Head-Fi.org › Forums › Equipment Forums › Sound Science › Building a Headphone Measurement Lab
New Posts  All Forums:Forum Nav:

Building a Headphone Measurement Lab

post #1 of 355
Thread Starter 
Hi Guys!

As many of you know I was the founder of HeadRoom. I'm finished with my bit
there (you can read about it here) and am now moving on to my next thing.

I've got some ideas and am talking to Jude and some of the other folks who's input I want,
and since I shouldn't talk about what's likely to change, I'll just shut-up ...
... except to say I want to work for the hobby.

What's not going to change is that it will have something to do with measuring headphones. I spent a lot of effort when I was at HeadRoom to put a cool measurement system in place .... because at heart I'm a lab rat. I love instrumentation. I used to fix scanning electron microscopes, and once worked at Hughes Research Labs in Malibu, CA fixing vacuum pumps. That was fun, man did they have some cool gear in there.

Anyways ... HeadRoom and I came to a very cool arrangement, IMHO, in which I ended-up with the headphone measurement gear. I will continue to measure headphones for HeadRoom when they need on a contract basis, but I am also free to start something new using the equipment.


This thread is about building that lab.
post #2 of 355
This will be exciting.

I wish you all the best Tyll. You cannot fail at this hobby, this I know.
post #3 of 355
Me like.
post #4 of 355
Woohoooo! :can_can_dancing_girls_wearing_lab_coats:

Are waterfall graphs for headphones practical to do?
post #5 of 355
Thread Starter 
Originally Posted by Ham Sandwich View Post
Are waterfall graphs for headphones practical to do?

Certainly the time scale would have to be much shorter than speakers.

We'll have to give it a go one of these days, eh?

Got a long way to go though.

Let's see ... where to start.

One of the things I want to do is make the lab portable.

Well ... transportable. Which means the enclosure I do the testing in has to be about the size of a washing machine. I'll start building that tomorrow!

Well, it was weeks ago ... but you get the idea.

post #6 of 355
Sweeeeeeeet !

Will you be measuring headphone cables as well ?

That would be cool !

Good Luck !
post #7 of 355
oh yeah... I'm REALLY looking forward to this!
post #8 of 355
Well done!
post #9 of 355
Man, this sounds like an AWESOME project!!!

Good luck Tyll!
post #10 of 355
Best of luck Tyll, I look forward to your results. Will you be publishing a methodology? Should we expect all headphones to be tested on a identical system allowing for what I would guess to be a first for the headphone community on something of this scale?
post #11 of 355
How about putting a little into measuring amps, as well? There isn't nearly enough of that in this hobby.
post #12 of 355
Originally Posted by Uncle Erik View Post
How about putting a little into measuring amps, as well? There isn't nearly enough of that in this hobby.
Excellent input from U.E as always.

This Headphone measurement lab Tyll is doing will REVOLUTIONIZE the way people write reviews...
post #13 of 355
Tyll first off this is awesome! Having someone work full time providing information to the hobby to help sort out the chicken poop from the facts is invaluable.

Possibly more important than the raw data is the analysis of the data. Typically raw data is misinterpreted by under-informed but well meaning hobbyists. So just not measuring but providing insight to what the data could mean is where the value lies in my eyes. Amp measurements would be very cool as well.
post #14 of 355
Thread Starter 
Mornin' all! *sip*

Originally Posted by Nick Charles
Will you be measuring headphone cables as well?
Of course, I'll give it a try, but I really doubt we'll see anything. Not because I believe something isn't there, but because the nature of a headphone measurement is relatively complex and happens on a much grosser scale than the artifacts that would be observed with the differences in cables.

With cables, my guess it that you'd want to look at the smearing in time of the various low and high frequency components of the signal. The amount of smearing you get in a 10 foot cable run would be minuscule when compared to the frequency dependent artifacts that occur in the acoustic coupling of the headphones to the ear. (Concha reflections, ear canal resonances, etc.)

The equipment needed to characterize differences in cables would generally be much faster (capable of analizing stuff up in to the mega-Hertz range). The gear I've got is very good in terms of low noise and wide-band (audio signals encompas many octaves), but it's not high frequency.

Still, gotta prove that that's the case, so sure, we'll have a go at measuring cables.

Originally Posted by thisbenjamin
Will you be publishing a methodology? Should we expect all headphones to be tested on a identical system allowing for what I would guess to be a first for the headphone community on something of this scale?
You bet, and you bet.

The method in general will evolve from what we've been doing at headroom. There is impulse response and phase data that we've been collecting for years but not showing because I'm not sure if it's meaningful. One of the great things about doing this work now as more of a community project, is that there are some very accomplished engineers/scientists among the members, some of whom have already expressed an interest in helping to analyze the data.

The first goal is to ensure the raw data being gathered is a comprehensive and useful set. There's a myriad of issues to deal with on that issue alone. For exmple: How much resolution, accuracy, and freedom from noise does the data have to have. There's a tendency to say, "As much as possible," but that's a problematic answer because one technique for accomplishing that is to average many sequential measurements over time. Do you really need a 500 point sweep with +/-1dB repeatability on the measurement of how well a pair of headphones isolate from outside noise. Probably not.

Right now it takes about 40 minutes to get through the test sequence on a pair of headphones, I'd like to cut that down to about 10 minutes. Once we get a good data aquisition scheme set up, then the second half is developing some good methods for post processing the data to make it intelligible. This includes stuff like smoothing the data to further reduce noise; background subtractions for head-related-transfer-function, possibly analysing the data to come out with quality ratings like "neutrality" or "bass impact."

So, lots to do for both acquiring and analyzing the data.

The reason I have a goal of making the system transportable and faster when measuring headphones is I'd like to take it to large events like CanJam to get access to a bunch of headphones for measurement. I will be bringing it to CanJam this year, and will have it set up in my hotel room. It's important that the testing environment be fairly quiet. I'm currently reaching out to some folks who I know have cans that need to be measured to make arraingements and schedule measurements at CanJam. When the time comes I'll be publishing a schedule of those measurements, and will be requesting help to acquire certain other headphones for measurement. For example: R10, K1000, Leatherheads, Qualias, etc.

The problem is at 40 minutes a pop, and long setup times in the case of headphones that have their own amp, it might be very difficult to get more than 10/day done. So the current goal is to get the data acquisition part down pat and as fast as possible while retaining the desired accuracy and resolution. Post processing of the data will come after.

Maybe by next year we'll have it so dialed in that I can set up in a more accessible area and just measure headphones willy-nilly for folks.

Lots to do, eh?

Off to my morning walk, and when I get back, I'll post up the box build.

post #15 of 355
Wow, a new exciting venture!

If you are going to have a web-site, then here is my (very selfish!) wish-list:

1. Test results

Sensitivity in dB at 1mW
Frequency response
Impedance AND phase
Distortion FFT (from a 1mW input at 1 kHz)

2. Info store

Collect, store and make available documents relating to headphones, hearing research, audiology, etc. (i.e. a publicly accessible on-line reference library, a bit like Pete Millett does with his tube books).

Good luck with it all Tyll.
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Sound Science
Head-Fi.org › Forums › Equipment Forums › Sound Science › Building a Headphone Measurement Lab