif
Figure 11

With any limiter, if the audio signal is conditioned with high frequency pre-emphasis (see the analog tape compressor above), then de-emphasis will reduce the high frequency distortion components due to clipping. The circuit in figure 11 cancels low frequency distortion. The audio signal pre- and post-clipping is compared in a differential amplifier and the output is put through a low pass filter to extract the low frequency distortion. Since the output of the low-pass filter is phase shifted, the clipped signal is delayed through an all-pass filter (assuming a minimum phase low-pass) to compensate and then differenced with the output of the low-pass filter. In a multiband limiter, the output of each band may have its own distortion cancellation circuit.

References:

Burger, John Robert, “A Voice-Balancing Audio Peak Clipper,” QST, July 1998, p. 45.
Chandler Jr., James, “Simple Projects,” Electronic Musician, January 1990, p. 24.
Fukushima, Isao, “Limiter Circuit For Removing Noise From Demodulated Signals,” U.S. Patent No. 4,256,975, March 17, 1981.
Orban, Robert, “Apparatus and Method for Peak-Limiting Audio Frequency Signals,” U.S. Patent No. 4,208,548, June 17, 1980.
Orban, Robert, “Multiband Signal Processor,” U.S. Patent No. 4,412,100, Oct. 25, 1983.
Schlesinger, Eugene, “Circuit Having Adjustable Clipping Level,” U.S. Patent No. 4,138,612, Feb. 6, 1979.
Werrbach, Donn R., “Split-Band Limiter,” U.S. Patent No. 5,737,432, April 7, 1998.

Addendum

12/9/2007: Corrected figure 5b.

c. 1999 Chu Moy.

The Collected Grado Headphone Mods.

by “Skippy” et al.

Editor: Although Grado headphone owners enthusiastically praise the sound quality of their headphones, they are equally likely to complain about the styling, fit and spotty reliability (such as the infamous “Grado grattle”, a distorting condition that occurs when the transducer diaphragm develops a crease or wrinkle). Over the years, DIYers have devised several modifications to improve the sound, comfort and even the reliability of their headphones.

This collection of Grado headphone modifications comes from posts by Skippy, Beagle, TimD, Voyager, Neruda, Chych and Squirt in the HeadWize forums and from Kevin Gilmore’s Pure Class A Dynamic Headphone Amplifier project. Most of the mods apply to the Grado SR-60, SR-80, SR-325 or the Alessandro-Grado MS-1, but may be adaptable for other Grado models (and for other brands of headphones). Some of these mods (especially the transducer mods) can permanently damage the headphones. Neither HeadWize nor the authors accept any responsibility for the destruction of Grado headphones resulting from any of these mods.

The Transducer Mods

Damaged headphones should be sent back to Grado Labs for repair. On occasion, it may be necessary to open the earcups either to repair the transducers (if returning them to Grado Labs is not an option) or perform other modifications to improve the sound. The components inside the earcup are fragile, so DIYers must exercise extreme care to avoid damaging the transducer further. Again, neither HeadWize nor the authors accept any responsibility for the destruction of Grado headphones resulting from these mods.

grado_mod1.jpg

Be especially careful working with metal tools near the strong magnet of the voice coil. The pull from the magnet can cause a tool to rip the diaphragm or a piece of metal can snap off and attach itself to the magnet, altering the sound.

Opening the Grado Earcups (SR-60)

Method 1 (Skippy): Pour boiling water onto a cookie sheet (about 1/8-1/4″ deep) and put the Grado earcups grill-side down into it. Keep them there for about a minute, and then remove them. The glue softens and the earcup sections should slide apart.

grado_mod2.jpg

Method 2 (Skippy): Insert a small screwdriver (a jeweler’s screwdriver will do) into the crack between the 2 halves of the earcup and carefully pry. Insert a second small screwdriver a bit to the right of the first one. Carefully pry. Take the first screwdriver and insert it a bit to the right of the second one. Pry carefully. Repeat this process, going around the cup until the seal loosens. Finally pull the cup apart.

Method 3: (Beagle): “Blow dry” the plastic enclosures with a hair dryer for about 1 or 2 minutes. Point the hairdryer at the sides of the enclosure, not directly at the front or back. If you fire from 3 or 4 inches away and take a little more time, you get the same results with no risk of affecting the plastic diaphragm. This softens the glue, and you can they pry or twist the enclosure apart.

Note: Beagle says that this method has also worked to open the earcups of his SR-325 headphones, but NOT to use a screw driver: “The aluminum actually conducts the heat better so you have more time to work. Don’t use the screwdriver method as it would make marks on the housing. I heat the sides of the air chambers, then use the pull/twist method to extract the driver portion. I remove all the original glue and use my gluegun to re-seal. This glue is easier to soften if I need to go in again.”

grado_mod15.jpg

Method 4 (Mole): Thanks to the collected Grado mods page I got the courage to take apart my Grado SR125s. It was nerve wracking to say the least, because none of the suggestions on how to take Grados apart seemed to work. First I tried prying with a small screwdriver, but I didn’t have the heart to do that much damage to the housing. Then I tried lowering the grilles in hot water. No effect. Still couldn’t twist the cover off.

Finally I came up with another idea: I found a teaspoon with a sort of tapered handle that I used as a wedge to gradually ease the two parts of the cans apart. You keep the handle of the spoon pressed flat into the crack between the two halves of the Grados using your thumb, and then you pull on the “bowl” of the spoon so that the wider part of the handle forces the two halves apart. You start of where the handle is narrow enough to fit and then force the halves apart. It’s all a matter of finding a teaspoon with the right shape.

One of the reasons I think the other methods didn’t work was that the SR125 has some kind of rubbery superglue applied where the edge of the transducer half meets the grille. From what I understand other Grados (SR80, SR60 etc) have superglue applied BETWEEN the two halves of the housing, so by prying them apart the glue will “snap”. I also doubt the rubbery stuff (somewhat like Liquisole used for repairing shoes) is affected by moderate heat in any substantial way.

Instead of resealing with a glue gun, I placed a strip of black electrician’s tape along the inside of the grille part to make the fit nice and snug. That should make it easier to take the cans apart again if the need should arise. (I’m planning on applying the “mini-phono jack” mod in the future.)

Removing the Transducers (SR-60)

grado_mod3.jpg

Method 1 (Skippy): There are little bumps on the back of the earcup that serve as clips to hold the transducer in place. Sand the bumps down just enough to take out the driver, so that they can go back into position with a bit of force. If the bumps are sanded down too much, the driver will not stay in place when reinstalled. Beagle wrote that he glued them in place. When I installed the submini jack for the replaceable cable (see below), the jack pressed against the back of driver and holds it in place.

Note: the voice coil wires are fragile. Be careful when removing the transducer to avoid breaking a signal connection.

Method 2 (Beagle): Using a medium size screwdriver, going in from the back of the driver, insert the screwdriver along the edge between the driver and the plastic housing. Turn the screwdriver with a slow wiggling motion, gently attempting to “pry” the housing away from the driver. Do this every 1/4 inch or so. You should hear a cracking sound as the glue (probably superglue) lets go.

You may get the driver itself loose or the driver with the perforated cap still attached. If the cap is still attached, the driver should come out through the front of the plastic enclosure. If you have pried just the driver loose, you will either have to remove the plastic bumps on the enclosure and bring it though the back or try prying the driver cap from the enclosure. Use your judgment.

Fixing the Rattle Noise in the Diaphragm (SR-60/SR-325)

Method (Beagle): One of my SR-60 drivers developed a “rattle” on bass notes at somewhat high volumes, and it also caused a channel imbalance. Since they were out of warranty, I decided to try and fix it myself. I think the “wrinkle” occured when the diaphragm was hit with a sudden or prolonged low bass resonance at high volume. I know that his was indeed the case with the SR325, while I was playing the HDCD remaster of Mike Oldfield’s “Tubular Bells”. There is a lot of low bass energy and extreme dynamic range from quiet sections to loud. I think it is probable that the energy caused the diaphragm to snap back suddenly and retract a tad too far.

Once I got in view of the diaphragm, I could see a “collapse wrinkle” on the diaphragm. Using a small piece of duct tape, I VERY CAREFULLY applied a small area of the tape VERY GENTLY on the affected area of the diaphragm and pulled up slowly and gently. A couple of attempts and the diaphragm “snapped back into place” i.e. the “wrinkle” was gone.

I then used the hair dryer (held about two inches away) to “cure” the diaphragm by “stretching back” the plastic. I applied the heat for about 15-20 seconds. Before gluing everything back together, I placed the driver back in the enclosure, placed the rear enclosure on and gave it a listen. I used superglue to adhere the drivers to the housing and glue gun to hold the enclosure halves back together. This method worked to fix a similar problem with my SR-325 a few months back.

Reattaching the Transducer’s Signal Wires (SR-60)

Method (Skippy): A voice coil wire broke underneath the diaphragm of my SR60, so I peeled off the diaphragm, and extended the wire to repair the driver. The transducer consists of the diaphragm, the coil, the magnet assembly, and the big plastic cup part. The coil is glued to the diaphragm, and the magnet assembly is plastic cup. the magnet assembly consists of a circular magnet in the center, and a metal ring (more like a cup actually) surrounding the magnet. the voice coil sits inside the gap in between the magnet and the ring.

First, I applied some masking tape to the diaphragm to keep it from ripping or stretching during the operation. The shape of the diaphragm isn’t flat, so one big circle just doesn’t work. I put little strips in a radial fashion overlapping a lot. And I put some ring-shaped pieces on top of those, and then some more just X-ed all over the place – a lot of masking tape. It is very important to use masking tape; The adhesive on other tapes would cling too strongly to the diaphragm, and you’d probably ruin the diaphragm removing the tape.

The diaphragm is stuck on to the big plastic cup part using glue. This glue isn’t too strong, around the same adhesiveness as scotch tape. I used a pin to separate the diaphragm from the plastic cup – moving very slowly and very carefully.

I slowly peeled the diaphragm off, exposing the voice coil. Carefully, I unwound a loop of wire from the coil. Then I replaced the diaphragm back onto the transducer. The glue residue was still sticky enough to hold the diaphragm without using any extra glue. That part was hard, but it gets harder.

I couldn’t dissolve the enamel coating on the wire. Usually I scrape the stuff off, but the Grado wire is so thin and fragile that any scraping would rip the wire to shreds. Unfortunately, scraping the enamel was the only method I had left. I shredded the wire many times and almost had to remove the diaphragm again to unwind another loop, but I was finally able to scrape off the enamel without breaking the wires.

Soldering those wires is hard because they’re so thin and fragile. I couldn’t even hold them with tweezers without snapping them. I ended up using the tip of the soldering iron to control the wire. It actually worked very well.

Note: the wires that connect to the transducer are uninsulated. When the cord moves around the uninsulated wires brush up against each other, causing clicking sounds. A little electrical tape fixes this easily.

The Earcup and Headband Mods

Replacing the Plastic Grill with Wire Mesh (SR-60)

Regarding the value of having a mesh grill, Beagle had the following comments comparing the SR-225 to the SR-325: “The rear screen on the 225 is metal and more open than the plastic one on the 125. This lets more the rear firing sound out. The plastic one, being plastic and not as open, causes some sound to bounce off of it and causes a slight honky quality compared to the SR225.”

Method (Skippy): The drilled plastic comes of easily from the inside. I cut out a circle of wire mesh (called screen fabric at hardware stores) and put it in place of the drilled plastic grill. I was hoping that the mesh would improve sound, as Grado Labs advertises mesh grills on the higher level models; however to my ears, no improvement was recognisable.

grado_mod4.jpg

Adding a Cable Jack (SR-60)

Method (Skippy): I made the cable of my SR-60s replaceable by mounting a mono 2.5mm sub-mini phone jack to each earcup, and a 2.5mm phone jack on the end of the cable for each channel. Just open up the earcups. Desolder the cable. Solder wires from the transducer to the phone jack. Mount the small 2.5mmm mono phone jack to the front part of the earcup. The back piece of the earcup may need a little filing (I used mini files that I got at Radio Shack).

grado_mod5.jpg

I can remove the cable easily, so swapping cables is easy. My Sennheiser HD490s have a similar plug. I kinda like the fact that when the Grado cable is tugged, it falls out. When i used my HD490s portably, the cable would get snagged in odd places. If the cord didn’t come out, either i would have fallen, or some serious damage would have befallen my headphones.

Adding a Collapsible Headband (SR-60)

Method (Skippy): I replaced the Grado headband with a earmuff headband and bent pieces of a wire coat hanger to hold the transducers to the headband. My Grados are now collapsible, very comfy and very portable. It’s surprising how small SR-60s actually are. When the cable is removed and the headband is collapsed, they’re small enough too fit in a minidisc bag!

grado_mod6.jpg

Now for the bad part: the earpieces no longer swivel, nor do they have the cool looking “antenna.” For my next project, I will attempt to preserve the swiveling feature and the antenna, while keeping them collapsible.

Removing the Model Number Button on the Earcup Grill (SR-60/SR-80)

grado_mod9.jpg

Method (Skippy): The model number button comes off very easily. Just grab the button with a pair of pliers and twist it off. Now they look like those Alessandro-Grados. Using some bent-tip needle nose pliers (probably easier to grasp the buttons than normal pliers) in a twisting motion should be safer.

Note: Grado may now be gluing the button to the grill. Soften the glue by applying hot air from a hair dryer before carefully twisting/lifting off the button. If the glue is not softened first, attempting to remove the button could destroy the grill.

An Earcup Facelift for the Alessandro-Grado MS-1

Method (TimD): I don’t like the cheap looking silver paint on plastic decals. I also am apathetic to the bump texture as well. This mod came to me in accident, I actually tried repainting the decals, screwed up pretty bad, decided to sand the thing down to a smooth black beveled face, and said, “I like that look!”

grado_mod7.jpg

This mod is only for plastic versions of Grados. You’d be insane to even attempt to harm the wooden or metal ones. First, put tape over the grill so the shavings don’t ever get inside. I used different grades of sand paper, twisting and turning the front side of the Grado over it until it became a smooth black surface (using rougher to really fine sand paper until you are left with a nice thin black surface with no decals). It is a very easy process, however OBVIOUSLY it is NOT reversible.

An Earcup Facelift for the Grado SR-325

grado_mod14.jpg

Method (M Rael): I removed the button the SR325 and took the black paint off the driver vent screen. I changed the cord so it terminates in a mini jack (and made it shorter), and lastly, I took the black paint off the driver housings. The driver housing needed to come apart in order to take the paint off the screen. I took the screen entirely off and used lacquer thinner and a small wire brush – this also removes the blob of glue underneath the button.

My SR-325’s were all black when new. The driver housings are aluminum. To get the paint off the housings, I tried a few things. The best was a sanding drum attached to a Dremel tool. With a steady hand, it’s actually pretty easy. The remaining paint came off with a fine metal file. It’s worth noting that the screen and the housing could now be sprayed virtually any color at this point. I like the natural metal look; it reminds me of the dearly departed HP-1.

Padding the Grado Headband (MS-1)

Method (Tim D): The paper thin vinyl of the Grado headband comes off as “cheap” to me. You have two big earcups connected by paperthin vinyl and metal wiring – looks weird.

I glued the rubber cushion from the headband of my Sennheiser HD490 to the underside of the vinyl strip of the Grado headband. It makes the headband look much beefier and fits better. The black coloring of the rubber headband is PERFECT and matches perfectly with the black coloring of the Grado. Plus, it makes the Grados seem light as a feather on top of your head. The HD490 headband and cushion can be ordered as a replacement part for $8.60 from Sennheiser USA.

This mod may or may not be reversible depending on the way you “glue” the cushion to the band. I used a thin layer of acrylic caulking. When the caulking cures, it becomes clear, barely noticeable unless under really close inspection. The cushion is bonded well to the bottom of the vinyl. I never tried undoing this modification myself, but I should be able to pull off the cushion and caulking on the vinyl should the need arise. The caulking is very sticky, but it doesn’t make a HARD bond, so pulling off the caulking should be easy.

I thought about using the entire Sennheiser headband, but decided not to, although it is also very easy to do. However I couldn’t think of a way of removing the vinyl headband that wouldn’t be “permanent” and didn’t want to fiddle with the metal attaching to the plastic pieces. Also I am not brave enough to fiddle around with cords and wiring. Besides, the rubber headband pad of the HD490 fits very well right under the existing vinyl. Someone who never saw a Grado before would not think I modded my phones.

New Earcups for the Grado SR-80

grado_mod20.jpg

Mike Lin replaced the earcups of his SR-80 with larger earcups comparable in size to the ones on the Grado RS-2 headphones. He first disassembled the original earcups using Mole’s method. The new earcups were constructed from three parts cut from pvc piping and aluminum screening: a rear chamber, a vent screen to replace the drilled plastic plate in the back of the original earcup and a small ring glued to one end of the chamber to hold the rear vent screen. The chambers are made from 1.5-inch repair PVC couplings and the rings are made with 1.5-inch schedule 40 PVC pipe.

The larger rear chambers were machined from 1.5″ PVC Repair couplings. Standard PVC couplings could be used but they have a tapered inner profile, contain middle stop lips, and their outer surfaces are generally less uniform.

The 1.5″ coupling is what this particular size of coupling (an adaptor to join two pieces of pipes together) is called. It is what one would ask for in the plumbing store to obtain this item. But the 1.5″ does not correspond to either the inside or the outside dimension exactly. By plumbing industry standards, 1.5″ pipes would actually have an outer diameter (OD) of approximately 1-7/8.

grado_mod21.gif
grado_mod22.gif

The new rear chamber dimensions are: 1-29/32″ in diameter, 11/16″ from the transducer housing (the outer part that contain the four notches) to the aluminum screen. The 1.5″ PVC repair couplings were first shortened using a miter saw, and the resulting cut-ends were hand sanded square.

grado_mod23.jpg

A router fitted with a home-made jig was used to enlarge the inside diameter of the cut-ends to fit the transducer housing. The distance between the two bolt heads determines the routed diameter. Small adjustments are made by rotating the heads of the bolts (ie, coupling contacting the angle part of the hex head will result in smaller diameter compared to coupling contacting the edge part of the hex head).

One other way to do this is to use a lathe, but if I had access to a lathe, I would have made the chambers out of mahogany wood. Anyways, mounting the PVC couplings/pipe to the lathe could get complicated. I considered using a sanding drum attached to a drill press, but a router bit cuts much faster and much more precisely.

People may attempt to hand-shape the coupling using tools like Dremel tool. I highly recommend AGAINST this. The cuts need to be exactly square, and the enlarged hole needs to be nearly perfectly round. In addition, the cuts need to be precisely reproduced in the two chambers; any difference would introduce misbalance between the L and R transducers. Unless a person has super-human steady hands, hand-shaping will produce inferior results.

On the other hand, a Dremel tool mounted to a stationary stand could be used if a similar jig is employed.

grado_mod24.jpg

In the picture above, the white circles represent the coupling and the pipe viewed from the end. One of the bolts needs to be moved to accommodate the different diameters. In each case, the bolts are spaced apart so that the router bit would route away the appropriate amount of material from the inside. One would simply hand-turn the coupling/pipe against the two bolts (while the router is running) so that the router bit would cut away the entire inner diameter, enlarging it.

Fine adjustments were made by two steps: I made the bolt’s holes slightly larger so that I could slide the bolts around a little; when the approximate position is reached, the bolt’s correponding nut on the other side of the wood (masonite) board is then tightened to secure the bolt head. Further fine adjustments could be made by rotating the bolt heads – an angle of the bolt head contacting the coupling/pipe would push the coupling/pipe slightly away from the cutting edge of the router bit (compared to if an edge of the bolt head were contacting the coupling/pipe).

grado_mod27.gif

The depth of the cutout is obviously determined by how far the router bit protrudes above the masonite board. This was set at 8.8mm for the couplings (for the rear chambers), and about 1″ for the pipe (for the lip rings discussed below). With 8.8mm-deep cuts in the rear chambers, enough of transducer housing is exposed to accept the foam ear pads.

I took great care to make sure both chambers had the same portion routed – this is critical to ensure that the chambers are the same size. The chamber is routed so the opening provides a tight fit with the transducer housing; no adhesives are used to attach the two together.

grado_mod25.gif

To make the rings, a small section of 1.5″ schedule-40 PVC tube was routed to create a section of pipe having ~3/16 inch thickness, and two rings were cut with a miter saw from this section of thinned pipe. These rings are sanded squared and glued (PVC glue) to the not-yet machined ends of each of the couplings to create lips that will accept the aluminum screens. The rings can be cut with a coping saw or hack saw – just be careful, because the rings are thin and can easily break. Hand-cutting will also require more sanding to get them square and to match the left and right rings.

grado_mod26.gif

The resulting lip ends were then sanded square, and an angle was routed into the inner and outer edges prior to securing the screens using Superglue. Appropriate sized cut-outs and holes were introduced into the resulting rear chambers to accept the mounting pins and wires. For the rear screen, I used aluminum window screening available at most hardware stores (probably similar to the screening described in Skippy’s mod).

Since I don’t have access to equipment that measure frequency response, and since this mod is readily reversible (no glue used), I performed some listening tests with the aid of a MCM Electronics diagnostic test CD. This CD (MCM 80-815) contains a frequency sweep from 5Hz to 20KHz (-15dB). The tests were performed with Sony Discman D-25; a stereo to mono adaptor jack was used to ensure that both transducers were driven by the same channel (to eliminate any difference between the Discman’s R and L channels). The tests were the following:

1. sweep with OEM chambers
2. sweep with new chambers
3. sweep with OEM chamber on L and new chamber on R
4. sweep with new chamber on L and OEM chamber on R

Tests 1 and 2: during both of these tests, the tone localized in my head slightly deviated left and right from the center, indicating driver mismatch. There is a gross mismatch at low frequency (mid 30 Hz, I think) and the rest of the more slight mismatches occurs mostly in high frequencies (higher than 5KHz).

Now the peaks (resonances): Both tests 1 and 2 revealed a significant peak (resonance) somewhere around 3-5KHz, and a couple of smaller peaks are I think around 8-10KHz. My impression is that the higher peaks (the ones around 8-10KHz) are slightly less noticeable with the new chambers. Keep in mind that I could not do a blind test (I always knew whether I was listening to the new or the OEM chambers), so I cannot rule out subjective bias.

Tests 3 and 4: these tests indicate that the new chambers changed the sound at the higher frequencies (above 3-5KHz I estimate). The tone deviated slightly more dramatically from the center (compared to tests 1 and 2), and the deviation occurred in both directions (moved back and forth 4 times) in both tests.

My conclusion is that the new chambers sound different, but do not have any significant peaks or resonances. The modified headphone sounds more open and has less midrange coloration. The bass is subtely enhanced and now extends slightly deeper.

The Ear Cushion Mods

Over the years, there have been three types of Grado ear cushions: the original doughnut-shaped flat cushions, the “comfy” cushions on the SR-60 and the bowl-shaped cushions on the other Grado models. The shape of the cushions has an effect on the sound quality.

Converting the Bowl-Shaped Pads to the Old-Style Doughnut Pads

Method 1 (Beagle): The old style pads have been discontinued. By trimming off the cupped area of the new bowl-shaped pads, the driver moves closer to the ear and back comes the warmth, bass power and smoother top end.

Make the old pads out of the new bowl-shaped ones by cutting the pad to a 1/4″ thickness (from the driver). You end up with a basically flat pad that pretty much simulates the old one physically and sonically. Take a very sharp knife (or better yet, a blade from a disposable razor) and glue it to a popsicle stick (keeping the sharp side outside the surface of the stick). You may need two blades, as one might not be sharp enough after you do the first pad.

Either draw a line along the outside of the pad (at the thickness you would like – you’d probably be trimming about 3/8″ thickness off the top) or put a strip of thick tape such as duct tape around the outside. This will act as your guide.

Keeping the blade flat, VERY slowly “saw” the blade along the top (face up) side of the pad, shaving off the upper layer of the pad, moving forward, slowly along to keep the cutting level. If you go SLOWLY, and keep the blade level with the top surface edge of the inner hole and the outside line (or tape edge) you should be able to make a fairly clean, even cut. When you come around to reach your starting point, you should have a pad similar to the old style, but a lot sturdier.

WARNING: You should ONLY attempt this if you are desperately seeking the older pads and feel that you are up to the task. THIS IS BASICALLY A ONE-SHOT ATTEMPT! If you bungle the job, the pad is ruined and you have to buy a new pair.

Method 2 (Skippy): When I sliced up bowls, I used a Ginsu knife – it made quick work of that pad. Take time to mark off where you are cutting to avoid cutting the pad too thin. I’ve seen the original doughnut pads, and they look very similar to the sliced up bowls. They’re made of the same material and the shape is practically the same.

Some people find the old doughnut-shaped pads too dark sounding, and some people find the new bowls too bright. Slicing up the bowl pads a bit thicker than the old doughnuts might be a good compromise in sound.

Mods and Substitutes for the Grado “Comfy” Pads

Method 1 (Kronsteen): Making a hole in the center of the Comfy pads to get more detailed sound: I had great success with the “coin” technique. I used a nickel for mine and it worked perfectly (just the right size). Just take your Grado pad, line up the nickel with the center of the pad and hold it down firmly with your thumb while carefully cutting around the nickel’s edge with an x-acto (sp?) knife. My pads now look great – as if they were designed that way. I recently upgraded from the Grado SR-60 to the Grado RS-1 (ten times the price) and (I hate to confess this) my biggest impression is just how good the SR60’s are. Give them a chance.

Method 2 (TimD): The Radio Shack RS 33-379 replacement cushions ($2 US) are cheaper than Grado’s Comfy cushions for the SR-60 ($10 US). They have a smaller footprint than the Grado’s (3/4th the thickness of comfy’s and are a smaller diameter) but they look much better (IMO), and you wouldn’t notice that you were using non-Grado pads at first glance, since they fit perfectly.

Method 3 (Strap): The replacement cushions for the Sennheiser HD410 are available at Sennheiser USA for about $3 US. In terms of comfort, the Senns are softer, not scratchy, tapered, the inner rim diameter a bit smaller than the outer (2-3/4″ vs. 3″). Because they are softer, they sit closer to the ears. The central aperture is about 1-5/8″ for the Senns and 2″ for the Grados.

Circumaural Earpads for Grados

Method 1 (Voyager): The sock pads for Grados are the best compromise between sound and comfort. Its not really a compromise though, because they are much more comfortable than the original pads, being circumaural on my ears, and not changing the sound in any profound way. I used 100% cotton socks for this. You must have bowl pads for this to work exactly as described and it will not hurt them in any way. (I have not tried them with other pads, but the bowl pads seem as ideal as possible) If you do not like the new pads, simply remove them and you are only out a pair of socks.

grado_mod8.jpg

The first thing to do is get a pair of clean tube socks. Then cut the end off of each sock, leaving a tube that is 10 inches (I think about 25.4 centimeters for you metric people) in legnth. The next thing to do is flip the 2 tubes inside out. This will make it so that the nice outside is seen when you are done with the pads. Then fold the cut end over the rest of the sock about 1 1/4inches (3.175 cm). Repeat the folding process until the entire sock is rolled up. When you are finished, try to not have any slack left over at the end. If you do it’s ok, but it is neater looking without the slack.

Stretch the pads out and over the bowl pads so that they both start at the same point in the rear and the sock pads extend abot 1/2 inches (1.27cm)above the bowl pads. It may not look good at first, but with fine adjustments they can actually look like the belong there. In my opinion they are more comfortable than all of the Beyers and Senns I’ve tried (HD545, HD565, HD570, HD600, DT330, DT831, DT990) because they don’t press as hard against the head, and still 99%, if not all, of the Grado sound is intact.

Neruda’s comments: First off, less is more when wrapping these things it seems. The more you use the denser the pads are and the harder they become. The less you use the softer the pads are. Also, the more you use the darker they sound. If you try hard and decrease the length of the sock in small increments, you can find something of a “sweetspot” where the sound is detailed but not too bright (as some complain about) and the pads are comfortable. When unrolled, the tubes I used are only four inches in length. Also, the ones I made are not nearly as big as the ones voyager made as far as diameter goes. mine are closer to the size of the original pads, whereas his are much larger. All in all, I’m very pleased with what i’ve got so far. (See below for Neruda’s update.)

grado_mod10.jpg

Chych’s comments: Just did the mod myself. I rolled a 4″ cotton sock and put it over my sliced bowl pads. It was a bit of a pain to put it over the sliced bowl pads, since they are a bit unstable and can’t take the pad pressure that well. Anyways the sound is much darker, more emphasis on the bass, richer and deeper soundstage, more involving and all, BUT I can’t hear mid-highs, and voices are a tad bit distant. Also the bass can be slightly muddy – just try and listen to them with mega bass on full. They are probably like unmodded ex70s… makes you dizzy quite fast. I’ll have to try a different material (this is 4″ of thick cotton black sock, looks like the HD590 pads).

Method 2 (Neruda): I just did the mod again, by wrapping the sock around the bowl pad. I cut it so that it wouldn’t even wrap around the entire pad once – there’s still a bare section where the pad connects with the earcup to improve the fit of the pads. Whoah! Even more comfortable! They look even better now and look like they were really meant to be on my SR80’s. Of course, my pads are perfectly round. Also, they’ve gotten thick enough so that they’re almost circumaural on me. Even though they’re not quite, they’re still one of the most comfortable headphones I’ve ever had on. I was using my pads all day yesterday in the sun, and my ears never heated up. It seems the socks keep nice and cool, for some reason. A nice plus, since the foam pads always got my ears hot.

Method 3 (Squirt): As a variation of the Voyager sock mod, I tried taking off the ear pads from my old AKG 240 and placing them over the donut pads. It works ok, but you will need to use two-sided tape or something to hold the AKG pads in place, or just leave them as is and just use the spring pressure to hold them in place on your head.

grado_mod11.jpg

I was shocked to find how much this widens the sound stage and also improves the high frequency details. I’m hearing high frequency details i’ve never heard with the RS-1 before and sounds at the outer edges of the sound stage now sound like they are outside your head. You lose some bass and mid-range impact though i’ve felt my RS-1 are a little too punchy and bassy anyway.

grado_mod12.jpg

Method 4 (Chych): I glued the vinyl (or plastic, some material I’m not too sure about but were somewhat comfortable) earpads from some freebie or cheap headphone I found in my basement onto the Grado pads. I just used a few drops of super glue. The adhesion probably isn’t too strong; I kept it this way in case I wanted to change the pads once again.

The Headphone Cable Mods

Converting Standard Grados into Dual-Mono Grados (SR-80)

This mod comes from Kevin Gilmore’s article for A Pure Class A Dynamic Headphone Amplifier, which describes an option for a bridged output. Some commercial headphone amps, such as HeadRoom’s BlockHead, have bridged outputs. Bridged amps require separate ground connections to each transducer, instead of the common ground of standard headphone cables.

gilmore3_11.jpg

Method (Kevin Gilmore): Dual mono Grado SR-80s have more and tighter bass when driven with the bridged dynamic amplifier, which can output twice the voltage, twice the current (not at the same time) and double the slew rate, giving better control of the diaphragm and a higher damping factor.

I cut apart a pair of Grado SR-80s. I had to take them apart anyway – one side was damaged by my toy terriers. There are two ways to wire Grados for dual mono operation: the easy way and the harder way. The easy way is to cut the wires just above the “Y” of the headphone cord. Carefully splice new mono connections to each of the wires.

The hard way is to take the headphones completely apart and rewire all the way to the transducers.

  1. Take the transducers out of the headband.
  2. Take the ear pads off.
  3. Very carefully, with a sharp jeweler’s screwdriver, pry around the side seal. It is superglued in only a couple of spots. Comes apart real easy.

    gilmore3_10.gif

  4. Buy a set of 20-ft, 1/4″ headphone jack extender cables from Radio Shack. Cut one end off at the length you want and then solder to the transducer. I use the tip and sleeve part of a stereo 1/4 jack, and leave the ground connection unconnected.
  5. On the amplifier, use 2 separate stereo 1/4″ headphone jacks (one for each channel) and wire the tip and sleeve to match the headphone plugs. The ground sleeve IS connected to ground in the headphone jacks only (actually decreases the noise level a bit).

Portions c. 2001 Chu Moy.

4/30/2002: Added tips by Mole and M Rael.

6/22/2002: Added MLin’s replacement earcups for the SR-80.

1/21/2004: From forum member Reeseboise, another technique to fix the Grado “grattle” problem:

I was reading the page of Grado mods not more than a few minutes ago looking for the easiest way to fix the low-frequency rattle in my SR60s, when I realized that all that was suggested could be accomplished by much simpler means. So, I had an idea: I sucked on it. Now apparently this “Grado Grattle” is caused by a small crease in the diaphragm, and what was suggested to fix it involved removing the drivers, preforming some careful duct tape artistry to correct the crease by pulling it outwards, and then putting the drivers back in, all of this without damaging something (obviously). So, I figured, if the basic premise of the fix was to pull out the crease, why not do so with air pressure? I removed the pad from the offending earcup, put my mouth on the mesh, and VERY gently breathed in. Heard a little snap, got a hair dryer and “cured” the diaphragm, as per Beagles instructions, performed a couple of tests, and now they work like champs. Massive Attack is once more enjoyable.

Well, I can confirm that this method actually works for sure now, since I just did the same thing to the right earcup (it was the left before), which got the Grattle pretty badly today, after some very loud listening sessions with very bass intensive music. Now, it sounds grand.

1/16/2004: Forum member Hollowhead submitted this idea for trimming Grado ear cushions so that they’re more comfortable:

I stumbled across the article on Grado mods, and found it to be fairly interesting. The primary thing I want to change about mine is their comfort, or rather uncomfort. I read the section re: trimming headphone pads, and just have a little technique to share that I used when constructing prototypes for industrial design projects. I’m describing the basic concept, you’ll have to make it work for your particular pad.

For trimming them to a thinner profile, that is, getting closer to the diaphragm:

  • Attach the headphone pad to some sort of circular item, with double stick tape, carpet tape (heavy duty white foam double stick tape), anything to make it snug enough to resist a little tug. A plastic lid, tuna fish can, etc..
  • Drill a hole near the center of the can, lid, etc.. it doesn’t have to be dead center, but the closer the better.
  • I would use a utillity knife razor blade (i.e. Stanley), or replacement blade for a “snap-off” knife (Stanley makes these too). Square razor blades aren’t very big, and the extra thickness at the back can be annoying.
  • Set the headphone pad flat on a table, such that the pad is facing up, and the disk thingy you’ve put in the center is contacting the table, flat. If the disk thingy isn’t tall enough, find a new disk thingy.
  • Use anything you can get your hands on to make a small block, to which you will tape the utility blade. Something shaped similar to a pack of cigarettes would be pretty decent. It can be a small piece of wood, plastic, etc.. but should be flat, and not shift around much if it’s made of several layers.
  • Tape the blade to the block. When the block is set flat on the table, the blade should be touching the pad at the height you need them trimmed to.
  • Hold the pad steady, and rotate the block around the pad, until you have to rotate the pad and block together, to continue the motion. The blade should probably be close to tangent to the earphone pad, meaning, let it slice the foam, as though you were cutting a tomato, not cheesecake.
  • WATCH OUT for your hand holding the pad.. the blade is headed straight towards you! To minimize the chance of getting cut, use a NEW, SHARP, blade, and try to never let your rotating hand be in a straight line with the blade. Imagine it suddenly slipping, it should pass by your hand without coming near.

Depending on the density of the foam, this whole procedure will be more or less difficult. Snap off blade replacements will cut wider pads, but may tend to “swim” in the foam, and make imperfect cuts. Use as short an amount of the blade extended off the block as in possible. This is also safer.

You can place a shim under the the blade to create angled cuts, or make multiple cuts at different angles to make a contoured pad surface.

I hope someone finds this useful!

6/2/2004: Forum member Shell submits this creative mod that makes the cord removable and allows the Grado SR80 to operate in regular stereo or dual mono mode. The headphone cable is removed and two mini jacks (with switches) are installed on each earcup. The left channel jack can receive either a standard stereo signal or a mono signal. If a mono signal is then inserted into the right channel jack, it plays only that mono signal (the right channel information from the stereo signal is cut off).

I recently got me a pair of Grado SR-80s. This is, technically, my first pair of high-end headphones, and i must say the 80-somehting dollars that they cost me were well-spent. These things are just begging to be modded though. [And yes, I’ve read the Collected Grado Mods page] First off, I plan on “velvetizing” the earpads, and perhaps putting a very thin velvet fabric over the white fabricky mesh of the headphone, with the circular holes matched. Next is the cable. The original cable is a bit too long and bulky for me, and the conector is a bit bulky as well. However, rather than doing just a mono-plug mod, I decided to do something a bit more useful.

shell_grado2.jpg

I ordered a pair of 3.5 mm stereo jacks with 2 NC (normally-closed) switches, along with a bunch of other stuff. I put each of these in each of the headphones so that I have both the capability to use them as 2 mono plugs and one Stereo plug coming from only one side.

shell_grado1.gif

How? This is where the 2 NC switches come in. The right headphone’s jack is wired through the NC switches to the left headphone’s stereo jack (the wires go through the headband). This way, I can plug in a stereo cable to the left headphone’s jack to use the headphones, and if i plug in a mono (or stereo, doesn’t matter – as long as I plug something in) connector into the right headphone, the NC switches open and I’ve got 2 separate headphones with separate ground.

The first step was to take apart the driver housing. This was actually very easy with the spoon method discussed on the collected grado mods page. The cable is desoldered from the transducer. Then the jack is mounted on the earcup in the hole where the cable used to be. There was a lot of filing involved to get the jacks to fit in the earcups. To connect the right channel, I used about 15″ of 20AWG wire (way too thick – overkill, really).

shell_grado1.jpg
shell_grado3.jpg
shell_grado4.jpg

Installing the jack for the left channel is done similarly. After I had finished with the right channel, I “threaded” the wire inside the headband so it came out on the other side. Then I connected the right channel to the stereo jack in the left channel.

After putting everything together, everything worked except the right headphone in mono mode. The reason? I thought i could get away without soldering the center band (“right channel” of the jack) and the ground together, but apparently the contact didn’t reach ground on a mono plug. So, insert small wire connecting the center band and ground. Voila: I have, if I want, two completely separated headphones (they don’t even share common ground), or a stereo cable in the left headphone for both of them.

shell_grado7.jpg
shell_grado5.jpg

8/10/2004: Forum member Wallijonn transplanted the tranducers from his Grado SR-80 to an Aiwa AK100 headset assembly. He writes:

walli1

found the Aiwa AK100 on Ebay, and I paid $30 for two. The headband is light flexible plastic, mimicking the AKG K501 and ATH AD10 type of headband – dual plastic covered wires. It has very comfortable over-the-ear earpads and cloth-like (top) covered leather (bottom) headband tensioner. Even though they are of an open design, the middle of the transducer leaks very little sound. If you know what a dome tweeter sounds like, you should have some idea of what these headphones sound like. The cable is Sennheiser thickness like and is 15 feet long, making it perfect for television and movie viewing. I used the Aiwa AK-100 headphone for a basis and used Beyer pads. I made two aluminum donuts 3 & 31/32″ inch in diameter and made a 2″ hole in the centre. I turned the finish into brushed aluminum. Black & Silver. GORGEOUS. I unsoldered the transducers and mounted the transducers in the aluminum hole. I then resoldered the wires. I cut the Grado pads so that the angled portion of the bowl fits inside the Beyer DT831 pads. The sound: SOOO MUCH BASS! This thing shakes when it’s playing bass notes. I’m wondering if adding cotton batting to the cup side will tame it. I doubt it as cotton batting tends to increase volume space and therefore increasing bass, but I may be able to push it towards an infinite baffle type enclosure. This thing has to blow away DT770s! The transients are all there, as is the forward sound of the midrange. The highs are super smooth. Maybe it just matches well with the Crown D60 amp, as I have found that the K501 has a LOT of bass, as does the K1000s when I play them through my Crown D60 amp. So let’s chalk this one up to my modifications AND the Crown amp. Phase 2 is the tuning. Phase 3,The final phase, would be the headband. I’m still undecided if I want to destroy the Aiwa AK100s. It should be easy enough to mount it on new pivot points. The bass is super TIGHT: sub-woofer tight. I can’t believe my ears! My head is shaking from the bass notes. I feel like I’m inside a car with a $5000 stereo – the ghetto blaster woofers being heard from 10 blocks away! HOORAY! Michael Jackson’s “Billie Jean” sounds like it would at Studio 54 on a saturday night! I can feel my ear lobes shaking!!! AND THOSE HIGHS! AND THE MIDS! WOW!!!!!! MAN, OH MAN!!! Kiddies, throw away your dt770s and V6’s. There’s a new king in town. Since the AK-100s have “angled” drivers, I removed the drivers, extended the Grado as far as possible (a little over 9/16″) and put wires over the inner hole (inside the Aiwa ear cup) while I glue gunned all the parts together. In this configuration the Grado driver comes into contact with the ear lobe (as is normal). But now it is angled. It took a few rubber bands to hold everything in place while I glue gunned all the parts together (both the Grado driver housing, and the Grado driver to the Aiwa inner hole). Yes, the wires were removed after the glue dried. I kept the original driver cloth cover as it does not rub as much as stock. The sound? Well, it sounds just like a Grado. (A sound which I have outgrown). But it is a lot more comfortable, thanks mainly to the beyer pads.And some of the high end harshness is gone. I find that the drivers, when they fire directly into your ear canal, are too harsh. I expect the bass and midrange to improve slightly as the pads compress a little more. (What some people mistake for burn in). The wires no longer get tangled. The cable is still too short. It just makes it to my soundcard. And it is way too short for 12′ viewing from a TV set. I definitely like my DT880s a lot more.

walli2.jpg

walli3.jpg

walli4.jpg

The headband is somewhat flimsy as it is flexible plastic. I took off the wimpy pads and put on the Beyer DT831 pads on it. It would be interesting if other headphones can be adapted. I would think that any headphone which presently has angled drivers could possible work. The trick will be to see if a closeout item can be adapted. I would wait until they go for $15.

The Psychoacoustics of Headphone Listening.

by HeadRoom Corporation

Listening to music on headphones is different, and acoustically less satisfying, than listening to speakers. The reason for this has to do with a human’s ability to locate sounds in space. This ability is described by the science of psychoacoustics, and specifically by the set of formulas called “Head Related Transfer Functions” (HRTF).

hdrm1.gif

For the sake of our argument, we will use the special condition of two speakers set 30-degrees to either side of the listener.

hdrm2.gif

Now let’s say you turn off the left speaker. Both ears will continue to hear the right speaker. Because the left ear is slightly farther away than the right ear, it hears the sound slightly later. This time difference between ears is called the inter-aural time difference (ITD), and is the primary cue with which your brain figures out left-to-right position of sounds.

Now imagine you are wearing a pair of headphones, and you turn off the left channel. Only the right ear hears the right channel. The left ear hears nothing. This never happens in nature (except maybe if you have a fly buzzing in your ear ), and leads to that annoying blobs-in-your-head audio image normally associated with headphone listening. The reason it’s annoying is because your brain doesn’t have enough information to correctly localize sounds, and it keeps struggling to figure things out. In the end (usually after about two hours), your brain starts screaming, “Let me outta here!”, and your headphone istening is done for the day.

Enter HeadRoom. We employ as little circuitry as possible in an attempt to fix the localization problem without mucking up the signal. We won’t take all the credit; a guy named Ben Bauer figured a lot of this stuff out way back in the 60’s. He came up with a basic theory of time delays and other factors to compensate for headphone listening.

hdrm3.gif

The state of the art in electronics today permits the effective design of Bauer’s headphone compensation circuit. All HeadRoom amps have such a circuit. A little bit of each channel is crossfed through a time delay over to the other channel. This circuit, which we call our Audio Image Processor models the basic inter-aural time differences found when listening to a pair of speakers.

Although the audio image remains inside your head, it now has a sense of continuity from left to right. The most common customer comment is that the music just sounds more natural with HeadRoom. The correct time delays are heard by the ears, and your brain is able to process the audio information in a normal manner.

But what about something more complicated like surround sound? For that you would need a five channel processor. On pages 32 and 33 of this catalog you will see home theater products called the Auri and the Lucas. These are much more complex digital processors that perform a five channel head related transfer function mapping to headphones.

These processors take into account inter-aural time differences between ears for five channels. They also calculate acoustic properties for elevation and add room acoustic reverberances. All of these parameters are calculated for common head sizes and shapes. The resulting audio presentation remains in (or slightly outside) your head and, for home theater applications, is quite an involving and effective synthetic acoustic environment. Unfortunately, because of the extraordinary amount of processing, we feel these amps lack the precision and clarity needed for excellent music reproduction.

You can stop reading right here if you’re not interested in the technical details associated with the HeadRoom Audio Image Processor.

There are a number of other phenomena that the brain uses to aid in localization. One of these is pinna reflections. The ear nearest a sound source shows increased high frequency response. Reflections off the outer part of the ear (the pinna) cause comb filter effects that aid in the determination of elevation.

Reflections off the walls of the room and foreknowledge of the difference between the way various floor coverings sound also contribute to localization.

The most important cue of all is the way all these factors change as your head moves relative to sound sources. It was found in experiments that when a subject’s head is clamped, the number of times a sound coming from the front was mistaken as coming from behind went up by a factor of ten. It was also found in homing experiments that the blindfolded subjects would precisely guide themselves to sound sources.

Still with us? OK, here goes with the numbers: With the speaker 30 degrees off axis, you get about a 300µs delay between left and right ears.

hdrm4.gif

There is an amplitude difference between ears, too. The far ear is in the shadow of the head, so it hears the sound at a slightly lower volume.

hdrm5

This amplitude difference also varies with frequency. Though the lows are not attenuated by the head much, the highs begin to roll off rather quickly around 2kHz.

Then a funny thing happens: the highs start reaching the far ear by traveling along the skin surface. This effect causes a broad rise centered around 5kHz. The result is a hump in the far ear curve response.

The near ear experiences better high frequency sensitivity, and hence sees a smoothly rising frequence responce curve.

The described phenomena are the primary left-right localization cues that the HeadRoom Audio Image Processor tries to model, although it accomplishes it in a completely different way.

hdrm6.gif

The HeadRoom Audio Image Proccessor circuit consists simply of two-stage active filter which controls both the Frequency response and phase of the crossfeed circuit. The complex phase delay this circuit introduces creates a delay very similar to the ITD up to about 2 kHz.Because ITD information becomes less meaningful above 2 kHz, the circuit also rolls frequency responce off above that point. The crossfed signal is simply summed in with the opposite channel at a 10 dB lower amplitude.

Because of the strange and complex way these two signals sum together, the frequency response of the HeadRoom process reasonably matches the hump in the far ear response curve.

When a signal is delayed then sum it back on itself, a cancelation results every time the signal is 180 degrees out of phase. This causes a series of notches up the frequency response curve which is commonly called a comb filter. In the HeadRoom processor, we take a bit of right channel, delay it, then sum it with the left. The resulting comb filter effect is weak because we sum the delayed signal in at a lower level and with rolled off highs. Because the delay has its first 180 degree out point at 2.5 kHz, the HeadRoom processed mono signal has a gentle dip centered at 2.5 kHz.

As a result of the factors described above, there is an overall warming of the mono signal. The Filter switch on the front of HeadRoom amps is designed to compensate for this warming by slightly boosting the high frequencies. This boost is roughly equivalent to the near ear response.

The real question is, “What does all this sound like?” You’ll notice a number of things when you listen to a HeadRoom amp. First, the processor is very subtle. Because your ears can’t move relative to the sound sources in headphones, it takes a long time for your brain to “learn” this new and synthetic acoustic environment. However, you will notice right away that you can listen for much longer periods of time without any listening fatigue.

After about 40 hours of listening, your brain will have learned the new acoustic trick and will become adept at localizing sounds very much like with a pair of speakers. There is a smooth left-to-right continuity, and the audio image will have reasonably good depth. HeadRoom amps simply give your brain a more natural headphone listening experience

c. 1998, HeadRoom Corporation.
From HeadRoom Corporation. (Republished with permission.)

 

Understanding Headphone Power Requirements.

By Dennis Bohn, Rane Corporation

Much confusion abounds regarding headphone power requirements. This RaneNote is intended to disperse some of the mist surrounding headphone specifications and hopefully give you a clearer understanding of how much power is really needed for your application.

HEADPHONE SENSITIVITY

Headphone manufacturers specify a sensitivity rating for their products that is very similar to loudspeaker sensitivity ratings. For loudspeakers, the standard is to apply 1 watt and then measure the sound pressure level (SPL) at a distance of 1 meter. For headphones, the standard is to apply 1 milliwatt (1 mW = 1/1000 of a watt) and then measure the sound pressure level at the earpiece (using a dummy head with built-in microphones). Sensitivity is then stated as the number of dB of actual sound level (SPL) produced by the headphones with 1 mW of input; headphone specifications commonly refer to this by the misleading term dB/mW. What they really mean is dB SPL for 1 mW input.

Think about these sensitivity definitions a moment: headphone sensitivity is rated using 1/1000 of a watt; loudspeaker sensitivity is rated using 1 watt. So a quick rule-of-thumb is that you are going to need about 1/1000 as much power to drive your headphones as to drive your loudspeakers since both of their sensitivity ratings are similar (around 90-110 dB-SPL). For example, if your hi-fi amp is rated at 65 watts, then you would need only 65 mW to drive comparable headphones. (Actually you need less than 65 mW since most people don’t listen to their loudspeakers at 1 meter.) And this is exactly what you find in hi-fi receivers. Their headphone jacks typically provide only 10-20 mW of output Power.

Take another moment and think about all those portable tape players. Ever hear one? They sound great, and loud. Why you can even hear the headphones ten feet away as the teenage skateboarder that ran over your foot escapes.

Power output? About 12 mW.

THE LIST

As an aid in finding out how much power is available from the MH 4 Headphone Console, we have compiled a listing of popular headphones. Included is a column giving the maximum SPL obtainable using the MH 4 and any particular headphone. Ultimately, it all gets down to actual SPL. The power rating really doesn’t matter at all. Either it’s loud enough or it isn’t (of course it has to be clean power, not clipped and distorted). The SPL numbers shown are for maximum continuous SPL; for momentary peak SPL add 3 dB.

Note that the maximum achievable SPL varies widely for different models and manufacturers, ranging from a low of 107 dB to a harmful 146 dB! The table also shows there is very little relationship between headphone impedance and sensitivity, and that power output alone means nothing, since in one case 80 mW produces a maximum SPL of 107 dB, yet in another case the same 80 mW yields an SPL of 124 dB!

Sensitivity dB is measured sound pressure level with 1mW of power. The Max Power mW columns are typical continuous average (RMS) power, 20 Hz-20 kHz, with THD less than .4%.

If headphones are not yet owned, or replacements are desired, use this listing as a guide for selecting headphones with sufficient sensitivity for the maximum desired SPL.

Table of Headphone Specifications

Disclaimer: The headphone specifications were supplied to us by the respective manufacturers, subject to change without notice. 

Manufacturer Model Impedance
(Ohms)
Sensitivity
(dB)
Max Power
(mW)
Max SPL
(dB)
AKG K141M 600 98 80 117
K240M, K240DF 600 88 80 107
K270S 75 92 380 118
K301 100 94 285 119
K401, K501 120 94 290 119
Audio-Technica ATH-COM1, COM2, ATH-908 40 90 440 116
ATH-910 40 92 440 118
ATH-P5 40 100 440 126
ATH-M40 60 100 400 126
ATH-D40 66 102 295 127
ATH-M2X, ATH-M3X 45 100 435 126
Beyerdynamic DT150 250 97 175 119
DT211, DT311 40 98 440 124
DT250 80 98 360 123
DT411 250 102 175 124
DT531 250 95 175 116
DT431, DT331 40 86 440 112
DT770PRO, DT990PRO 600 96 80 115
DT801, DT811,DT511 250 94 175 116
DT901, DT911 250 98 175 120
Fostex T-5 44 96 435 122
T-7 70 98 385 124
T-20 50 96 425 122
T-40 50 98 425 124
Hosa HDS-701 40 91 440 117
Koss Headphones A/250, A/200, A/130, TD/80 60 98 320 125
R/200 60 84 400 110
R/100, R/45 60 85 400 111
R/90, HD/2, SB/15 60 100 400 126
R/80, R/35S, R/20, Porta Pro models 60 101 400 127
R/70B, R/55B, SB/50, SB/35 60 101 400 127
R/40 60 90 400 116
R/30S 60 106 400 132
R/10 60 103 400 129
TD/75 60 95 400 121
TD/65 90 101 340 126
TD/61 38 93 440 119
Sennheiser HD433, HD435 32 94 450 121
HD25 70 120 380 146
HD445 52 97 390 123
HD25SP 85 100 350 125
HD265, HD525, HD535, HD545, HD565 150 94 190 117
HD455, HD475 60 94 400 120
HD465 100 94 285 119
HD580, HD600 300 97 80 116
Sony MDR-V100MK2 32 98 450 125
MDR-85 40 102 440 128
MDR-V600, MDR-D77 45 106 435 132
MDR-CD10 32 96 450 123
MDR-CD550, CD750 45 100 435 126
MDR-CD6 45 110 435 136
MDR-CD850, CD950 32 102 450 129
MDR-CD1000, CD3000 32 104 450 131
MDR-D33, MDR-D55, MDR-7504 45 104 435 130
MDR-7506 63 106 400 132
MDR-7502 45 102 435 128
Stanton ST PRO, DJ PRO 1000 32 100 450 127
Telex PH-6 600 105 80 124
Yamaha RH5M 32 98 450 125
RH1 32 90 450 116
RH2 32 95 450 122
RH3 60 95 400 121
RH10M 40 102 440 128
RH40M 32 103 450 130

c. 1983, Rane Corporation

Understanding Headphone Power Requirements PDF version.

c. 1983, Rane Corporation
From Rane Corporation Site. (Republished with permission.)

 

Designing A Pocket Equalizer For Headphone Listening.

eq1

In audiophile circles, equalizers are prescribed only as a last resort to correct imbalances in recordings and room acoustics for loudspeaker playback. The case for using equalizers with headphones is simply this: headphones by their very nature change the tonal balance of music. The acoustic shaping that results from sound interacting with a listener’s head and outer ears varies from person to person and is missing in headphones, which play directly to the eardrums. Therefore, not only does music in headphones have a different tonal balance than was intended, but each listener hears a slightly different presentation. If the listening system includes an headphone virtualizer (such as Dolby Headphone), equalization may help restore the imaging if the headphones have a different frequency response than what the virtualizer expects.

While equalizers for home stereos and studios are available in range of prices and features, are there any options for portable users? Koss Corporation makes an inexpensive, small 3-band equalizer (EQ-30 and EQ-50) for headphones that runs off 2-AAA batteries, but it has received less than favorable reviews (poor filter action and poor headphone drive). Some portable stereos have built-in equalizers, which are mostly inferior gimmicks. These products obviously are not designed with audiophile concerns in mind. The best way to get a high quality portable equalizer in these times of audiophile minimalism is to build one. This article is a collection of opamp-based equalizer designs that are all suitable for portable use (except for one vacuum tube circuit, included for the benefit of glass audio fans). Appendix 1 describes a DIY project: building a mini tilt equalizer/amplifier for headphones. Appendix 2 shows how to use OrCAD circuit simulation software to model and customize equalizers.

EQUALIZATION AND HEADPHONES

biphonic.gif

In addition to correcting tonal imbalances in headphones and recordings, equalizers can, to a certain extent, alter the perceived spatial characteristics of headphones. Headphone acoustic simulators, such as virtualizers, electronically process stereo and multi-channel signals so that they image outside the listener’s head – as though they were being projected by real loudspeakers. When an acoustic simulator is not available, equalization can mimic some aspects of normal hearing. For example, Ron Cole and John Sunier at the Binaural Source recommend a “biophonic” curve (shown above) as a guide to setting equalizers to correct for ear canal resonance and other differences in the spectrum between speaker listening and headphone listening. (See Taking Sound in Another Direction for more information.) The curve suggests boosting three center points and is the basis for this article’s focus on 3-band equalizers.

fletcher.gif

Another reason for equalizing headphone sound has to do with human loudness perception. When listening at safe volume levels, headphone sound may not be as satisfying because the perception of loudness is frequency and volume dependent. As shown in the Fletcher-Munson curves above, human hearing does not become “flat” until the SPLs are at the threshold of pain. Since the curves change shape at different loudness levels, accurate loudness compensation would have to be dynamically adjusted to avoid overcompensation – an unlikely feature in a rack-mounted commercial equalizer, let alone in a hand-built pocket EQ. Nevertheless, a small treble and bass boost can improve the perceived frequency balance of low volume headphone sound.

PASSIVE 2-BAND BAXANDALL

eq2.gif

The passive circuit in figure 1 is a variation of the famous Baxandall circuit. It features a smoothly increasing ±6 dB/octave slope of boost or cut. The bass and treble filters have a shelving response, although the “shelves” are outside the audible range with these component values. The filter equations in figure 1a predict the threshold and shelving frequencies. When the bass control is rotated for maximum boost, the wiper shorts out the .033uf capacitor. R3 and C4 form a frequency dependent voltage divider that determines the shelf frequency of the boost. When the treble control is set for maximum boost, the wiper bypasses the 10K resistor. C1 and R2 form a high pass filter.

eq2a.gif

In addition to simplicity, the circuit uses commonly available parts. Even the 100K log taper pots, which are usually hard to find, are sold by Radio Shack as part number 271-1732 (in a stacked configuration for stereo). At the midway point, the pot will have about 10K on one side and 90K on the other. For the treble control, the side with 10K parallels R2. For the bass control, the side with 10K parallels C4. Note that this circuit could be divided for an individual bass or treble control (R5 may then be omitted – it helps isolate the bass from the treble circuit when the two are put together).

The passive Baxandall must see a low impedance source and drive a high impedance load to avoid loading the network and affecting the response curves. Low impedance sources include the output of a portable stereo or preamp. It can drive a headphone amplifier or high impedance headphones (possibly with a series resistor to increase the load impedance), but typical low power portables may not be able to drive headphones to adequate volume. Since passive networks are not amplifiers, the gain is simulated by dropping the “flat” response 20dB below the input, for which a portable stereo may not be able to compensate.

eq2b.gif

The active version of the 2-band Baxandall shown in figure 1b incorporates the frequency shaping circuit into the feedback loop of the opamp. The feedback network is almost identical to the passive Baxandall. The pots are now linear taper, and the double capacitors in the bass and treble sections have been simplified to one per section. The input stage serves as both an impedance buffer and as a phase correcter, so that the output of the EQ is in phase with the input. The equations for this circuit are shown in figure 1c. Unlike the passive version, this equalizer has true voltage gain: ±10 (±20dB).

eq2c.gif

ACTIVE 3-BAND BAXANDALL

eq3.gif

A 2-band equalizer is limited in its ability to correct trouble spots in headphones. The circuit in figure 2 is an active version of the Baxandall with an adjustable resonant filter acting as a midrange control. Again, the inverting buffer at the input ensures that the equalization stage sees a low source impedance and that the output phase of the equalizer is non-inverting. The total gain/cut for any band is ±20dB.

As with the 2-band version, the treble and bass filters are shelving equalizers (figure 4) – although the “shelves” occur outside the audible range with these component values. A better shelving characteristic can be obtained by moving the shelving frequencies closer together. With small modifications to the center frequencies, this circuit could generate a broad facsimile of the biophonic curve. One important characteristic of the biophonic curve is the “dip” at 7.5kHz to simulate the ear canal resonance of normal hearing. If diffuse-field headphones are used, they probably already have a response curve with the ear canal resonance compensation, and then the EQ will help adjust it for the individual listener.

The treble and mid bands then are the most critical. They must be spaced far enough apart to simulate the response dip at 7.5kHz. Moving the treble shelving frequency higher to 20kHz will help. Because the operating range of the midrange resonant filter overlaps with those of the bass and treble controls, no simple equations can describe how changes in component values will exactly affect the response curves. Applying the equations for the 2-band version to the 3-band, they appear to predict the bass and treble crossover frequencies on the 3-band, but not the gain for the treble control.

eq4.gif

The midrange control parameters must be set by experimentation. These following changes are designed to have minimal impact on the bass and treble responses:

 

  • R4 affects the overall midrange gain/cut. Increasing R4 will reduce gain and shift the midrange center frequency higher. Decreasing R4 will raise the gain and set the center frequency lower.
  • Changing C4 and C5 will shift the midrange center frequency without affecting the bass or treble. To shift the center frequency higher (lower), decrease (increase) C4 and C5 in proportion such that C5 = 5C4. The frequency shift is proportional to the shift in the value of C5: f1/f = C51/C5 – where f1 and C51 are the new values.

 

eq5.gif

This design has the virtue of being a single IC project (with the quad version of the opamp), uses easier-to-find linear-taper pots, and is free of inductors. R1 and R2 are chosen for unity gain, but the circuit can also function as a headphone amplifier, if the R2/R1 feedback ratio is made greater than 1. For example, setting R1 to 10K ohms (and C1 to 10uF) would give the equalizer a “flat” gain of 10. Then a volume control in front of the inverter would be a good idea. In selecting the opamp, it should have a very high slew rate for high bandwidth and be able to run off standard battery voltages.

Most quality, FET-input opamps will work (such as the LF353 from National Semiconductor) when run from two 9-volt batteries or even one 9V battery configured as a virtually grounded dual supply (see figure 10 below). The Burr-Brown OPA132 or OPA134 is especially attractive, with a 20uV/sec slew rate, no phase inversion and a power supply as low as ±2.5VDC. For the active equalizers circuit shown in this article, the recommended battery supply is ±9V. Smaller power supplies will likely cause an equalizer to clip frequently because of the high gains involved.

BANDPASS EQUALIZERS

Although the Baxandall circuits are good for general purposes, there are discriminating headphone connoisseurs who insist on doing the least sonic harm when applying equalization. In particular, the shelving characteristic of bass and treble bands can make recordings sound bass-heavy or excessively noisy (noise is less of a problem when listening to digital recordings). Parametric equalizers have a bandwidth control (also called a Q control) for each band to further narrow the response of a resonant filter. Since each band has a minimum of 3 adjustments (gain, frequency and bandwidth), a 3-band parametric would have at least 9 controls and a barrel of parts and is impractical to house in a truly pocket-sized enclosure.

The bandpass filters of an octave equalizer can be preset with a high Q. A 3-band graphic equalizer with high Q bands is rare, because 3 bands usually do not offer enough flexibility when the filter action is narrow. However, there are situations where a 3-band graphic equalizer might be adequate, such as when the EQ is application specific and the center frequencies are known in advance. Again, the number of parts to build a high Q 3-band equalizer is probably more than can fit into a small enclosure (although there are “graphic EQs on a chip” that might work).

eq6.gif

For the purposes of improving headphone listening, such complexity is not mandatory. The effect of a resonant filter with standard 6 dB/octave slopes (figure 5) is broad only at maximum gain. Thus, the action of the filter can be confined by not setting the gain to extremes.

eq7.gif

The Wien bridge equalizer in figure 6 uses dual resonant filters to isolate the effect of the bass and treble controls to the center frequencies. A voltage divider at the input sets the signal level into the filter which is then varied by feeding back the bandpass response into the differential input of the opamp. The per-band gain is ±9 dB. Each bandpass consists of a pair of RC filters, so the slope of boost or cut is 6 dB/octave on either side. Even when the controls are set at maximum gain or cut (figure 7), the filters have little or no effect on the low, middle and high portions of the audio spectrum.

eq8.gif

Going back to the active Baxandall, that circuit’s midrange resonant filter could be duplicated for the bass and treble bands to target the audio spectrum more specifically. To avoid the problem of overlapping bandpasses complicating the design of a single EQ stage, each band should have its own EQ gain block. Unfortunately, this strategy increases the number ICs (2 quad and 1 dual IC) but the whole circuit should still fit into a pocket-sized enclosure (although the pocket may be a little bit on the large side).

eq9.gif

The 3-band resonant equalizer in figure 8 ensures that the EQ stages see a low source impedance by buffering the input signal via a non-inverting voltage follower. (To make this circuit an amplifier/EQ, change the voltage follower to a non-inverting amplifier.) Each EQ stage (figure 9) is configured with a low, mid or high center frequency and a ±10 dB boost/cut. The outputs from these EQs are then mixed back at unity gain for an overall gain of 3. This mixing stage is another point where gain can be added to the system. Since each EQ stage and the summing stage are inverting, the equalizer output has the correct phase.

eq10

Each EQ stage is a combination of low frequency and high frequency shelving filters. In fact, C1 and C2 can be alternately switched out to restore the circuit’s low or high frequency shelving characteristic. The center frequency and gain of a resonant filter are set by overlapping the responses of the low and high pass filters. To design the filter, first calculate the C1/C2 ratio by solving equation 1 (figure 9) for a given maximum voltage gain (say 3 or about 10dB). Solve for C1 as a proportion of C2 and substitute this equivalent for C1 in equation 2. Then choose a center frequency and solve for C2 in microfarads. The shown C values set the center frequencies suggested by the biophonic curve. BEFORE BUILDING WITH THESE C VALUES, check out the sound with an existing equalizer.

Note that equation 1 for calculating gain does not work if the design attempts to attain the same level of gain as a single-sided shelving EQ. For example, the maximum gain of this circuit configured as a shelving EQ (by removing one of the capacitors) is approximately (R2 || R3)/(R1 || R2) which comes to about 10 or 20dB. However, solving for C1/C2 by plugging “A = 10” into equation 1 results in a negative ratio for C1/C2. These equations limit the maximum gain to about 8 or 18dBs – which is plenty for most applications.

TILT EQUALIZER

quad34.gif

The equalizer circuits shown so far are of the traditional type: they divide the audio spectrum into bands with a separate gain control for each band. The audiophile community has decried the use of such EQs, because they are difficult to set properly without introducing coloration into the audio signal – especially in the midrange. In the early 1970s, Quad Ltd. believed that the proper role of equalization was to fix subtle tonal flaws in the audio system. They developed a “tilt” tone control, which first appeared on their model 34 preamplifier. Unlike the Baxandall controls, the tilt control “tilts” the frequency content of the audio signal by simultaneously boosting the treble and cutting the bass frequencies or vice-versa. The effect is subtle because the control has a maximum boost/cut of 3dB.

eq_tilt.gif
Figure 10

The tilt EQ can be very beneficial for correcting tonal flaws in headphone sound, such as excessive brightness or darkness, without being too sonically obtrusive. The above schematic (figure 11) is from a preamplifier design by Reg Williamson and Alan Watling. It is a tilt control with a center frequency of 900Hz and a maximum boost/cut of 6dB. The circuit produces a shelving characteristic on either side on the center frequency (see the graph below). When the pot is in the center position, the EQ’s response is flat (the bypass switch takes the EQ filter out of the audio path entirely). Turning the control to the left (counter-clockwise) lightens the sound; to the right (clockwise) darkens it. Note that at the extreme settings, the tilt EQ does result in a phase inversion.

eq_tilt_curve.gif
Figure 11

TUBE-BASED EQUALIZERS

Although vacuum tube EQs can hardly be “pocket-sized,” they are prized by tube audio aficionados and sound professionals especially. All of the active EQs described so far use solid state opamps. Although not well known, vacuum tube opamps such as the Philbricks were available back in the Golden Age of tubes, when analog computers required their precision. Designing an Opamp Headphone Amplifier includes an introduction to building tube opamps (theory and schematics). These devices will work very well with the resonant filter EQ designs shown in figure 10. In some cases, however, it may be necessary to scale the feedback resistors up or down to meet circuit preferences.

eq11.gif
Figure 12

The circuit in figure 9a is a biophonic EQ that contains a simple tube opamp made of an input differential pair and a 12AT7 follower (the 33K resistors and the pots form the feedback network). Frequency shaping is accomplished with 3 triode-based resonant filters that vary the amp gain (each band can be adjusted ±12dB). The center frequencies can be approximated from equation 2 in figure 9 with these modifications: C1 = C2 and R1 = 10K. Only 6 tubes are needed: 5 12AX7 and 1 12AT7. The filament voltages are either 6.3VAC at 4A or 12VAC at 2A (twist the filament wires for lowest hum). The dual 150VDC supply should be able to supply 60mA and does not need to be regulated. The THD is less than 0.001% at 1V RMS.

Appendix 1: Build a Pocket Tilt Equalizer/Headphone Amplifier

eq_tilt4.jpg

Of all the circuits listed in the article, the tilt equalizer may be the most acceptable to audiophiles in terms of the simple circuit topology and the subtlety and effectiveness of its action. Tom Edney built a pocket version of the tilt equalizer, which he modified to drive his low-impedance Grado headphones. This section describes Edney’s modifications to the tilt EQ circuit and provides construction details.

Edney built the tilt EQ to compensate for the brightness in his Grado SR-60 headphones. He writes:

A couple of months ago when I got a pair of Grado SR-60s (my first headphones costing more than $30), I was immediately struck by how bright they are. I’d heard that Grados were bright when I had been shopping for headphones, but not being an audiophile, I didn’t really understand the significance of “brightness”. I tried burning them in for about 60 hrs, but they showed only slight improvement. The more I listened to them, the more displeased I was. The bass was lacking, and the treble was way to prominent for my taste. This was aggravated by the fact that I listen at very low volume, and as the Fletcher-Munson Loudness Curves diagram indicates, the loss of low frequencies is affected more strongly than other portions of the audio spectrum at low volumes.

So because of the combination of bright headphones, low volume, and personal preference, I had very fatiguing headphones. I could only go for about 30-60 minutes at a stretch before I had to stop using them. I considered returning the headphones, but first I looked around on the HeadWize site for an answer, and found it in the Tilt Equalizer.

eq_tilt1.gif
Figure A1

Figure A1 is the circuit that Edney used. It has a direct bypass, and R3 is omitted. Because Edney’s headphones are low impedance Grados, they need more current drive than the OPA134 can output when the EQ’s bass response is boosted. Thus, the output of the EQ section is buffered with a Burr-Brown BUF634 to provide high current drive for low impedance headphones. The feedback capacitors were changed to 2200pF to move the center frequency of the EQ up to about 2.3kHz. Edney based this change on listening tests with his Windows Media Player equalizer. He says:

Before building the Tilt EQ, I tested the idea using Windows Media Player with its 10-band equalizer. I played around a bit with the sliders until I got the sound I wanted, and saw that they formed an almost perfectly straight negative slope crossing at the 7th band from the left. Bingo! I knew then that the Tilt EQ was what I needed. Although I hadn’t worked on a DIY electronics project since high school (over 20 years), the circuit diagram looked simple enough that I decided to give it a shot.

The [original] Tilt EQ circuit seems to underboost the bass (and over-reduces treble) compared to my settings on the Windows Media Player EQ. I wondered if that was because the Tilt EQ’s pivot point is at 900 Hz, and the 7th band on the software EQ is about 2 KHz. I just tried replacing the 5600pF caps (C3/C4) with 2200pF caps, which should put the pivot point at around 2KHz. At first there didn’t seem to be a huge difference, but after a while I found I could dial in a satisfactory sound with the 2200pF, where I couldn’t with the 5600pF. Then I took the CD I was listening to and put it into my computer to see how my Windows Media Player EQ settings sounded. They were identical. Now the ratio of bass to treble seems higher, which is what I was looking for.

eq_tilt2.gif
Figure A2

Edney increased the gain of the first stage to 1.5, so that the bypass volume level would be the same as the flat EQ volume level, when connected to his Casio PCDP. Normally, gain matching is not required as all the stages of the original EQ have unity gain, and the output of the EQ should be at the same level as the input. However, when he hooked the tilt EQ to another source, the volume levels became unmatched again.

Figure A2 is an alternate circuit which functions as both an equalizer and headphone amplifier. The bypass only switches out the EQ network, so the volume level is the same regardless of the bypass setting (R3 is required). The input stage has been configured for a gain of 2. The value of the volume control should not be higher than 10K ohms (audio taper). Otherwise, the pot’s impedance could load the feedback network and change the gain of the first stage. Since the EQ is powered by two 9V batteries (see below), the gain of the first stage should be no higher than about 4 to avoid clipping when the EQ’s tilt is at full bass or treble boost.

eq_tilt3
Figure A3

Figure A3 is the power supply. The two 9V batteries in series connect to an RC network to create a ±9V output with a virtual ground. A single 9V battery will not provide enough headroom for the EQ’s output at full bass or treble tilt. (For more information about virtually-grounded supplies, see Designing an Opamp Headphone Amplifier.) Edney decoupled the power supply pins on each of the ICs with a pair of 10uF electrolytic and 0.01uF ceramic disc capacitors.

CONSTRUCTION

eq_tilt5.jpg

The tilt control pot is a dual, linear type. Low value capacitors, such as the 2200pF, can be hard to find in the preferred audiophile film types. Silver mica (or dipped mica) capacitors commonly have this value, but can be expensive; however, many surplus electronic outlets sell them at very reasonable prices (for example, All Electronics has dipped micas for about $0.35 each). If film or silva mica types are not available or are not affordable, NPO-type ceramic capacitors can be used. NPO ceramic capacitors are more stable than other ceramic types.

eq_tilt6.jpg

The PACTEC enclosure holds two 9V batteries. The bypass toggle switch is mounted on one side of the enclosure. The illuminated power switch (with green LED) is mounted on the other side. The circuit board is a protoboard from Radio Shack. Edney listed the steps that he followed to build the equalizer:

  • I successfully breadboarded one channel without the BUF. Then soldered it to the circuit board. Audio worked and so did the pot.
  • Then I breadboarded a BUF634 to the one channel with success (good audio, pot worked). I did not solder the BUF to the circuit board at that point.
  • I soldered the second channel to the circuit board, no BUF. That channel worked, and then I checked it with a breadboarded BUF. It worked fine.
  • I added both BUFs to the circuit board.
    I built the whole thing with solid wire I’d had on hand, but while trying to fit the EQ into the enclosure I broke off a number of wires. I had to completely rewire the EQ with stranded wire to keep this from happening. Obviously a mistake only a rookie would make.The dual 50K pot was almost impossible to find, especially in the small 16mm configuration. I finally found it at Main Electronics in Vancouver, B.C. (part number 08-1755). Excellent, fast service. I mounted it with the middle of the rotation at 9 o’clock, so that it would work much like the picture of the Quad Ltd. Tilt EQ shown in the article. Since my knob is round, I considered adding a cosmetic straight line (instead of just the little arrow) to the knob to more intuitively reflect the slope of the tilt. I may still do that.

    My biggest problem was that my headphone jacks were *mono*, not stereo. Because they were switched they had three leads, and I trusted the guy at the electronics supply store, when he told me they were stereo. I was getting no audio, apart from a clicking sound in the left ear at the rate of about 2 per second. Anyway, I’ve got stereo jacks now and verified that it’s all functioning properly. The darn thing works!

eq_tilt8.jpg

The Results

Because Edney’s EQ does not have a volume control, the audio source must have its own volume control. He had no trouble using the headphone output of his Casio PCDP, but the EQ generated hiss when connected to the headphone jack of his receiver. He says:

There is no discernable loss in sound quality with the Tilt EQ. I find that for most CDs I set the Tilt EQ to about 10 o’clock (9 o’clock is flat/no change). This small change warms up the audio just enough to lose the fatiguing brightness of the Grados.

Since I’m not using a headphone amp, I plug the EQ into the headphone jack on my PCDP rather than the line-out so I can use the built-in volume control of the PCDP. That is what I matched the volume to. I’m very pleased with the results. I really like being able to dial in the brightness in response to the music, the headphones, and the audio source (PCDP, Walkman, portable radio).

I consider this EQ to be just barely pocket size. It is quite portable and is perfectly suited to my desk at work, but for use in a pocket, I would make at least two changes:

    1. try to make it even smaller (using the quad version of the opamp, eliminate the BUF634 if possible, and use smaller power and bypass switches or eliminate the bypass altogether),
    2. put both the input and the output jacks on the front next to the knob. This would fit better in a pocket than having the input coming in on the side of the unit the way I’ve done it.

Someday soon I hope to change the bypass so that it includes the input stage, as the original article intended. The true bypass seemed at the time to be the best way since it completely removed the EQ, but gain mismatch is a real problem when I switch audio sources.

Appendix 2: Simulating a Tilt Equalizer in OrCAD PSpice

Circuit simulation software can be invaluable in designing or customizing equalizer circuits. With complex EQ networks, this process is, at best, a hit-or-miss proposition for hobbyists who do not have audio test equipment like a signal generator and oscilloscope, and is still not easy for those who do. EQ networks can quickly become too complex to describe with simple equations. In multi-band equalizers, the action of each band can influence the responses of the other bands. Again, there may be no simple equations to represent the overall response of a multi-band equalizer. While not always accurate, predictions of the frequency response from simulations can guide the design process.

This section discusses how to use OrCAD Lite circuit simulation software to simulate the performance of the tilt equalizer. OrCAD Lite is free and the CD can be ordered from Cadence Systems. At the time of this writing, OrCAD Lite 9.2 is the latest version. OrCAD Lite 9.1 can be downloaded from the Cadence website (a very large download at over 20M) and should work as well. There are 4 programs in OrCAD suite: Capture, Capture CIS, PSpice and Layout. The minimum installation to run the amplifier simulations is Capture (the schematic drawing program) and PSpice (the circuit simulation program).

Download Simulation Files for Tilt Equalizer

Download OrCAD Burr-Brown Simulation Libraries

After downloading tilteq_sim.zip and orcad_bb.zip, create a project directory and unzip the contents of the tilteq_sim.zip archive into that directory. Then extract the contents of the orcad_bb.zip archive into the <install path>\OrcadLite\Capture\Library\PSpice directory. The files burr_brn.olb and burr_brn.lib are libraries containing simulation models for several popular Burr Brown opamps, including the OPA134 used in this equalizer.

Note: The Burr-Brown libraries contain some very large models (such as the BUF634) that will not run in OrCAD Lite.

The two basic types of simulation included are frequency response (AC sweep) and time domain. The time domain analysis shows the shape of the output waveform and can be used to determine the harmonic distortion of the circuit’s output or to visually inspect the waveform for anomalies like clipping. They both run from the same schematic, but the input sources are different. For the frequency response simulation, the audio input is a VAC (AC voltage source). The time domain simulation requires a VSIN (sine wave generator) input. Before running a simulation, make sure that the correct AC source is connected to the amp’s input on the schematic.

orcad_tilteq1.gif

The following instructions for using the simulation files are not a complete tutorial for OrCAD. The OrCAD HELP files and online manuals include tutorials for those who want to learn more about OrCAD.

Frequency Response (AC Sweep) Analysis

  1. Run OrCAD Capture and open the project file “tilt_eq.opj”.
  2. In the Project Manager window, expand the “PSPICE Resources|Simulation Profiles” folder. Right click on “Schematic1-freq_resp” and select “Make Active.”
  3. In the Project Manager window, expand the “Design Resources|.\tilt_eq.dsn|SCHEMATIC1” folder and double click on “PAGE1”.
  4. On the schematic, make sure that the input of the amp is connected to the V3 AC voltage source. If it is connected to V4, drag the connection to V3. By default, V3 is set to 0.5V.
  5. To add the Burr-Brown library to the Capture: click the Place Part toolbar button (orcad1.gif). The Place Part dialog appears. Click the Add Library button. Navigate to the burr_brn.olb file and click Open. Make sure that the analog.olb and source.olb libraries are also listed in the dialog. Click the Cancel button to close the Place Part dialog.
  6. From the menu, select PSpice|Edit Simulation Profile. The Simulation Settings dialog appears. The settings should be as follows:Analysis Type: AC Sweep/Noise
    AC Sweep Type: Logarithmic (Decade), Start Freq = 10, End Freq = 100K,                      Points/Decade = 100
  7. To add the Burr-Brown library to PSpice: Click the “Libraries” tab. Click the Browse button and navigate to the the burr_brn.lib file. Click the Add To Design button. If the nom.lib file is not already listed in the dialog list, add it now. Then close the Simulation Settings dialog.
  8. To display the input and output frequency responses on a single graph, voltage probes must be placed on the input and output points of the schematic. The probes should already exist on the schematic. If not, here’s how to add them: Click the Voltage/Level Marker (orcad2.gif) on the toolbar and place a marker at the junction of R3a, R4a and C3. Place another marker just above RLoad.
  9. OrCAD does not have a functional model of a potentiometer. R5a and R5b represent a 50K-ohm pot. When R5a = R5b = 25K ohms, the pot’s wiper is at the center. To “rotate” the pot to a clockwise or counter-clockwise position, make R5a << R5b or R5a >> R5b, but in all cases, the sum of these resistors must total 50K. For example, in the schematic shown above, the equalizer is set for a bass tilt with R5a = 1K and R5b = 49K.
  10. To run the frequency response simulation, click the Run PSpice button on the toolbar (orcad3.gif). When the simulation finishes, the PSpice graphing window appears. The input and output curves should be in different colors with a key at the bottom of the graph.

    orcad_tilteq2.gif

  11. The horizontal axis of the graph does not have adequate markings to determine center frequency of the tilt by eye. To display the PSpice cursor, select Trace|Cursor|Display from the PSpice menu. A vertical cursor line appears on the graph, and the Probe Cursor window appears over the graph. Drag the cursor to the point where the output curve (shown here in green) intersects with the input curve (shown here in red). Then, the exact coordinates of the intersection are shown in the Probe Cursor window on the A1 line. The first number (2.3101K) is the center frequency. The second number (500mV) is the corresponding voltage for that frequency.

Time Domain (Transient) Analysis

  1. On the Capture schematic, make sure that the input of the amp is connected to the V2 sinewave source (the default values are: VAMPL=0.5, Freq. = 1K, VOFF = 0). If it is connected to V3, drag the connection to V2.
  2. In the Project Manager window, expand the “PSPICE Resources|Simulation Profiles” folder. Right click on “Schematic1-transient” and select “Make Active”
  3. From the menu, select PSpice|Edit Simulation Profile. The Simulation Settings dialog appears. The settings should be as follows:
    Analysis Type: Time Domain(Transient)

      Transient Options: Run to time = 10ms, Start saving data after = 0ms, Max. step size = 0.001ms
  4. To display the input and output waveforms on a single graph, voltage probes must be placed on the input and output points of the schematic. If the probes are not already on the schematic, follow the procedure in step 8 of the Frequency Response Analysis to add them.
  5. To run the time domain simulation, click the Run PSpice button on the toolbar (orcad3). When the simulation finishes, the PSpice graphing window appears. The input and output curves should be in different colors with a key at the bottom of the graph.
  6. To determine the harmonic distortion at 1KHz (the sine wave frequency), harmonics in the output waveform must be separated out through a Fourier Transform. In the PSpice window, press the FFT toolbar button (orcad7.gif). The PSpice graph changes to show the harmonics for the input and output waveforms. The input and output curves should be in different colors with a key at the bottom of the graph.
  7. The fundamental frequency at 1KHz will have the largest spike. The other harmonics are too small to be seen at the default magnification. In the PSpice window, press the Zoom Area toolbar button (orcad8.gif) and drag a small rectangle in the lower left corner of the FFT graph. The graph now displays a magnified view of the selected area. Continue zooming in until the harmonic spikes at 2KHz, 3KHz, etc. are visible.
  8. Harmonic spikes should exist for the output waveform only. The input is an ideal sine wave generator and has no distortion. To calculate total harmonic distortion, add up the spike values (voltages) at frequencies above 1KHz and divide by the voltage at 1KHz (the fundamental).

Additional Simulation Tips

  • To change the value of any component on a schematic in the Capture program, double-click on the value and enter a new value at the prompt.

Note: Simulations only approximate the performance of a circuit. The actual performance may vary considerably from the simulation as determined by a number of factors, including the accuracy of the component models, and layout and construction techniques.

References:
Berlin, Howard, Design of Operational Amplifier Circuits (1984).
Berglund, Rickard, “Ultra-Low-Distortion Graphic EQ,” Glass Audio, 6/96, p. 40.
Burr-Brown Applications Notes (OPA132, OPA134).
Graf, Rudolph Encyclopedia of Electronic Circuits, Vols 1-6 (1985-1996).
Horowitz, P. and Hill, W. The Art of Electronics (1989).
Jung, W. Audio IC OpAmp Applications (1988).
Meiksen, Z. H. Complete Guide to Active Filter Design, Opamps, and Passive Components (1990).
National Semiconductor Application Notes (LF353, LM349).
National Semiconductor Audio/Radio Handbook (1986).
Texas Instruments Application Notes (TLE2426).
Williamson, Reg and Watling, Alan, “A New Control Preamp,” Audio Amateur, 4/91, p. 10.

c. 1998, 2000, 2001, 2002 Chu Moy.

Addendum

11/15/1998: Added section on active 2-band Baxandall. Updated figures 10c,d. Lowered voltage divider resistor values to 5K.

9/30/1998: Corrected error in text about how to change the midrange center frequency of the 3-band active Baxandall and added further discussion re implementing the biophonic curve. Also added section on tube EQs.

5/1/2000: Revised section on battery supplies. Added figure 14.

8/6/2000: Added section on tilt tone control.

12/22/2001: Corrected capacitor values in 28Hz-band section of tube equalizer schematic (figure 10). Original value was 180pF. Revised value is 180nF. Thanks to Steven Zielinski for finding this error. Also revised section on power supplies.

2/15/2002: Steven Zielinski built a 100W tube guitar amp with a 4-band equalizer based on the circuit shown in figure 10. He substituted 5K ohm pots for the 10K ohm pots and calculated the capacitor values for the center frequencies of the equalizer section: 40Hz, 200Hz, 1000Hz and 5000Hz respectively. The amp is called “Big Bloo”. He writes:

zielinski1.jpg

The amp is one-channel, and power varies from about 85W to 130W depending on whether I have the power supply configured as choke-input (580V) or Pi configuration (~700V, but sags a bit). I think I chose the centre frequencies poorly, as not many 5kHz tones come from an electric guitar. There’s certainly enough boost/cut though. I think I will eventually trim the EQ stage back to 3 bands, use some typical corner frequencies of other guitar amps and use the triode as a mixing stage for a power amp input. Here are the capacitor values for each of the EQ stages (see figure 10):

Band #1 40Hz 100nF
Band #2 200Hz 25n9F (22nF||3n9F)
Band #3 1kHz 5n17F (4n7F||470pF)
Band #4 5kHz 1nF

I also have a hum problem due to this chassis having so many modifications. This will be addressed before the centre frequencies are moved. For those unfamiliar with tube circuitry, the bias circuitry is not shown. Any adjustable source from -40V to about -100V will do (negligible output current required).

zielinski2.png
zielinski3.png

For the power supply for the EQ section, I thought it best to show from ‘mains in’ to output, as we’re on a different wall voltage to you guys. The power supply utilises a capacitor from +Ve to -Ve rail without referencing ground. This mirrors noise more evenly across positive and negative. Your opamps aficionados should take note (of course you still have two R-C networks referenced to ground preceding the final capacitor).

zielinski4.jpg

The power transformer was custom made – at least 600VA (the winding is 600V). All I did was multiply maximum current in the output valves by four and add in some headroom. This transformer also had 3 separate heater windings for pre-amp valves (so they could float at whatever voltage was needed to keep within cathode-to-heater limits). Add in 6A times 6.3V for the main heaters and you can see how it ended up with such huge VA rating. Yes it was overkill but it was a work in progress and needed flexibility. Obviously you can use separate transformers, in fact I had to when I broke off the wire for the bias winding.

About the power amp: Morgan Jones wrote the book that gave me several of the ideas used in BigBloo. Overall, it’s a pretty serious piece of gear to build if you haven’t done tube equipment before. (Caution: Only advanced DIYers with an understanding of high voltage construction techniques should attempt this type of project.)

What is the value of the choke at B+? Dunno, it was in the junkbox, recovered from an old B&W; TV power supply. Probably in the tens of mH range at a guess. C- is the bias supply. For 6550’s I think it is around -55V, but is always adjusted on test setting the bias. That is another advantage of the cathode being the power amp “switch”, the bias is present before the tube tries to conduct. Some early Marshall amps suffered from this problem.

The RC relay contact is part of an automated standby switch. The heaters come on at initial power-up, and after ~90 seconds, the relay contact closes and the cathode is connected to ground and the valve can operate. This enables the use of low voltage relays instead of switching on the screen voltage. All I used was a 555 timer to control the relay via a BC548. The main purpose of the delay is to protect the output valves from cathode damage when cold.

5/2/2002: Added Appendix 1 and Appendix 2.

7/31/2004: Forum member stereth built this 3-band Baxandall eq/headphone amp in a Fossil watch box. He writes:

This is my first DIY audio project and I thought I’d share how it went. Almost all of the parts for this came from Radio Shack. I’ve learned my lesson – next project, I’m getting everything online. I probably could have saved $10-15 by doing so, out of $50 or so. The dual pots were the only things I had to get online.

The circuit board I used is a Radio Shack project board with two bus paths. I soldered them together and used them for ground. Unfortunately, most of the connections are between three holes, and the circuit has a lot of four-branch nodes. So I connected them in pairs and had groups of 6.

I managed to come across the perfect container at home. It’s from Fossil, basically an oversized Altoids tin. And it’s still borderline pocket size.

stereth_eq_1.jpg
stereth_eq_2.jpg
stereth_eq_3.jpg

Creating a 2nd Headphone Feed When Only One Is Available.

by George Kourounis

Sometimes when you are working on a recording session, it’s not uncommon that one musician wants to hear a different headphone mix of the music than the rest of the band. The drummer may want to hear lots of the bass guitar and kick drum, while the guitarist wants to hear mostly himself (not a rare thing among guitarists). Achieving this task isn’t difficult, provided you are in a studio that has two discrete headphone cue feeds. If, however, you are in a studio that has only one headphone feed, then you’re forced to improvise with the resources you have available.

If you’re lucky, the studio will have an extra amplifier sitting around. If not, you may need to borrow the amplifier from the stereo system in the lounge or the studio manager’s office (always ask permission first). It doesn’t need to be a Bryston or anything like that, but it must have enough juice to power a couple of sets of headphones.

If the studio’s cue system is fed from, for example, auxiliary sends 5 and 6, then use them to send your first stereo mix to whichever musicians want it. Then, for the second mix, take the outputs of another pair of auxiliary sends from the patchbay (3 and 4 will work fine) and route them to the imputs of your “borrowed” amplifier. Connect the outputs of that amplifier to a headphone cue box via a banana plug to XLR adapter cable (or something similar). Connect your headphones to the cue box and – abracadabra! – instant second stereo headphone mix!

An important note, though: caution must be taken when setting the levels on the power amp and attention to the impedance of the headphones should be observed so that you don’t end up blowing up a set of headphones by mistake. Musicians hate that, studio owners hate that even more and toasting headphones is generally considered an engineering no-no.

Going to the extra trouble of creating a second cue system can be a little time-consuming while setting it up, but the benefits can be rewarding. If the musicians have the exact mix in their headphones that they want, they will likely give a better musical performance. The better the musical performance, the better the song will be and the happier everyone involved will be.

c. 1994, Professional Sound Magazine.
From Professional Sound Magazine (Spring 1994). (Republished with permission.)

Headphone Monitor Switch.

by Rudy Trubitt

trubitt1.gif

I wrote an article on field recording techniques in the November 1997 issue of Electronic Musician. In that story, I promised the plans for a simply do-it-yourself (DIY) headphone monitor switch I use in my work. Here it is! Please note I am not providing a complete, detailed, step-by-step guide to creating this little wonder box. It is not a complicated project, but some previous experience with DIY electronics is required. If you’ve done this sort of thing before, you’ll be able to fill in the details easily.

In any recording situation, monitoring is critical to make sure you’re getting what you want on tape. This is just as true in field recording, but in most cases, one’s monitoring options are severely limited–stereo headphone is the only choice. Since I often use dual-mono mics, hearing a stereo feed of the two is not always convenient. I wanted the option to hear JUST the left mic in BOTH ears, or just the right mic in both ears, as well as a normal stereo signal. This is simple enough to do with a big rotary switch. When completed, you can create a little box that your headphones plug into, which in turn is plugged into the stereo phone output of your deck. Then, by turning the knob on the switch box, you can hear normal stereo, left-only mono, right-only mono, left+right mono and even left-right reversed stereo (or normal stereo again).

trubitt2.gif
Figure 1: Wiring diagram of headphone monitoring box

Figure 1 is the wiring diagram for the box. Figure 2 shows a slightly more conventional schematic of the same switching circuit. The common contacts are shown at the far left. These go to and from the left and right outputs of the deck and to the left and right inputs of the phones. Ground connections, while necessary, are not shown. If you build the project into a metal case, that might provide the ground you need. I ran a wire between the sleeve contacts of the input and output connectors of the box.

trubitt3.gif
Figure 2: Schematic of headphone monitoring box

Note the use of summing resistors in the left+right mono section. This was an attempt to prevent the two outputs from “fighting” each other if there were very different voltages in left and right outputs. I used 8 ohm resistors here, but a higher value might be better. Maybe ~20 ohms? Also, I initially decided to put normal stereo on both ends of the switch’s travel so I’d always be able to find it without looking. However, I sometimes wish to have left-right reversed. If you’d like to try this, simply swap the leads on one of the “normal stereo” connections.

I also put a 1k Ohm linear taper pot across the output. This gave me a knob for setting the level on my Sony D8, rather than their inconvenient push-button volume control the deck provides. If you do this, you might find, as I did, that it is more convenient to leave the deck’s switch in “line out” position, which gives you maximum level from the deck without having to ever wonder where its volume level is set. Then just turn down the pot on the switch box to a comfortable level. Of course, this is at the expense of battery consumption, but since I’m using an Eco-charge battery, this hasn’t been an issue.

One final caveat: The left only/right-only mono positions are -6dB down, since only one half of the deck’s headphone amp is driving your phones when the switch is in those positions.

Good luck!

c. 1998, Rudy Trubitt
From Rudy Trubitt’s site. (Republished with permission.)

A Compact 50W Integrated Amplifier with Meier Headphone Section.

by Tim Harrison

harris2_0.jpg

My reason for constructing this project was to develop a design for a compact integrated stereo amplifier suitable for use by a poor (but sound quality conscious!) student living in a university or college dorm. The amplifier drives a pair of loudspeakers using two LM3876 integrated power amp ICs (50 watts per channel), or a pair of headphones via a Meier crossfeed filter and an OPA2134 dual opamp. It provides four switchable line level inputs, and an unbuffered line level output for recording purposes. The design uses readily available good quality components, and is based around four separate PCBs; one for each power amp channel, one for the power supply board, and one for the preamp/headphone driver.

THE CIRCUIT

harris2_10.gif
Figure 1

The block schematic for one channel of the design is shown above (figure 1). The preamp and the first stage of the headphone amp are separate in this application, ‘straddling’ the gain across the volume control. There is an initial gain of 2.5 before the control, followed by a further gain stage of x3 after it. This arrangement allows the power amp to be driven directly from the output of the volume control without further gain, and makes for lower noise operation of the headphones. The input selector switch is a 4-way, 3-gang type, so one gang isvused for each channel, and one gang is used to switch the input indicator LEDs.

harris2_1.gif
Figure 2

Above is the schematic for part of the preamp board (figure 2). The output of the selector switch is sent to pins J1 and J3. Looking at the left channel, C1 and R2 form a low pass filter with a -3dB point of 40kHz, which rejects any RF interference picked up on the interconnects. R2 also sets the input impedance of the unit, in this case 47k ohms. R1 ensures the opamp U1 is presented with an equal impedance at both its inputs, helping improve its distortion performance as outlined on the OPA2134 datasheet. The value of R1 (9k1) is the nearest commonly available value to the parallel combination of R3 and R4 (22k and 15k respectively). R3 and R4 set the gain of this stage, just under 2.5 in this case. This value allows ample headroom for a wide range of source signals, which could be as much as 3VRMS. In this case, the peak output voltage of 10.6V would be fine with the suggested ±15V power supply.

This initial gain brings the signal up to a level whereby the output from the volume control can drive the power amp circuits directly, with no further gain, and allows the headphone driver circuit to operate with a lower gain, giving lower noise performance. C7 forms a 100kHz low pass filter with R3, rolling off the gain to unity at very high frequencies, and helping promote stability of the opamp. It is not strictly necessary with the suggested OPA2134 device, but allows the drop-in substitution of a cheaper but more oscillation prone device, such as the NE5532, if budgets are tight. C19 AC couples the output from this stage to the volume control, and with a 50k potentiometer, sets the -3dB point of the headphone amp’s response at 1.4Hz (the power amp has further high pass filtering). This capacitor is very important, as all the other stages are DC coupled, and C19 prevents any DC offsets from source components being amplified and presented to the headphones or speakers.

The resistor R9 links the output of the input selector to a recording device, such as a tape deck or minidisc recorder. It helps prevent the source becoming too loaded down feeding both the input gain stage and the recording device, and protects the source should the output become shorted to ground for any reason. The outputs from J5 and J6 are fed into the volume control pot, which should be a good quality type. Finally, C3 to C6 provide local decoupling of both the power supply rails, C5 and C6 decoupling the high frequencies, with C3 and C4 decoupling the lower ones.

harris2_2.gif
Figure 3

The output of the pot feeds the power amp and the headphone driver, which is also mounted on the preamp board. Looking at the above schematic for the headphone driver (figure 3), we can see that the opamp U2 is used in a similar configuration to the input amp U1. In this case, R24 matches as closely as possible the parallel combination of R11 and R12, helping reduce distortion as before. Again, C21 allows compatibility with cheaper opamps. R11 and R12 set the gain of the stage at just over 3, bringing the signal up to a level sufficient to drive a pair of headphones. This stage also acts as a buffer, isolating the Meier crossfeed filter from the varying output impedance of the volume control. C8, R14, (with C10, R21, and R15) form a crossfeed filter, which in this case is permanently wired in circuit. A detailed description of the operation of this circuit can be found in Jan Meier’s article A DIY Headphone Amplifier with Natural Crossfeed.

Basically, the circuit performs a frequency selective mix of the two channels into each other, allowing recordings meant for speaker listening to sound natural on headphones. I had built projects with the filter made switchable in the past, but I never turned it off, so the switch was omitted here. Finally, the opamp U3 forms a simple noninverting buffer to drive the headphones. R17 forms a minimum load when the phones are disconnected, and helps prevent pops and clicks when they are connected with the unit powered up. While it is possible to substitute cheaper opamps in other parts of the circuit, the device used here needs to have a high output current capacity, and must remain stable when driving difficult loads. J10 and J12 are the output to the headphone socket, which should have its ground isolated from the chassis so as not to defeat the ground loop breaker circuit. Again, C11 to C18 provide local supply decoupling for the opamps.

You can find more information on the detailed operation of opamp based circuits, such as the preamp and headphone amp circuits presented here, in Chu Moy’s article Designing an Opamp Headphone Amplifier. Figure 4 is the power amp schematic for one channel (both channels are identical – and use one power amp board each).

harris2_3.gif
Figure 4

The circuit given here is similar to the one presented in the article, Single Chip 50 Watt / 8 Ohm Power Amplifier, on Rod Elliott’s site, ESP. The LM3876 is a good quality component capable of delivering 56W continuously into an 8 ohm load and 100W peak – enough for any dorm! It has a quoted distortion figure of 0.06% at 40W output, and offers good sound quality in a simple design. It has comprehensive output protection circuitry, preventing not only thermal runaway, but protecting the device from short circuits on the output, and voltage spikes from inductive loads.

Looking at the circuit, R3 and R1 set the gain of the power opamp at 23, and C1 limits the DC gain to unity. It also forms a low pass filter with a -3dB point of 7.2Hz. R2 draws roughly 1.5mA from pin 8, disabling the internal muting function of the LM3876, and C2 provides a large time constant for the action of the muting circuit. R4 should be a 1W resistor, and has 10 turns of 0.4mm enamelled wire wound round it, with its ends soldered to the resistor leads, giving a roughly 0.7uH inductor in series with the 10 ohm resistance. The inductor acts to promote stability of the power opamp, by ensuring a minimum 10 ohm load at higher frequencies. Likewise, the low pass zobel network formed by C7 and R5 (which should also be a 1W type), helps prevent oscillation should any RF appear on the output. C3 to C6 provide local supply decoupling for the power amp IC.

To enable the power amplifier to deliver its full rated power (56W/ch) continuously, and to cater for the potential 100W peaks, I decided to build a good quality power supply for the project, capable of supplying 200W. The main power supply for the speaker amps was built directly into the chassis, and is a fairly standard design. It supplies ±35V, and is capable of just over 3A continuous per rail for both the power amps. A ±15V supply for the preamp and headphone driver is provided from the main supply by the PSU board. Firstly, I will describe the main power supply, whose schematic is shown below (figure 5):

harris2_4.gif
Figure 5

The mains enters the chassis via a filtered IEC inlet, and the live line is fed through a 1A antisurge type fuse mounted in an insulated chassis fuse holder, before both the live and neutral lines are fed to a DPST rocker switch mounted on the front panel. The mains feed from the switch is connected to the primary of the power transformer, and a pair of transient suppressors are wired in parallel with it (only one is shown in the diagram). They should be rated for the mains voltage where you are, and should be mounted securely on the base of the chassis, I used two sections from an insulated terminal block.

The secondaries of the transformer are wired in series, and the wires from the toroidal types can be connected directly to a heavy duty chassis mounted bridge rectifier. The output of the bridge rectifier is sent to a pair of reservoir capacitors, C2 and C3, connected in parallel with C4 and C5, which provide high frequency decoupling. The only other point about the power supply that needs explaining is the ground loop breaker circuit. The 0V rail is connected to chassis ground and mains earth via R1, a 10 ohm wire wound resistor, in parallel with C1, a mains rated 100nF capacitor. The resistor prevents any currents flowing round the loop created by the mains earth and the ground in unbalanced phono interconnects. The 100nF capacitor shorts the resistor at high frequencies, allowing any RF to flow to ground in the normal way. I placed C1 and R1 on the underside of the stripboard I used to mount the reservoir and decoupling capacitors.

The output from the main PSU is fed to the power amp boards via a front panel DPST switch, allowing the speaker amps to be switched off for headphone only listening, and also (unswitched) to the preamp PSU board. Below is the schematic for the preamp PSU board (figure 6):

harris2_5.gif
Figure 6

The ±35V rails must be reduced to around ±20-22V before they can be fed to standard three pin regulators. I simply used a potential divider comprising R3 and R6 for the positive rail, and R4 and R7 for the negative rail. Simply placing a reverse biased 12-15V zener diode in series with the supply, i.e. in place of R3 and R4 (and omitting R6 and R7), would be an alternative option, and probably simpler – this option didn’t occur to me until after I built the prototype! C1 to C4 decouple the output of the regulator, and R1, R2, and R5 set the current flowing though the LED indicators, around 15mA in this case. The stabilised ±15V supply is presented on pins J1-J3, and the remainder of the pins provide supplies for a pair of power on indicator LEDs (mounted next to the mains rocker switch), and the input selector LEDs. These are mounted above the input selector switch, and light to show which input has been selected. They are controlled by the remaining gang of the three gang rotary switch.

CONSTRUCTION

I have provided my PCB artwork for you to use to make your own PCBs if you are interested in building all or just part of this project. Below are links to the artwork files, and the relevant placement guides (all GIF format):

harris2_9.gif

Download PC Board Artwork and Placement Guides 1
Download PC Board Artwork and Placement Guides 2
Download PC Board Artwork and Placement Guides 3
Download PC Board Artwork and Placement Guides 4
Download PC Board Artwork and Placement Guides 5
Download PC Board Artwork and Placement Guides 6

The artwork will print the correct size if you set your graphics software to output 600dpi to your printer. The placement guides should be printed at 300dpi. If you have trouble getting them the right size, the power amp boards should be 41mm wide, the PSU board should be 113mm wide, and the preamp board 132mm. I made the artwork for the lead pitch and size of the components I could source, so I suggest you print out the placement guides real-size (300dpi), and compare the sizes and lead pitches of the components you can source, selecting the ones that best fit the board.

As a guide to component selection, I used 0.6W metal film resistors throughout, except for R3, R4, R6, and R7 in the PSU, which should be 5W wire wound radial lead types, and R4 and R5 in the power amp, which should be 1W wire wound types. For the decoupling capacitors use ceramic disc types, and for capacitors in the signal path (C19, C20, C8 and C10 in the preamp), I used the Wima MKS4 250V series, although any metal film type will do (but may not fit).

harris2_6.jpg

The line level signal from the sources is received via an array of gold plated phono plugs on the rear of the unit. The plugs I used had red and black identification bands on them to indicate which channel should be connected to them. This was important, as I was not planning to print any lettering onto the case, so the connections and controls had to be fairly self-explanatory. The phono plugs should be mounted using an insulating bush, as the design uses a ground loop breaker circuit, and the signal and earth (chassis) grounds are separated.

The source signals are routed via screened cable to a rotary selector switch mounted on the front panel which is used to select the source to be listened to (and recorded from). The switch should be a good quality part, as a positive tactile response from it enhances the feel of the finished project. The part I used was a 3-gang 4-way type, allowing 4 stereo inputs to be accommodated, leaving one gang free to switch the source indicator LEDs mounted above the control. I used a cheap part by a company called Alpha, their SR2611 series. This switch works fine and only cost a few pounds (roughly $5 US).

For a volume control, I used a 50k ALPS pot, but a cheaper type of any value between 10-100k could be used. A conductive plastic track type is preferable to a carbon track, and should be logarithmic law (also called audio taper). The ALPS RK27 series pots (the blue ones), while pricey, come highly recommended, as they have a very nice tactile feel to them, and exhibit good tracking between the gangs.

harris2_11.jpg

For the preamp and headphone section opamps, I recommend the OPA2134 by Burr Brown, and DIL sockets are a good idea to help prevent heat/static damage during soldering. Note, the LM3876T power opamp in figure 4 must be used with my PCBs, the T suffix denotes the package type. The power opamps share a large 2 degrees C per watt heatsink mounted on the rear panel in the prototype and are mounted using greaseless silicone insulators and insulating bushes. Make sure the metal tab of both the power amp ICs is isolated from the chassis – this is very important.

Power supply

The value of the mains fuse in figure 4 varies depending on what type of transformer you use, and the supply voltage in your country. Since I live in the UK where the mains supply is 230V, and I am using a 225VA rated toroidal transformer, a 1A antisurge fuse was used. Take care to get this value right, as if it is too low, you will suffer nuisance blowing, and if it is too high, you will not get proper protection in the event of a fault. The fuse rating can be calculated in the normal way using I = P / V. A double pole type switch is preferable to a single pole type, as it allows the unit to be completely isolated from the mains when it is switched off. The mains rocker switch used should be rated to handle the in-rush current of the transformer, anything over 4-5A should be fine in this case.

harris2_8.jpg

I don’t normally include transient suppressors in the power supplies of audio projects I build, as I run the mains supply to my home audio system through a filter which includes them. However, as this integrated amp was designed to be used away from home, they are included here. You can use a pair together in parallel as suggested to increase their dissipation capacity.

The toroidal power transformer was made by Nuvotem and sourced from RS components in the UK, at a cost of about 23UKP (or about $35 US). In general, the power transformer itself should be a good quality type, and I recommend a 225VA toroidal part for this project. If the budget is tight, a lower value toroidal (say, 160VA) could be used, or even a conventional EI laminate type. Although the standard EI transformers are cheaper than toroidal types, and exhibit a lower inrush current, they are less than ideal for a compact unit. They tend to be quite bulky, and emit strong electromagnetic fields, leading to hum pickup in adjacent circuitry. A toroidal transformer is both compact, and emits a far less strong field.

Although the rated current of the power supply is only 3A, the charging current of the reservoir capacitors will be much higher than this at times. I recommend using a 35A type bridge rectifier, such as the KBPC3506. A pair of heavy duty insulated terminal blocks should be mounted nearby, and the centre tap of the secondaries connected to this. The terminal block will now form a star grounding point, and should be the place all the 0V rails in the unit are connected together. This method of grounding ensures hum free operation. Save yourself a lot of grief and use this method the first time – hum free results are almost guaranteed.

If the budget is tight, 4,700uF capacitors can be substituted for the 10,000uF ones specified in figure 4, especially if a 160VA transformer is used. I had trouble fitting 10,000uF caps into the chassis, so I used two 4,700uF caps in parallel per rail. I couldn’t get any capacitor mounting brackets, so I simply soldered C2 – C5 onto a small piece of stripboard. You could use either method, but be sure to take your DC output from the capacitors and not the rectifier.

For the 5W resistors in figure 5, I used vertically mounting ceramic, wire wound resistors, but you could use standard axial types, with one leg bent down the side, if you find the radial types hard to get. C5 to C8 decouple and stabilise the output of the potential divider (or zener diode), before it is fed into a pair of standard voltage regulators. These should be mounted with a pair of small flag type (clip-on) heat sinks with a thermal resistance of around 20-25 deg. C per watt. I used Redpoint Theramalloy PF752.

Chassis

I mounted the project in a compact instrument case (300mm W x 150mm D x 100mm H), which has a removable internal chassis. The case was supplied painted grey, but once I had drilled it, I decided to repaint the chassis blue to make the project look more individual. I prepared the chassis by sanding it down thoroughly, making sure that all surfaces would provide a good key the paint could adhere to. I then cleaned all the surfaces with white spirit, and applied three thin coats of standard car spray paint. I used diffused blue 3mm LEDs, black rocker switches with blue markings, and black aluminium knobs to complete the effect.

You can see the layout I used pretty clearly from the pics inside the unit, all the boards were mounted on the base of the chassis, except for the two power amp boards which were mounted on the rear panel. The bridge rectifier is bolted to the bottom of the metal chassis. I used 4mm binding posts for the speaker terminals, two black ones for the ground connection, a green one for the left channel, and red for the right. All the signal wiring should be done using shielded cable, with the screen grounded at one end only. Ribbon cable can be used for LED wiring, 32/0.2mm hookup wire should be used for power amp supply and speaker connections, and 7/0.2mm hookup wire can be used for other low power connections.

If there is hum on the output of the completed project, the problem is almost certainly to do with the ground scheme used. Make sure that there is a 10 ohm resistance between the chassis and signal ground (i.e. that you have not defeated the ground loop breaker), and make sure you have not accidentally grounded a point by two paths simultaneously. The star grounding scheme as outlined earlier is highly recommended. The path to ground on the volume control pot is particularly critical, in the prototype the unit refused to stop humming until the far end of the wiper had its own separate connection to the star ground point. It should be possible to set the volume control to zero and, with the unit on, put your ear to the speaker and hear nothing but a faint hiss.

RESULTS

My impression of the project overall is very good, it sounds good, and is very compact. The performance from the IC power opamp is impressive, and I think my prototype looks nice, too! Listen to your favourite cans through it late into the night, or let it provide some serious slam through speakers for a small room or dorm.

c. 2002 Tim Harrison.

The Art of Monitoring and Mixing With Headphones.

Is it really an art to monitor and mix with headphones? There is certainly no dearth of discussion on the matter. Based on hours of reviewing newsgroup postings, articles, white papers and other pro audio literature, it must be so, since opinions on the subject range from gung-ho support of headphones for all monitoring and mixing to banning headphones outright from the studio and the stage. Performers worry about one set of concerns, while recording engineers fret over another. Musicians tend to be more accepting of headphones than recording engineers. Since consumers currently prefer loudspeaker fidelity over headphone fidelity, engineers who successfully monitor and mix with headphones must be able to correlate the two very different soundfields.

This article does not take a stance on whether headphones are better than loudspeakers in professional settings (even the most ardent champion of cans will say “it depends”), and instead examines techniques for using headphones in all monitoring situations. It may one day be possible to dispense with monitor speakers entirely, but the consensus agrees that day is not here yet. Nevertheless, when used with forethought and grounded in practical experience, headphones can challenge loudspeakers on every front. Further, in particularly demanding applications, a little technological wizardry can go so far as to fool the ears into believing that headphones are loudspeakers.

SETTING UP FOR A LIVE PERFORMANCE

Equipping The Performers

Hearing preservation as well as higher quality sound have been the driving forces behind the popularity of headphones and in-ear monitors for performers and audio engineers alike. Monitor speakers (wedges and sidefills) are the traditional means for musicians to hear themselves during a performance. The wedges are the primary sound sources, and the sidefills help to maintain uniform coverage across the stage. Most of the time, the sound image from such setups is washed out stereo, the image being best at a “sweet spot” on stage, which degrades if the performer moves around. Monitor speakers also have the bad habit of “spilling” off stage (due to wall reflections and the radiation pattern of the loudspeakers themselves) and compromising the soundfield from the house system. They are also prone to feedback if the microphones are not carefully placed. Perhaps the biggest concern is that monitor speakers (and concert sound systems generally) play so loudly that the musicians suffer hearing damage. Ear plugs are not an option since musicians onstage need to hear each other clearly.

Headphones can avoid most of these problems. The acoustic isolation of headphones allows each performer to listen to a monitor mix at a comfortable volume, with improved fidelity and dynamic range, and, when the equipment provides, to customize the monitor mix without affecting what others hear. During live performances, monitoring with standard headphones is pretty much restricted to rhythm sections and instrumentalists. Closed-ear headphones improve the intelligibility of a monitor mix and provide some attenuation of ambient noise as well. There is no feedback since the output from the canalphones is not audible to onstage microphones. For greater realism and interactivity, one or two stage monitors may be installed to expand headphone sound, or a performer may wear the headphone with one earcup off. However, these techniques can place hearing at risk.

Performers, such as vocalists, who like to move around the stage are not good candidates for headphone monitoring – primarily because headphones are too bulky. Instead, in-ear monitors permit mobility and provide acoustic isolation that is better than closed-ear phones. The popularity of in-ear monitors has displaced most, if not all, on-stage wedges and sidefill monitors. The earpieces of in-ear monitors are typically canalphones with excellent attenuation of ambient noise. (See A Quick Guide To Headphones and Preventing Hearing Damage When Listening To Headphones for more information about canalphones.) Performers are free to roam about the stage, toting wireless receivers to drive the canalphones and listening to a personalized mix in aural privacy. Without the higher SPLs of stage monitors, vocalists additionally benefit from reduced vocal fatigue, as they no longer have to strain their voices singing over the mix. See STUDIO MONITORING AND MIXING for a special discussion about vocalists and headphone monitoring.

Tips on Setting Up The Monitor Mix

A monitor mix for headphones and in-ear monitors has different requirements than one for loudspeakers. Headphones sound more detailed, offer true stereo (as opposed to the washed out image of stage monitors), and isolate the listener from ambient sounds. Consequently, a good headphone mix is more difficult to achieve. At the same time, since they can hear better and are listening to higher quality sound, performers are more likely to demand custom mixes. Here are some tips from Steven McCale and other engineers familiar with the art of mixing with ear monitors:

  • The mix must have everything that a musician would hear without headphones: drum overheads, left and right keyboard feeds, video sound.
  • Use audience microphones to lessen the isolation of headphones.
  • Always provide a stereo mix, for greater intelligibility and so that each musician has a spatial location.
  • For a more natural sound, add special effects such as reverb, harmonizers and delays to the mix.
  • Be prepared to provide for a large number of custom mixes.
  • Every ear monitor should have a limiter to prevent hearing damage – especially important in live performances where spurious noise in a sound system could amplify to deafening levels.
  • Install gates on drums and compress bass guitars slightly for lively, controlled sound.

A setup designed for headphone monitoring is vital for creating a good mix. The average console may not have enough inputs or outputs. Headphone mixer amplifiers have multiple inputs for a main stereo mix, various subgroup mixes (such as drums, background vocals or keyboards) and possibly a separate effects loop. Individual levels controls let musicians dial their own custom mixes for each headphone output. Remote mixers, such as the The Psychologist from Intelix, are the most convenient of all, and free up console outputs besides. As performers may be using headphones with different efficiencies and listening requirements, booster amplifiers can augment the individual outputs of the mixer. The better headphone distribution amplifiers have two inputs per output – one for the main mix and another for a custom mix, and musicians can select between them at the flick of a switch.

Enhancing Headphone Sound

Two major sources of dissatisfaction with headphone monitoring are the quality of the low-end response and the distorted spatialization of headphone soundfields (“in-each-ear” and “inside-the-head”), which many performers find uncomfortable. Regardless of the frequency response, headphones do not convey the strong sensation of bass of loudspeakers, which some musicians (like bass players) demand, because low frequency perception is more physical than aural. Bass notes are conducted through bones in the body, and merely hearing them lacks impact. In that case, headphone monitoring is still an option if supplemented with vibration transducers such as “shakers” and subwoofers, which add a physical sensation to bass. See A Quick Guide To Headphone Accessories for more information about vibration transducers.

The distorted perspective of headphones can be mitigated by first processing the mix through an acoustic simulator such as a crossfeed filter. Where crossfeed processing is not sufficient, an auralization processor (virtualizer) applies more complex processing to achieve true 3-D spatialization. Virtualizers were once implemented with expensive computers and software, but are now available in consumer audio gear. They can be added as an outboard to an existing monitoring system. Acoustic simulators are sold separate devices or as components of headphone amplifiers, of surround sound decoders and even as accessories with headphones. Many PC sound cards feature 3D sound outputs for headphones. Be careful to distinguish between acoustic simulation for for headphones and for loudspeakers (acoustic simulation for loudspeakers generates surround sound from stereo loudspeakers). Some in-ear monitors, such as AKG’s IVM1, have a built-in virtualizer. For more information about acoustic simulators, see A Quick Guide To Headphone AccessoriesAn Acoustic Simulator For Headphone Amplifiers and Technologies for Surround Sound Presentation in Headphones .

STUDIO MONITORING AND MIXING

More Monitoring Options For Performers

In the studio, setting up the monitor mix for performers may be slightly less complex, since the performance is not live. Most of the principles of setting up a mix for a live performance still apply. In this less formal atmosphere, there is more flexibility in configuring the headphone system. If hearing conservation is not an issue, then open-air phones are more comfortable than closed-ear types. However, if played too loudly, open-air types are prone to leak or bleed sound into microphones, so should be offered to performers as a second choice. Remote mini-mixers, which could be a distraction onstage, are a blessing-in-disguise in studios and let musicians instantly customize their mix, thus freeing engineers to focus on other things.

Vocalists are most comfortable hearing their own voices. Where hearing conservation is not an issue, vocalists can monitor with open-air headphones (again, being careful to avoid sound bleeding into the mike). Closed-ear headphones are also workable, with one earcup off the ear to let in ambient sound (mute the channel to the floating earcup to minimize sound bleed). If a vocalist wears in-ear monitors or closed-ear headphones without compromising the acoustic isolation, avoid making the voice so prominent in the mix that it sounds close-miked and unnatural – which can then cause singers to restrict their sound. In particular, vocalists who use in-ear monitors (canal-type headphones) can hear their own voices very clearly due to the occlusion effect. A compressor in the mix can help in these situations. (Apparently, some engineers raise the level of the vocalist’s mix as a natural form of compression for vocalists.) If vocalists report that they cannot hear themselves in headphones, try reversing the phase on the microphone to see if the vocalist’s voice is in phase with the voice in the headphones.

Acceptance From Recording Engineers

Headphone monitoring is also gaining converts among recording engineers, many of whom have discovered the advantages of monitoring with headphones over loudspeakers. From the console operator’s point of view, the soundfield of headphones is more detailed, so that any problems in a mix are easier to spot. However, engineers will often draw a line between using headphones for tracking and for mixdown. Headphone mixes can sound terrible when played back over loudspeakers, due to the different characteristics of the soundfields such as frequency response, interchannel crosstalk and spatialization.

Whatever the reason (hearing conservation, budget, equipment, preference), there are success stories about mixdowns done through headphones. While it isn’t easy to correlate headphone sound with loudspeaker sound, it can be done. An understanding of psychoacoustics is a good beginning. Good mixdowns with headphones are a matter of practice (and a few tips don’t hurt either). Of course, the final result should always be checked on loudspeakers.

The Challenge of Mixing with Headphones

The close proximity of headphone transducers to the ears affects how the audio spectrum is perceived. The lack of physical sensation of deep bass in headphones was discussed earlier. Headphones also tend to be brighter than loudspeakers, because the air attenuates high frequencies from speakers before they reach the ears. Headphones direct all sounds straight to the eardrums, bypassing the acoustic shaping that occurs when sound interacts with the listener’s head. Many headphones are now “diffuse-field” equalized so that they sound flat from within the ear canal – although that equalization is based on an average head shape and may not be a good match with every listener. For more information on HRTFs and diffused-field equalization, see A 3-D Audio Primer and A Quick Guide To Headphones.

Loudspeakers play in a real acoustic space. Headphones sound artificial because each audio channel is isolated to one ear. Sound waves from loudspeakers interact with each other (interchannel crosstalk), with wall reflections and with the listener’s head before they reach the ears. The resulting soundfield is a complex amalgam of phase-shifted amplitudes, which may amplify, cancel and/or delay select frequencies. It is impossible to determine through standard headphones how the phase and amplitude variations in one audio channel will affect another when played back over loudspeakers. Consequently, an otherwise smooth headphone mix can have a decidedly rough quality when heard through loudspeakers.

Acoustic simulators can improve the distorted perspective in headphones. Even the simplest of acoustic simulators, a crossfeed processor, can recreate interaural crosstalk in headphones for a more natural presentation. Beyond mere crosstalk is the whole issue of true spatial perception. Binaural recordists must monitor with headphones to hear spatial information, but the narrower soundfield of headphones can result in regular stereo mixes sounding almost monaural over loudspeakers. Until surround sound came into vogue, few audio engineers spoke of mixing for true spatial placement. Yet, with a good microphone configuration to capture localization cues, a stereo soundfield from quality loudspeakers can reproduce a sense of spatial depth and height as well as left-right width. The average headphone is not capable of re-creating these spatial artifacts in a stereo recording.

In terms of spatial editing, engineers have had a limited set of spatial options for stereo recordings: pans, delays, reverberation and other special effects. Headphones are perfectly good for auditioning spatial effects, unless the effects phase shift signals so that they sound different when heard over loudspeakers. And of course, standard headphones trap the soundfield in a straight line between the ears, so are of little value for directing placement of voices and instruments in a 3-D surround field.

Engineering A Realistic Acoustic Environment In Headphones

While many of the same techniques for making headphones sound more natural to musicians are easily adapted to recording engineers, the importance of having a close correlation between headphone and loudspeaker sound demands careful selection of headphone equipment and application of techniques. First and foremost, if they don’t already own a pair, engineers should audition diffuse-field equalized headphones, which are designed to sound flat inside the ear canal. Diffuse-field equalization is a fairly common product feature nowadays (for example, many of the AKG and Sennheiser phones are diffuse-field equalized).

If more than one engineer is participating in a session, giving everyone the same brand and model of phones set at the same gain will help with consistency of perception (if not peace). Closed-ear phones and in-ear monitors have the clearest and most extended reproduction, while attenuating ambient noise. Also, phones with good acoustic isolation are better for monitoring low-level (85-90db) mixes, which sound better on consumer audio systems. See A Quick Guide To Headphones for more information about diffuse-field equalization.

When monitoring with headphones, Fred Ginsberg of the Equipment Emporium recommends learning to set levels by ear instead of eye. The headphone level should be adjusted so that the 0 VU reference tone sounds as loud in one’s head as a loud telephone conversation – uncomfortable, but NOT painfully loud. Shouts and emphasized vocals should only briefly jump into the zone on a VU meter. LXH2 in his article Thoughts And Processes On Mixing With Headphones suggests the use of a tuned bass circuit when mixing low frequency content.

Whether or not the phones are diffuse-field equalized, binaural recordist Ron Cole suggests equalizing headphones with the biophonic curve (shown above) as a guide to compensate for ear canal resonances and other spectral differences between loudspeakers and headphones. Biophonic EQ (as well as any other signal processing mentioned below) is for listening purposes only, so the equalizer should be inserted just before the headphone amplifier. The biophonic curve is only a guide, and experimentation is encouraged. For more information about the biophonic curve, see Taking Audio In Another Direction.

Most headphones tend to image between the ears or in the back of the head. A binaural microphone system can help to create a realistic headphone sound field (see Thoughts And Processes On Mixing With Headphones). A simple technique for pulling the soundstage forward with supra-aural (on-the-ear) phones is to wear the earcups slightly lower and forward on the ears. Try out various positions to get the best localization and depth. The goal is to get the sound to enter the ears at an angle and engage more of the HRTFs of normal hearing. Unfortunately, this trick does not work as well with circumaural phones, which are designed to remain in a fixed position on the ears.

Acoustic simulators (crossfeed filters and virtualizers) electronically recreate the properties of a true acoustic space in headphones. The inability to hear interchannel phase effects on a recording is a major obstacle to using headphones for mixing. A simple crossfeed processor can mimic this aspect of acoustic space by introducing crosstalk between channels, so that phase effects can be heard. In fact, a good crossfeed processor tries to avoid overemphasizing phasey artifacts on recordings. At the same time, the processing smooths out the sonic image inside the head – no more stereo echos bouncing off the ears or holes in the soundstage. Moreover, by reducing the exaggerated stereo effect of headphones, crossfeed can help produce a better stereo mix.

For the most part, crossfeed simulators are electronic devices, but there are other options: headphone designs that provide an acoustical form crossfeed (such as the AKG K1000s) and PC-based crossfeed applications. See A Quick Guide To Headphone Accessories and HeadWize Projects Library for more information on crossfeed processors.

Note: Spatial enhancers are circuits that phase-invert the crossfeed to achieve a more spacious sound in headphones by adjusting the amount of ambient sound (such as reverberation) in a recording. Because these types of acoustic simulators dramatically alter the phase of the crossfeed, they are NOT suitable for checking interchannel phase interaction in recordings. In general, audition and experiment with acoustic simulators to become familiar with the characteristics of the sound fields that they create.

If an engineer insists on the utmost realism when mixing with headphones, then a headphone virtualizer (auralization processor) may be just the ticket. Virtualization takes acoustic simulation a huge leap beyond crossfeed to completely externalize the headphone soundfield outside the listener’s head. Virtualizers simulate a virtual, loudspeaker array inside regular headphones. Virtual reality headsets would not be convincing without them. For example, a stereo virtualizer will simulate a soundfield of two in-front loudspeakers. A surround virtualizer recreates five (or more) virtual speakers around the listener’s head.

Virtualizers also let the listener adjust many of the acoustic characteristics of an audio signal to mimic a variety of acoustic spaces – ranging from a large, reverberant concert hall to a small nightclub (add a vibration transducer to enhance the physical sensation of low frequencies for greater realism). Some virtualizers may be calibrated for each listener’s head-related transfer functions to generate highly accurate spatial cues. There are also special headphones with motion sensors (or standard phones with an add-on motion sensor) to vary the perspective of the sound field as the listener’s head moves.

Acoustic simulators are sold as separate devices, as a feature of headphone amplifiers and surround sound decoders, as headphone accessories, and more recently, as plugins for PC-based music players. Many PC sound cards offer 3D sound outputs for headphones, based on technologies from companies such as Aureal CorporationSRS Labs and Creative Labs. (Note: Be careful to distinguish between acoustic simulation for for headphones and for loudspeakers. Acoustic simulation for loudspeakers generates surround sound from stereo loudspeakers). There are also advanced PC cards and software (from companies like Lake Technology and WaveArts, Inc.) that turn PCs into full-blown acoustic modelling workstations.

When purchasing an acoustic simulator and, in particular, a virtualizer, careful and extensive auditioning is a must, as the quality of the image depends heavily on how well the processor approximates an individual’s HRTFs. For example, head-movement tracking may be critical for 3-D hearing for one person, but not another. Check the environmental adjustments to make sure that the various simulations are realistic. All Dolby Headphone virtualizers are pre-configured to simulate a Dolby Reference Room 1, which acoustically models a small, well-damped room appropriate for both movies and music-only recordings. Additionally, DH virtualizers may simulate a Room 2 (a more acoustically live room particularly suited to music listening) and/or a Room 3 (a larger room, more like a concert hall or movie theater).

QSound iQfx 2.0 plugin for Real players.

Fortunately, the popularity of PC-based music players (especially MP3-type players) has spurred the creation of low-cost and no-cost acoustic simulation plugins from most of the major 3D technology companies. QSoundSRS Labs and Lake Technology are just a few of the companies that have written plugins for PC-based music players such as Winamp and RealPlayer. These plugins are an inexpensive (often free for evaluation and less than $30 to purchase) means of evaluating the various competing 3D sound processing algorithms. However, poor performance from a plugin by a particular vendor does not necessarily reflect on the performance of any hardware implementations of that same technology.

Prices have fallen on hardware simulators as well. New (and improved) consumer devices, such as the Sennheiser DSP360, are appearing with MSRPs of less than $100. AKG’s Hearo wireless headphones have virtualizer circuitry inside the wireless transmitter. The Hearo 999 Audiosphere is full-featured enough for studio use. For more information about acoustic simulation and virtualizers, see A Quick Guide To HeadphonesA Quick Guide To Headphone AccessoriesA 3-D Audio Primer and Technologies for Surround-Sound Presentation in Headphones.

Engineers involved in surround-sound mixing have an alternative to virtualizers: surround-sound headphones. The standard 4-channel headphone, run directly from a 4-channel amplifier, does not produce a realistic surround field, because it does not integrate the listener’s HRTFs. Research indicates that these phones could sound more like loudspeakers with interchannel crosstalk and a small delay to the rear channels (around 30 to 50 ms) and the crosstalk feeds (5 to 10 ms). Another non-electronic solution for virtualization is Sennheiser’s Surrounder sound collar, which projects a sound field around the listener’s head. However, the Surrounder provides no acoustic isolation and is a different experience in feel and fit from headphones. For more information about surround headphones (including recent developments in 4-channel phone design), see A Quick Guide To Headphones and Technologies for Surround-Sound Presentation in Headphones.

Addendum

8/15/98: added section re: technique for improved in-front localization with supra-aural headphones.

7/10/99: updated section Engineering A Realistic Acoustic Environment In Headphones.

12/18/99: revised sections discussing acoustic simulation technologies.

5/1/00: added discussion of using Dolby Headphone from a mixing console.

8/4/00: added section on acoustic simulation plugins for PC-based music players.

References: Much of the research for this article came from a long (and laborious) review of audio newsgroups, where gabby engineers enthusiastically share their tips and techniques. There is no way to list them all, so I must thank them en masse.

__, In-Ear Monitoring, Garwood Communications c. 1997.
__, How To Mix In-Ear Monitors From The FOH Console, c. 1997, Garwood Communications.
__, How To Mix In-Ear Monitors From The Monitor Console, Garwood Communications c. 1997.
__, What The Pros Say About Garwood In-Ear Monitoring, Garwood Communications c. 1997.
Frink, Mark, “Monitor Lessons,” Mix, January 1998.
Ginsberg, Fred, Headphone Levels”, c. 1999
McCale, Steven, “Earphone Monitoring,” Mix, May 1996.
Santucci, Michael, PERSONAL MONITORS: What You Should Know , Mix, May 1996 (republished at Sensaphonics site).
Santucci, Michael, Musicians Can Protect Their Hearing, Sensaphonics Hearing Conservation, c. 1997.
Santucci, Michael, PROTECTING THE PROFESSIONAL EAR: Conservation Strategies And Devices , Sensaphonics Hearing Conservation, c. 1997.
Schulein, Robert, “Dolby Surround,” Mix, November 1987.

c. 1998, 1999, 2000, 2001 Chu Moy.

Judging Headphones For Accuracy.

caliper_head.gif

Evaluating headphones for accuracy can be difficult, because they present an unusual acoustic environment in which human ears were not designed to operate. The perception of 3-dimensional sound depends on the acoustic contouring that occurs when sound interacts with a listener’s head and outer ears. Because headphone transducers are pressed against the ears and because most stereo recordings are designed for playback over loudspeakers, this acoustic contouring is missing in headphone reproduction. The result is that the sound field has a “trapped inside the head” perspective, with a less than smooth frequency response. There are devices that attempt to mimic 3-dimensional hearing called acoustic simulators, but headphones can be judged for “accuracy” on their own terms.

The Language of Accuracy

The acoustic space in headphones is, by its very nature, distorted. However, audio reproduction is more than recreating an acoustic space. It is also the faithful reproduction of vocal and instrument qualities such as timbre, dynamics, transients and musical details. Acoustic simulators (and to some extent, binaural recordings) are becoming more widely available, so any discussion of headphone accuracy can include spatial fidelity, but the performance of acoustic simulators varies according to the listener, simulator circuitry and the headphones used. Without acoustic simulation, accuracy in headphone listening must be a viewed as a different concept from its loudspeaker counterpart.

In order to discuss issues of accuracy, there must be reference points in the language to convey analyses and descriptions. The world of audiophiles is replete with vivid adjectives that attempt to codify a universal audio quality vocabulary. Sound can be “steely” or “piercing” or “ice-cold” or “chocolatey” or “silky” or “grainy” or “yin” or “yang” and the list is endless. Does sound that is “steely” to one person come across as “silky” to another? Given the divergent reviews of audio equipment, these contradictions are a fact of life.

At the very least, a person who reads equipment reviews or gets buying tips from acquaintances should verify these assessments independently. When speaking of the sound of headphones, the following keyboard-frequency chart and list of terms from The Abso!ute Sound’s Guide to High End Audio Components (1994) conveniently divide the audio spectrum into bands:

keyboard.gif

Extreme bottom below 32 Hz
Low bass, bottom octave 20 to 40 Hz
Midbass 40 to 80 Hz
Upper bass 80 to 160 Hz
Lower midrange 160 to 320 Hz
Midrange 320 to 2,560 Hz
Upper midrange 2,560 to 5,120 Hz
Highs, lower highs 5,120 to 10,240 Hz
Extreme highs, top octave 10,240 to 20,000 Hz

Thus, if a review says that the midrange on a pair of headphones is “recessed,” the reader will know where in the audio spectrum to seek out this anomaly. These terms are not hard and fast rules, and perhaps when the reviewer specifies the midrange, it is actually the lower portion of the upper midrange that is under scrutiny.

Accurate Sound vs. Better Sound

 

hrtf.gif
Figure 1: Graph of a head-related transfer function.

Headphones that have built-in equalization to sound “flat” do not necessarily measure flat. Figure 1 shows how the ears hear a flat frequency sweep projected from a loudspeaker positioned slightly left of center. The hills and valleys in the two curves are due to the head-related transfer functions (HRTFs) that shape sounds interacting with the listener’s head and the pinna of the ears. The brain processes the amplitude and phase-shifts from the HRTFs to determine the nature and location of the sound. Diffused-field equalized headphones alter the frequency response of headphones to resemble a curve similar to those in figure 1, thereby restoring some of the HRTF contouring that is normally missing with headphones. Thus, diffused-field equalized headphones are supposed to sound natural and flat.

What about equalization that enhances the sound of headphones – not to make them sound flatter, but “better?” For example, some headphones are equalized with a treble boost. This high emphasis can make headphones sound detailed and balanced. In other models, the sound “sizzles,” which, though artificial, many people find initially pleasing. As with loudspeakers, headphones that appeal to consumer excesses may find their way into homes more quickly, since they sound impressive in fast-paced evaluations. Al Fasoldt, Technofile columnist for the Syracuse Newspapers, commented on headphones that sound better by design:

[I]n this age of fast food and quick desires, accuracy can be dull. Engineers who mix rock albums already know that and they usually add a punchy mid-range even before it gets to your amplifier. So the punch whether it’s added before or after can often add some life to the sound. That’s not bad, but it’s not necessarily good, either.

Take the case of a chef, for example. If the chef added pepper to everything, the food would taste pretty tangy all the time. You’d get tired of it fast…. And of course you’d hardly think the food tasted natural.

But back to headphones. You’ll probably be a lot happier with an accurate-sounding model than one that’s been spiced up. Unless the engineer’s twiddling has been nixed in the mix, you’ll end up with a double boost-once at the studio and once alongside your head.

The old adage “All that glitters is not gold” is equally applicable to aural and visual glitter. Accurate headphones may not make one’s ears prick up in excitement, but in the long run, the two will settle into a comfortable and lasting relationship.

construction.gif

Finding Low-bass Notes in Headphones

How well headphones reproduce bass depends on several factors such as whether the type of phones is closed-ear, open-air or in-ear. The open-air types tend to leak sound, so the bass may not be as defined as with closed-ear types. Canalphones that form an acoustic seal inside the ear canal have excellent clarity, but the close-coupling tends to highlight the lighter quality of headphone bass. Compared to loudspeakers, the best headphones seem to be bass-shy, regardless of how far down the frequency response extends. Headphone manufacturers may incorporate a bass boost to compensate. The overall response of a pair of phones can sound flat, yet the low notes lack heft, so these “mega” boosts appear to be justified.

When a person hears bass notes, the experience is both aural and visceral. The feeling of bass notes as they are conducted through the body contributes to the sensation of bass. Headphone listening limits the experience to the ears alone. In that respect, standard headphones will always have less dramatic bass than speakers (canalphones more so, since they radiate bass straight to the ear drums with little or no bone conduction). Therefore, it is important to distinguish between truly anemic bass and bass which feels light from the lack of physical impact.

The lightweight headphones that come with portables are likely to have truly anemic bass. When lightweight headphones have transducers too small to reproduce low bass, a boost in the midbass response can create the illusion of deep bass and a more balanced sound. Al Fasoldt recounts an incident that illustrates how the brain can synthesize missing bass notes:

I watched [a utility crew put in new light poles] as I listened to an old organ recording. The performance was full of extended pedal notes, down to 16 Hz. That’s low enough to qualify as thunder. After two or three minutes I began to realize that I was hearing more than just the usual Walkman-type sounds from my tiny headphones. As the organist stepped down onto the pedals, rumbles of the deepest imaginable bass poured into my ears. I became so excited that I tripped over the headphone cord….

With my headphones dangling near the ground, I was still hearing those bass notes. One of the workers walked over to the diesel generator and switched it off, and my bass notes disappeared. I had been hearing an ordinary engine’s chug-chugs and interpreting them as Bach’s mellifluous pedal-point. Frequency-response checks on my portable player’s headphones showed them to have almost no bass at all. And yet the bass sounded fine when I listened to music….

The solution to this mystery of the missing bass notes is found in the study of psychoacoustics. Music is a complex mix of different frequencies superimposed on each other. The sounds from a voice or an instrument are characterized by a fundamental tone and a series of overtones or harmonics that are mathematically related the fundamental tone. Thus, two different musical instruments are said to have the same pitch, if they are playing the same note. Yet, the sound of a violin is clearly distinguishable from that of a trumpet. They have different timbre because the harmonics from each instrument are not the same.

A fundamental note and its harmonics are mathematically related such that the brain can actually synthesize a missing fundamental note, so long as the harmonics are audible. The exact psychoacoustic process for this synthesis is still a subject of debate, but may be a combination of recognizing the timing intervals and the patterns of harmonics. Thus, low bass sounds from headphones that cannot physically reproduce low bass are nothing more than an illusion. Further, this synthesized bass often has a lighter timbre in comparison with the real thing. Nevertheless, it is this illusion that continues to drive the portable stereo industry. Lightweight headphones that could not create the sensation of low bass would probably have been doomed to commercial failure. For more information about the Missing Fundamental effect, see The Elements of Musical Perception.

vibes.gif

Musicians who monitor with headphones often miss the “tactile” feel of instruments such as thump of drums or the pull of a bow across a violin, which high volume loudspeaker monitors can convey. One of the benefits of headphone monitoring is being able to listen to a mix at safe volumes to preserve hearing. Vibration devices such as subwoofers, “shakers” and other tactile sound transducers can supplement headphone sound to create the physical sensation of high volume sound systems without the risk of hearing damage. Vibration transducers mount on floors and furniture (wherever there is contact with human bodies) and vibrate in synchronicity with the low frequencies in music. The vibrations travel in the body via bone conduction. The brain integrates these vibrations into the listening experience. Some headphones have mini-shaker transducers mounted on the sides, but these do not generate the same levels of deep, body-rumbling bass. See A Quick Guide To Headphone Accessories for more information about vibration transducers.

Tools of the Trade

The best tools for evaluating headphones are music CDs and audio test CDs (skipping those tracks that are specific to loudspeakers, such as speaker-setup tests). A good headphone test CD should have pink noise tracks, an assortment of frequency sweeps and/or chromatic scales as well as binaural tracks (or use separate binaural recordings). The music CDs should contain music that has not been overly processed. Recordings that exaggerate sounds by design are difficult to use as an accuracy reference. When listening to binaural recordings, be aware that closed-ear and circumaural headphones are usually better for binaural playback than other headphone types. Also, diffuse-field equalized headphones work best with binaural recordings that are made with the microphones mounted on sides of the dummy head, but not inside the artificial ears.

test_cds.gif

Listening to audio test CDs can be excruciatingly uninteresting, but are the best means for evaluating the individual nature of headphone performance (and are especially valuable for evaluating equalizers). Of the standard series of test tracks, the pink noise, slow spectrum sweeps and chromatic scales are probably the easiest to use. Listen to the pink noise tracks standing close to a reference-quality loudspeaker to have a baseline for comparing headphone sound. In a balanced system, the pink noise will sound smooth, with no frequencies standing out. Since loudness perception is frequency and volume dependent (see discussion below), listen to the pink noise at different volumes. This test will point out any rough areas in the headphones’ audio spectrum. The warbles and sweeps can identify any mechanical resonances in the headphone earcups.

If a headphone acoustic simulator or virtualizer (such as Dolby Headphone) is part of the audio system, listen to test CDs with and without the simulator engaged. However, be prepared for the result that the simulator may perform better with music than with test signals. The signal processing in acoustic simulators (and especially virtualizers) is sophisticated and based on psychoacoustic theory, and might be designed to work with complex and dynamic audio sources such as music. Thus, the best way to evaluate headphone acoustic simulators may be to compare their sound with loudspeaker playback.

diffuse.gif

Since measuring the frequency response of headphones by ear is difficult, those who are truly serious about evaluations might try measuring the response using an acoustic coupler (or better yet, an artificial ear). Headphone couplers (also available for earphones) are sold by such audiometric supply companies as Digital Recordings and must be fitted with a microphone connected to a real-time spectrum analyser (most equalizers have a built-in RTA and there are computer programs that turn PCs into spectrum analyzers). If a RTA is not available, the response curve could be manually plotted with a sound level meter from Radio Shack (33-2050 or 33-2053). Be sure to correct for any resonances from the coupler itself or the readings will be inaccurate. Diffuse-field equalized headphones will measure flat only from inside the ear canal of a model dummy head or an artifical ear. The procedure for measuring a diffuse-field equalized headphone is defined under the IEC 60268-7:1996 standard.

fletcher.gif

If test equipment is not an option for determining the frequency response (or as a follow-up to measurements), listening to test tones can still provide a useful assessment of how “flat” headphones sound. The Fletcher-Munson curves above indicate that loudness perception is a function of frequency and sound pressure levels. Unfortunately, the curve is flattest when the loudness is at the threshold of pain. Yet, there is a band of perception that fluctuates only about 5dB over a range of safe listening levels: 200Hz to 5kHz. Keeping in mind that loudness comparisons between frequencies are prone to subjectivity, tone tests between 200Hz and 5kHz can help narrow down irregularities in the response after a pink noise check. As each tone is played, check for variations in volume and character between tones. Again, if a headphone acoustic simulator is part of the listening system, check the frequency response of headphones with and without the simulator engaged.

Musical passages on test discs are usually selected to demonstrate a particular aspect of sound reproduction, and are more revealing of how headphones will perform with complex audio signals. Uncompressed vocals and instrumentals and special effects tracks can stress the dynamic range of headphones, but take care with the volume setting to avoid hearing damage. One safer way to test the headphones’ ability to play loud transients is to wear foam ear plugs to protect the ears while listening to the headphones. Get the kind that are rated to reduce noise by at least 29dB and absorb frequencies evenly across the audio spectrum. Another option is to hold the headphones off the ears at a comfortable distance and listen for distortion. The sound will be tinny and devoid of bass, but if the headphone transducers are poorly damped or overload, the distortion artifacts may be distinguishable. Do NOT pump too much volume or the phones could self-destruct (not to mention the potential damage to one’s hearing)!

The imaging on the binaural tracks depends on how closely the shape of the listener’s head matches that of the dummy head used to make the recording and on the positioning of the microphones on the dummy head. Further, some listeners localize sound with the help of head movement, which is not possible with standard binaural playback. Therefore, if a headphone does not reproduce binaural sound to stunning effect, the accuracy of the headphones may not the cause. Audition binaural tracks on many different headphones to get the best spatial reference. Acoustic simulators should be turned off when listening to binaural recordings. Also, closed-ear and in-ear phones tend to be better for binaural playback than open-air types.

Judging headphones for accuracy is not an intuitive process, but unlike loudspeakers that have to be moved around and positioned for optimal sound, can be done with hardly any exertion. The most important preparation is understanding the psychoacoustics of headphones, so that the listener does not misjudge the unique characteristics of headphone sound fields. Listening to headphones is an experience that can be as satisfying as listening to loudspeakers, if the listener evaluates the phones carefully before purchasing.

Addendum

12/14/98: Added discussion of using acoustic simulators during headphone evaluation.

12/18/99: Updated discussion of using acoustic simulators during headphone evaluation.

References:
__, Sensaphonics Product Literature: Bass Shakers, Tactile Sound Transducers, c. 1997, Sensaphonics Hearing Conservation.
Darwin, Chris, Perception, c. University of Sussex at Brighton.
Fasoldt, Al, “How (and why) to choose good headphones,” c.1988, The Syracuse Newspapers.
Fasoldt, Al, “Generating low-bass notes in my headphones and in my head,” c.1988, The Syracuse Newspapers.
McCale, Steven E., “Earphone Monitoring,” Mix, May 1996.
Murthy, V.S.Madhuri and Sethi, Simer Singh, Basic Acoustics and Psychoacoustics, c. 1996.
Pearson, H., The Abso!ute Sound’s Guide to High-End Audio Components, c. 1994.
Welsh, Norma, Demonstrations in Auditory Perception, McGill Univerity c. 1996.

c. 1998, 1999 Chu Moy.