shivohum
Keeper of the Quotes
- Joined
- Jun 21, 2001
- Posts
- 903
- Likes
- 12
The point of a beta is to get all the bugs out so the final product is more polished 
HeadRoom's New Web
- Thread starter Tyll Hertsens
- Start date
jphone
New Head-Fier
- Joined
- Jun 23, 2001
- Posts
- 41
- Likes
- 10
Tyll,
could you post some info on your test rig and procedures you use
to test. Also I think you need to provide more info on the graphs
as to what is being mesured and the units used.
could you post some info on your test rig and procedures you use
to test. Also I think you need to provide more info on the graphs
as to what is being mesured and the units used.
coolvij
Board geriatric
- Joined
- Jun 22, 2001
- Posts
- 2,139
- Likes
- 10
Lookin good, Lookin good - but i have to wonder if you'll ever give info on the only stuff i can buy, er, i mean, the cheap, er, ECONOMICALLY MINDED stuff......that is as good as the info u give one your high-end stuff....c'mon guys - PROVE TO ME YOUR HEADPHONE GEEKINESS!
JMT
JMT Audio:PPA/META42 Amp Factory
- Joined
- Jun 21, 2001
- Posts
- 6,733
- Likes
- 21
Was the purpose of the site redesign to send a subliminal message to my VISA card? If it was, it worked. Thankfully the site was a beta.
Great looking stuff Your Geekness, guess I really do need to follow the VISA prompt. Ok, I'll wait till the new site is up and running.
Great looking stuff Your Geekness, guess I really do need to follow the VISA prompt. Ok, I'll wait till the new site is up and running.
Neruda
Banned Moderator
- Joined
- Jun 20, 2001
- Posts
- 3,216
- Likes
- 11
I agree with apheared, pictuing the headphones on the neumann head was a fantastic idea. It's incredibly useful!
- Joined
- Jun 22, 2001
- Posts
- 4,111
- Likes
- 1,336
I have to tell you guys that I really am BURRIED at this red hot second as I am working directly on half the crap you're asking me for. If I answer too many posts, it will just take longer for the solutions to show up. But...
First, I intend to offer the service of measuring headphones for people for free IF I don't allread have the data. So, if there are headphones you want to see, but I don't carry them, you'll be able to ship them to me (your dime), I'll measure them, photo them, post the data, then send the cans back to you FedEx second day air (my dime) and include a catalog and a coffee cup.
That sounds like fun, ehh!
Then there are a couple of posts so far that I really should respond to as it's somewhat irresposible of me to put this stuff out there without giving you guys a leg up on getting the ongoing discussion rolling:
From jphone:
>>>could you post some info on your test rig and procedures you use
to test. Also I think you need to provide more info on the graphs
as to what is being mesured and the units used.
We use an audio precision system 2 cascade tester
http://audioprecision.com/products/index.htm
driving a new Max then to the cans. The cans are placed on Fritz ---the neuman dummy head binaural mike you see in the pictures (yes, that's a microphone).
http://www.neumann.com/infopool/mics...p?ProdID=ku100
The ouput from Fritz goes back into the AP. Fritz lives inside a big wooden box during testing. The box is a bit smaller than a chest freezer. It weighs about 300 lbs! Inside it has 2 layers of sheetrock all around on the inside then a couple of inches of sound absorbent matt, then Sonex foam anti reflection material. Total cost of this test rig (Box, Fritz, AP, Max, Computer, and Ccables: about $40,000.
He he he
From apheared:
>>>And a SPECIAL THANKS!! for whoever came up with the idea of photo'ing them while on the Neumann head...
That was me. And I thought of it after Ivy had photograph about half the cans! I couldn't walk for a week.
From shivohum:
>>>2) Why do some of the headphones seem to have much larger frequency response deviations on the new graphs than they did on the older ones? Sample variation?
We did all the measurements in the last two days befor we left for the show. The answer is "I don't know." Danny is on vacation right now. When he comes back July 1st, his first task is to re-measure everything for posting. Then we spend 3 months just measuring headphones and trying to DEEPLY understand them. I can't go into this at all right now, but suffice it to say that we have made some VERY interesting observations here and will be persuing them with vigor. Then we will re-measure all headphones again, maybe 4-5 months from now using testing proceedures that will be far more accurate at highlighting the performance of the headphones. We think we will be publishing AES papers on the techniques.
Once all that is done, then we'll do the "send in your headphone stuff." So, don't start asking for it soon.
On important note: Danny worked for about a week playing with the equipment to test the cans to ensure what we put up was useful and repeatable data. We can't say with a great deal of surity what the data means, but we can say that the relative accuracy of the data headphone to heaphone is quite good.
One of the goals of the project is to develope a background subtraction graph and apply it to all data before display. If we average the data from , let's say, the 15 best headphones we know of; then create a compensating graph to for the average flat; then apply the compensating curve to product data as it is served out; we should end up with something that looks a little more flat.
From Tim D:
>>>I think some graphs should be double checked though...like the Ety graphs. Also the square wave sampling doesn't seem aligned with one another when you compare making it confusing.
Good observation! Because the Ety relies on a seal in the ear canal, and Fritz doesn't have normal human ear canals the the acoustic impeadance comensation is not correct. We almost didn't measure them at all, but then stuck them in and they didn't look so bad as to be totally out of the picture, so we kept the data. There's an interesting story here though, we are developing a technique to provide a degree of absolute calibration to Fritz and it uses the Ety's because Etymotic has an increadable calibration lab and each Ety transducer is measured at the factory. On my way driving back from the show, I got a pair of Ety's and the calibration and compensation curves. But now I'm going to leave you hanging and say that I'll tell you how we do it when we really figure it out.
From Jude:
>>Would you kindly explain the various graph types? I'm guessing we all understand frequency response.
Oh CRAP! What in ten words or less?!
(I really shouldn't yell at the guy, I really do have to do this----and he did win that friggen naming contest)
Okay, hear goes:
Frequency Responce ... I think you guys know this one.
Sampled wave form: it's a square wave. Any rounding off of the wave form is loss of highs. And periodic wiggling is resonances. Basically, the cleaner the signal the better.
The missalignment of the waveforms is a triggering problem. When the final graphs are done though, this will be the graph that we'll use to detect whether the headphones are in absolute phase or not. Then the web site will calculate and set a flag so that one of the descrete items on the spec list will be an in phase/reverse phase spec.
SMPTE vs. Amp and CCIF vs Amp
It's intermodulation distortion products as you drive the headphones harder and harder. It's a measure of non-linearity in performance.
SMPTE uses two tones that are far apert and CCIF uses two tones close together.
FFT of 1Khz signal
Basically a frequency response plot of a single frequency. There should be one narrow bump at 1kHz. Any extra stuff is harmonics.
I'm sure some of you have some cool likes to places describing audio test basics.
Lastly, I did waste my day here .... workwize. But I'm still a geek ... headwize, and I'm stoked you guys like this. It's going to be real hard for me not to answer some things but I would very much appreciate it if you all would keep posting any comments and quyestions. I'll answer some of them, but I'll look at all of them. And it would be very helpful at this point, for all of us working on the thing to hear your comments. If you're luck we might even get Ivy and Matt to be sucked in for some answers.
I guess that's it.
First, I intend to offer the service of measuring headphones for people for free IF I don't allread have the data. So, if there are headphones you want to see, but I don't carry them, you'll be able to ship them to me (your dime), I'll measure them, photo them, post the data, then send the cans back to you FedEx second day air (my dime) and include a catalog and a coffee cup.
That sounds like fun, ehh!
Then there are a couple of posts so far that I really should respond to as it's somewhat irresposible of me to put this stuff out there without giving you guys a leg up on getting the ongoing discussion rolling:
From jphone:
>>>could you post some info on your test rig and procedures you use
to test. Also I think you need to provide more info on the graphs
as to what is being mesured and the units used.
We use an audio precision system 2 cascade tester
http://audioprecision.com/products/index.htm
driving a new Max then to the cans. The cans are placed on Fritz ---the neuman dummy head binaural mike you see in the pictures (yes, that's a microphone).
http://www.neumann.com/infopool/mics...p?ProdID=ku100
The ouput from Fritz goes back into the AP. Fritz lives inside a big wooden box during testing. The box is a bit smaller than a chest freezer. It weighs about 300 lbs! Inside it has 2 layers of sheetrock all around on the inside then a couple of inches of sound absorbent matt, then Sonex foam anti reflection material. Total cost of this test rig (Box, Fritz, AP, Max, Computer, and Ccables: about $40,000.
He he he
From apheared:
>>>And a SPECIAL THANKS!! for whoever came up with the idea of photo'ing them while on the Neumann head...
That was me. And I thought of it after Ivy had photograph about half the cans! I couldn't walk for a week.
From shivohum:
>>>2) Why do some of the headphones seem to have much larger frequency response deviations on the new graphs than they did on the older ones? Sample variation?
We did all the measurements in the last two days befor we left for the show. The answer is "I don't know." Danny is on vacation right now. When he comes back July 1st, his first task is to re-measure everything for posting. Then we spend 3 months just measuring headphones and trying to DEEPLY understand them. I can't go into this at all right now, but suffice it to say that we have made some VERY interesting observations here and will be persuing them with vigor. Then we will re-measure all headphones again, maybe 4-5 months from now using testing proceedures that will be far more accurate at highlighting the performance of the headphones. We think we will be publishing AES papers on the techniques.
Once all that is done, then we'll do the "send in your headphone stuff." So, don't start asking for it soon.
On important note: Danny worked for about a week playing with the equipment to test the cans to ensure what we put up was useful and repeatable data. We can't say with a great deal of surity what the data means, but we can say that the relative accuracy of the data headphone to heaphone is quite good.
One of the goals of the project is to develope a background subtraction graph and apply it to all data before display. If we average the data from , let's say, the 15 best headphones we know of; then create a compensating graph to for the average flat; then apply the compensating curve to product data as it is served out; we should end up with something that looks a little more flat.
From Tim D:
>>>I think some graphs should be double checked though...like the Ety graphs. Also the square wave sampling doesn't seem aligned with one another when you compare making it confusing.
Good observation! Because the Ety relies on a seal in the ear canal, and Fritz doesn't have normal human ear canals the the acoustic impeadance comensation is not correct. We almost didn't measure them at all, but then stuck them in and they didn't look so bad as to be totally out of the picture, so we kept the data. There's an interesting story here though, we are developing a technique to provide a degree of absolute calibration to Fritz and it uses the Ety's because Etymotic has an increadable calibration lab and each Ety transducer is measured at the factory. On my way driving back from the show, I got a pair of Ety's and the calibration and compensation curves. But now I'm going to leave you hanging and say that I'll tell you how we do it when we really figure it out.
From Jude:
>>Would you kindly explain the various graph types? I'm guessing we all understand frequency response.
Oh CRAP! What in ten words or less?!
(I really shouldn't yell at the guy, I really do have to do this----and he did win that friggen naming contest)
Okay, hear goes:
Frequency Responce ... I think you guys know this one.
Sampled wave form: it's a square wave. Any rounding off of the wave form is loss of highs. And periodic wiggling is resonances. Basically, the cleaner the signal the better.
The missalignment of the waveforms is a triggering problem. When the final graphs are done though, this will be the graph that we'll use to detect whether the headphones are in absolute phase or not. Then the web site will calculate and set a flag so that one of the descrete items on the spec list will be an in phase/reverse phase spec.
SMPTE vs. Amp and CCIF vs Amp
It's intermodulation distortion products as you drive the headphones harder and harder. It's a measure of non-linearity in performance.
SMPTE uses two tones that are far apert and CCIF uses two tones close together.
FFT of 1Khz signal
Basically a frequency response plot of a single frequency. There should be one narrow bump at 1kHz. Any extra stuff is harmonics.
I'm sure some of you have some cool likes to places describing audio test basics.
Lastly, I did waste my day here .... workwize. But I'm still a geek ... headwize, and I'm stoked you guys like this. It's going to be real hard for me not to answer some things but I would very much appreciate it if you all would keep posting any comments and quyestions. I'll answer some of them, but I'll look at all of them. And it would be very helpful at this point, for all of us working on the thing to hear your comments. If you're luck we might even get Ivy and Matt to be sucked in for some answers.
I guess that's it.
aos
May one day solve the Mystery of the Whoosh
- Joined
- Jun 21, 2001
- Posts
- 1,841
- Likes
- 13
No kidding, those graphs are useful. I assume dummy head has "ear canal" so this would be 'what you hear" measurements.
You can easily see why e.g. a simple 5 to 15 kHz RC filter (as given in a project on HeadWize) would make DT990 spectacular, why DT931 sounds fantastic as long as you don't hit the treble and why are HD580/600 so neutral and natural. Look at FFT - second harmonic is at -90dB for HD600, they look great here.
On the other hand, AKG-1000 has second harmonic at -50dB, and a few more quite visible ones?! That's not good at all for THD. Yet they still sound good?
They were not kidding, this must've taken many days to measure! Great work.
You can easily see why e.g. a simple 5 to 15 kHz RC filter (as given in a project on HeadWize) would make DT990 spectacular, why DT931 sounds fantastic as long as you don't hit the treble and why are HD580/600 so neutral and natural. Look at FFT - second harmonic is at -90dB for HD600, they look great here.
On the other hand, AKG-1000 has second harmonic at -50dB, and a few more quite visible ones?! That's not good at all for THD. Yet they still sound good?
They were not kidding, this must've taken many days to measure! Great work.
markl
Hangin' with the monkeys.
Member of the Trade: Lawton Audio
- Joined
- Jun 22, 2001
- Posts
- 9,130
- Likes
- 49
Hi Tyll,
I LOVE what you did with the graphs. I could play with that all day. (BTW: you can't compare the RS1 to the RS2, you get an error).
Some people tried to downplay the value of your graphs on another board, but to me, the mind boggles at how the audiophile headphone community could benefit. It just might force headphone manufacturers to try to produce phones with flatter response, if Headroom helps make that a buying criterion. Thanks a lot!
I am also surprised at how far from flat most of these phones appear to be. Very disappointing and eye-opening.
Now, what would REALLY REALLY cool would be the ability to compare the measurements of the headphone AMPS you carry.
One final question/comment: Why don't you carry Sony phones?
I LOVE what you did with the graphs. I could play with that all day. (BTW: you can't compare the RS1 to the RS2, you get an error).
Some people tried to downplay the value of your graphs on another board, but to me, the mind boggles at how the audiophile headphone community could benefit. It just might force headphone manufacturers to try to produce phones with flatter response, if Headroom helps make that a buying criterion. Thanks a lot!
I am also surprised at how far from flat most of these phones appear to be. Very disappointing and eye-opening.
Now, what would REALLY REALLY cool would be the ability to compare the measurements of the headphone AMPS you carry.
One final question/comment: Why don't you carry Sony phones?
Griffdog
New Head-Fier
- Joined
- Jun 23, 2001
- Posts
- 33
- Likes
- 0
I believe the two lines on each graph (red and blue) are the right and left channels (the graphs with four lines...ahh..your guess is as good as mine). You would have to ask Danny to be sure. When comparing headphone we are just using the red data from the individual graphs.
Great feedback so far... Keep it coming...
Griffdog (the HeadRoom web guy)...
Great feedback so far... Keep it coming...
Griffdog (the HeadRoom web guy)...
JohnActon
Put Ray Kimber Jr. through college
- Joined
- Jun 23, 2001
- Posts
- 220
- Likes
- 10
Very cool, Tyll - excellent site! It looks like a tremendous amount of work has gone into this endeavor. Your whole should be commended. I especially like the comparison graphs.
Eagle_Driver
Headphoneus Supremus
- Joined
- Jun 22, 2001
- Posts
- 6,496
- Likes
- 61
Now I noticed the frequency-response curves of two very popular Sony headphones... What a piece of crap the MDR-V600 is! They sound like a way cheaper sealed headphone! The MDR-V600's are almost bass-less (considering their size and price), have recessed mids and attenuated highs! The MDR-V900's are better, but still fall short - the bass isn't as nicely extended as even the Grado SR-60's!
Griffdog
New Head-Fier
- Joined
- Jun 23, 2001
- Posts
- 33
- Likes
- 0
Quote:
Give it a try now...
| Originally posted by markl BTW: you can't compare the RS1 to the RS2, you get an error. |
Give it a try now...
markl
Hangin' with the monkeys.
Member of the Trade: Lawton Audio
- Joined
- Jun 22, 2001
- Posts
- 9,130
- Likes
- 49
Look at the response of the one Stax phone they have up. Can that be an error? It does not look so good.
Eagle_Driver
Headphoneus Supremus
- Joined
- Jun 22, 2001
- Posts
- 6,496
- Likes
- 61
Tyll, I also saw some headphones having very extended bass response that have an obviously over-boosted mid-bass response (e.g.: Koss UR-20, Sennheiser HD-590). (By "over-boosted" I mean that the frequencies in question are at least 10dB higher than the reference level.)
As for Sony headphones, I saw the graphs of the other two models (MDR-CD2000 and MDR-CD3000 - they were mis-labeled as the MDR-V2000 and MDR-V3000), and - to my surprise - the MDR-CD2000 measured almost as well as the Sennheiser HD-580! (Though the MDR-CD2000 may not sound as well as the HD-580 to many listeners' ears.) And when I said that the MDR-V600 and MDR-V900 are lacking in deep bass response, I only had to take a look at the "Panasonic Crappy Headphones" (the ones that are bundled with the SL-CT580) to find out that even those crappy Panasonic 'phones have deeper bass extension than either of the two MDR-V### Sonys.
As for Sony headphones, I saw the graphs of the other two models (MDR-CD2000 and MDR-CD3000 - they were mis-labeled as the MDR-V2000 and MDR-V3000), and - to my surprise - the MDR-CD2000 measured almost as well as the Sennheiser HD-580! (Though the MDR-CD2000 may not sound as well as the HD-580 to many listeners' ears.) And when I said that the MDR-V600 and MDR-V900 are lacking in deep bass response, I only had to take a look at the "Panasonic Crappy Headphones" (the ones that are bundled with the SL-CT580) to find out that even those crappy Panasonic 'phones have deeper bass extension than either of the two MDR-V### Sonys.
markl
Hangin' with the monkeys.
Member of the Trade: Lawton Audio
- Joined
- Jun 22, 2001
- Posts
- 9,130
- Likes
- 49
There are in fact Sony MDRV200s and MDR V300s (both sub $50 phones), but no v2000 or V3000s.
As I have just ordered the Sony CD3000, I am dieing to know which Sony phones those really are.
Headroom-- can you clarify? Thanks.
As I have just ordered the Sony CD3000, I am dieing to know which Sony phones those really are.
Headroom-- can you clarify? Thanks.
Users who are viewing this thread
