DIY Headphone Measurement Contest Winners! Best Measuring

Best Measuring DIY Headphones
The big surprise here was how good some of these headphones performed. Clearly superior to the headphone prior to modification; and in a few cases measuring as well as good commercial cans. They all did seem to display some oddity, however, and this is where having DIY cans measured can really be helpful as they might point towards places small improvements can be made.

Judging this category wasn't easy. What I looked for primarily was well extended bass; flat frequency response to 1kHz and good linearity of the 30Hz square wave top; and lack of significant overshoot or ring in the 300Hz square wave, which typically indicates a smooth sound. I think the big surprise for me was how well many of the really old Ortho cans performed. It saddens me that these makers quit developing and producing these fine headphones.

One more note: None of the Grados made the cut here. Despite Kojaku's excellent efforts, even his Grados didn't have the heft in the low frequencies that the Orthos did.

5th Place - Yamaha HP-50A by dBel84

These small, on-ear, sealed headphone performed remarkably well, with a surprisingly flat response from 100Hz to 1kHz, and without any large peak in the treble. However, bass response was not well extended, and positioning on the ear was quite sensitive. It also looks like there may be a leak somewhere in the left earpad seal, which can be seen in the poorer response of that channel in the bass of the FR and THD+noise plots, and in the excessive bowing in the top of the 30Hz square wave.

Fifth place prize in this category gets one of my old PX 200 headphones, and ... oh my goodness ... dBel84 is in it again! Man, that guy made a haul here today. I guess that's what you get for sending in six DIY headphones for the contest.

4th Place - Fostex T20RP by scompton

My goodness, these T20RP headphones just shouldn't measure this good. Here's a terrific example of scompton doing just enough to aid the headphone without overdoing it.

Unfortunately, you can't make a silk purse from a sows ear, and these measurements do show some areas of underperformance. Overall response is somewhat mid-centric with a broad humped frequency response from 10Hz to 6kHz. A strong spike at 8-10kHz indicates a bright spot in it's response, and a moderate overshoot on the 300Hz square wave echos this. The downward slope of the 300Hz square wave top after the initial overshoot typically is heard as lean sound (not punchy).

You can read more about scompton's DIY efforts here.

Fourth place prize is the last of my old PX 200 headphones. As a testimony to the durability of these dandy little cans, this pair has no doubt spent hundreds of hours with four kids in the back of my van and continues to operate flawlessly. (No dried up rasins found within upon inspection ... but you might want to have a second look.)

3rd Place - Yamaha YH3 by rhythmdevils

This is the second Yamaha headphone to make it into the best measuring top five. Rhythemdevils' modifications show that this over 30 year old design had great promise ... it seems unfortunate to me that Yamaha didn't continue to develop it's Orthodynamic line to deliver on the surprisingly good performance rhythemdevils has shown them to possess.

Bass extension and frequency response linearity is very good to 1kHz for an ear-pad headphone, which is also shown in the nice flat top of the 30Hz square wave. This makes me think the ear-pads seal very well indeed, which is also indicated by the low level of THD+noise in the bass.

The only problem I have with these cans is the pronounced peak in the frequency response at 8-10kHz and in the strong overshoot of the 300Hz square wave. Coupled with the jagged impulse response, I tend to think these cans will be rather sharp sounding.

Third prize is a brand new, in-the-box pait of Sennheiser PX200-IIi headphones. Woot! Obviously, I've got a thing for PX 200s.

2nd Place - Fostex T50RP "Paradox" by LFF

I received about ten modified Fostex T50RP which used the original ear piece housings, and a few with housings replaced, which seem to perform better in some ways. It does seem to me that the stock housing is limiting the factor with LFF's "Paradox." I haven't measured any of Smeggy's new Thunderpants, but after hearing them I'm fairly certain we'll see significant performance improvement due to the better acoustic properties of the replacement earpiece.

LFF sent in three modified cans, all of which measured better than the other T50RPs sent in. Though the frequency response plots are a little rough looking, there are no large peaks or valleys below 4kHz. The bass extension is very good; it seems to me that his efforts at getting the earpads to work well has paid off handsomely. If you look at many of the T50RP plots, you'll see a fairly large feature somewhere between 60Hz and 100Hz. This is what I call the "main spring" of the headphones where it bounces on the springiness of the earpad cushions. One goal of DIYers should be to make/use earpads that are very squishy and damped to prevent resonance at these low frequencies. On LFF's headphones, you can see a small step feature just under 100Hz from this phenomena, but it's better than the other T50RPs measured, and about as good as that of Audeze's headphones. Nicely done.

The frequency response does start slowly rolling off at about 300Hz; ideally it should be flat to at least 1kHz. The large peak at 10kHz is a bit worrying, but you can see that the leading edge of the 300Hz square wave has a little wiggle but doesn't overshoot, which indicates a headphone that's not harsh. 30Hz square wave is a bit misshapen, but remains above zero over it's entire length, and the THD+noise remains under 1% to 20Hz, again indicating tight bass response into the lowest octave.

Bottom line: I think LF is squeezing about as much as can be squeezed from the stock ear piece housings of the T50RP. I don't think DIYers can expect much better performance from the T50RP with stock earpieces. The good news is that LLF's cans sound pretty good. A little uneven perhaps, but nicely warm and punchy, with a treble that's under control. With patience and practice DIYers can get quite good sound from this classic headphone.

Second place prize is a pair of Fostex T50RP headphones! Im pretty sure LFF knows what to do with them.

1st Place - Aiwa HP-500 by Kabeer

Wow! Spectacular measurements here. While one earpad was not sealing quite as well as the other, the frequency response is near ruler flat from 10Hz to 2kHz. Very, very nice! Likewise, the 30Hz square wave response has marvelously good shape. There is a peak at 10kHz, but it's smaller than most; and there is overshoot of the 300Hz square wave, but it's of modest size and relatively free of noise otherwise. The only thing troubling about these measurements are the rather noisy THD+noise plots. Not sure what to make of that.

The Aiwa HP-500 is a spectacular looking headphone. I didn't include it in my "Coolest Looking" category because they weren't of Kabeer's design, but I can tell you I lusted after these cans both for their great sound and superb styling. Color me jealous.

First prize is my pair of Audeze LCD-2 Rev 1 headphones ... which are currently in pieces in a box, but I'll be putting them back together this weekend right after I post this and get all the DIY measurements on the download page. Great job Kabeer!

Well, that's it, the contest is over, but remember U.S. headphone DIYers, I'm at your service. If you've got a pair of DIY modded headphones of any type, I'll be most happy to measure them for you. You ship them to me; I'll measure them and post the measurements; and I'll pay for the return shipping. Please remember to put your name, Head-Fi user name, email addy, return shipping address, and phone number on a piece of paper in the box. Just shoot me an email or PM and we'll get the ball rolling.

I'm looking forward to you future creations!