Skip to main content

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: Has anybody ABX'd headphone/speaker burnin? (Read 16300 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

Has anybody ABX'd headphone/speaker burnin?

I did a search, and could find nothing.

I am wondering if there have been any ABX tests of headphone burnin?

I would think it would be necessary to get say, 10 headphones of the same brand/model.

Burn in 5 with some protocol (and there do seem to be a few of these), and then try to ABX between the 10 sets.


Has this been done?

h

Has anybody ABX'd headphone/speaker burnin?

Reply #1
Not to my knowledge, except maybe in the AES, that have performed many scientific studies that the public is not aware of.

Anyway, you should not burn-in 5 out of 10, you must completely randomize the choice for every pair to be burned or not. At the end, there can be any number of burned-in pairs, from zero to ten.

Has anybody ABX'd headphone/speaker burnin?

Reply #2
Anyway, you should not burn-in 5 out of 10, you must completely randomize the choice for every pair to be burned or not. At the end, there can be any number of burned-in pairs, from zero to ten.


I dont think you are correct. Most statistical tests require a balanced design.  Your proposal has a 1 in 512 chance of having either all 10 burnt in, or none burnt in, a 1 in 256 chance of having only one burnt in or only one not burnt it, a 1 in 128 chance of having only 2 burnt in or only 2 not burnt in, etc.


So 5 burnt in and 5 not is best, but I agree that which 5 are burnt in should be randomly selected.

h

Has anybody ABX'd headphone/speaker burnin?

Reply #3
You are free to built the test setup you want.
The only difference is that there are 1024 ways of randomly burning-in some pairs among 10, but only 252 ways of burning in exactly 5 of them, which reduces the maximum significance of a success.

In both cases, two reference pairs, one burned in, and the other not burned in should be provided for comparison.

ABX tests are widely used, and they are not "balanced".

Has anybody ABX'd headphone/speaker burnin?

Reply #4
Can I just make the observation that, since headphones/speakers are electro-mechanical transducers (and therefore subject to fairly large manufacturing tolerances compared to electronics), it is likely that the sample variation amongst nominally identical models will probably be greater than any change due to burn-in. Therefore I don't see how you can hope to get meaningful results from an experiment like this.

Has anybody ABX'd headphone/speaker burnin?

Reply #5
  • Buy an anechoic chamber and a high-resolution digitizer
  • Buy a speaker
  • Measure the impulse response of the speaker
  • Burn the speaker in
  • Measure it again
  • Use the impulse responses to convolve a single selection of music
  • ABX the two versions of the selection with your choice of distortion-free transducers*

* transducers not included. However, I don't think anybody is going to chew you out too much if you just use electrostatics.

Has anybody ABX'd headphone/speaker burnin?

Reply #6
Can I just make the observation that, since headphones/speakers are electro-mechanical transducers (and therefore subject to fairly large manufacturing tolerances compared to electronics), it is likely that the sample variation amongst nominally identical models will probably be greater than any change due to burn-in. Therefore I don't see how you can hope to get meaningful results from an experiment like this.


Which is why I think you would need at least 5 replicates of just one burnin method....






, which reduces the maximum significance of a success.


Not sure what you mean here. Significance has a special meaning in statistical tests, are you referring to that?

h

Has anybody ABX'd headphone/speaker burnin?

Reply #7
I don't think ABX is needed.  You could do the impulse response before and after burnin, as suggested above.  If it changes significantly, then very likely the result could be detected.

I've actually got a sound proof chamber hooked up to LabVIEW and some very nice NI DAQ hardware.  How long does it take to burn in a pair of headphones if I wanted to play around with this?

Edit:  Well, not sound proof, but pretty good, at least above 1KHz, and still fairly good down lower.


Has anybody ABX'd headphone/speaker burnin?

Reply #9
In fact, I also had similar questions, but I don't know how to describe the process. Now I know it is called burn-in...

I would like to ask, is "burn-in" really scientific, i.e., is there any proofs (except ABX) that shows headphones behave differently before and after the "burn-in" process?

And no, I am not referring to the head-fi FAQ, because I think their FAQ is based on a statement that, as they say, "a pair of headphones may not sound as good as a well used pair", but I am asking the validity and the reason of this statement itself.

Thank you.

Has anybody ABX'd headphone/speaker burnin?

Reply #10
Hmm.  Well, I threw together a test program to measure the frequency response of a pair of headphones.  I guess whenever I get around to replaceing my broken MX500s, I'll try it.

Has anybody ABX'd headphone/speaker burnin?

Reply #11
Can I just make the observation that, since headphones/speakers are electro-mechanical transducers (and therefore subject to fairly large manufacturing tolerances compared to electronics), it is likely that the sample variation amongst nominally identical models will probably be greater than any change due to burn-in. Therefore I don't see how you can hope to get meaningful results from an experiment like this.


Meaningful results may be got if the changes due to burn-in are bigger than sample variation. As it is the case for speakers : http://www.hydrogenaudio.org/forums/index....11&#entry209911


, which reduces the maximum significance of a success.

Not sure what you mean here. Significance has a special meaning in statistical tests, are you referring to that?


Yes. If someone gets all the right answers, the probability of the null hypothesis, that is the probability that he got them right by chance, is 1/252 if 5 of them exactly are burned-in, and 1/1024 if any of them has one chance out of two to be burned-in.

I would like to ask, is "burn-in" really scientific, i.e., is there any proofs (except ABX) that shows headphones behave differently before and after the "burn-in" process?


For headphones I have no idea. For speakers, the link above shows that after 70 hours of burn-in, performances are significantly affected, and measurable parameters get closer to the manufacturer specifications.
However, it has been objected once that after several days not in use, the speakers get back to their original state.

Has anybody ABX'd headphone/speaker burnin?

Reply #12
Here is my thought regarding speaker burn-in:

Speaker manufacturers aren't going to be testing burned-in speakers during the design process, they are going to be testing fresh-off-the-line prototypes during the design process. They will/should probably run an endurance test to make sure they don't degrade noticeably over time, but that is it.

If there is a speaker engineer on here who DOES test burned-in speakers rather than newly minted ones, please correct me, but I doubt that is the case. This means that speakers, if anything, probably get a bit WORSE over time, not better.


Has anybody ABX'd headphone/speaker burnin?

Reply #14
Understand that I don't know what any of these listed specs actually mean, so the following is just speculation (as usual):

Is it not possible that the maufacturer does a burn-in, measures certain parameters, and then lists those? Depending of course on what these specs mean, the speaker may have just degraded to the expected values. This isn't the same thing as "improving" to the expected values. If i'm way off base, could you give me a quick primer on what VAS, Qts etc. mean?

Has anybody ABX'd headphone/speaker burnin?

Reply #15
I don't remember what they mean, but I remember having posted a link to a glossary in the discussion about burning-in linked above.

It is very possible that the manufacturer does a burn-in before studying the speakers. The measured specifications get closer to the expected values after burning in, however, they are still far from them.

Has anybody ABX'd headphone/speaker burnin?

Reply #16
You are free to built the test setup you want.
The only difference is that there are 1024 ways of randomly burning-in some pairs among 10, but only 252 ways of burning in exactly 5 of them, which reduces the maximum significance of a success.

In both cases, two reference pairs, one burned in, and the other not burned in should be provided for comparison.

ABX tests are widely used, and they are not "balanced".


What if one the reference pairs has some very minor manufacturing defect, or just unusually big variation in some component property that alters the sound slightly? For this reason, I think that in this type of test there's no sense in providing reference pairs, instead just the 10 samples. And the number of burnt-in ones should be same as new ones, for the same reason.

Anyway, lengthy listening tests with many subjects will likely "burn in" the headphones as well as pink noise or whatever. But maybe the non-burned-in ones could be changed to new ones as the test progresses...

I remember reading somewhere (can't remember the source, sorry) that burn-in has more significance with loudspeakers. This makes sense, as burn-in probably affects mainly the mechanical structures of the element: headphone elements are light and supple, opposed to heavy rubbery suspender elements in loudspeaker elements.

Has anybody ABX'd headphone/speaker burnin?

Reply #17
Speaker manufacturers aren't going to be testing burned-in speakers during the design process, they are going to be testing fresh-off-the-line prototypes during the design process.


Of course they are!! Jesus...

They are going to burn them in, then they are going to burn them some more. Then they will measure the so called Theile Small parameters (which are used to model the speaker for calculations). Then they will design the speaker box, crossover etc. Then they will build it. Then they will listen and modify the design, because calculations are one thing and real-life is another. Onyl after several of these cycles the design process will be finished.

I won't get into headphones, but speakers need to be burned-in prior to critical listening and to achieve their targeted sonic performance. There are not many (if at all) manufacturers that burn-in speakers themselves. Rather, they are just assembled in the factory and left to user to perform the burn-in. There are mechanical changes, particularly in the speaker's suspension and dumping, that will significantly change the sound of speaker when it's new from the factory and after burn-in.

Headphone drivers have different designs and suspension so the logic does not necessarily translate to them, but burn-in certainly won't hurt to "workout" the drivers initially and stabilize their characteristics.


Regards,
Goran Tomas

Has anybody ABX'd headphone/speaker burnin?

Reply #18
There are mechanical changes, particularly in the speaker's suspension and dumping, that will significantly change the sound of speaker when it's new from the factory and after burn-in.


Back on topic, are there ABX tests that backup this statement ?

Has anybody ABX'd headphone/speaker burnin?

Reply #19
Back on topic, are there ABX tests that backup this statement ?


I don't know... It's pretty obvious, so I don't think anybody made such test. Why don't you make one?

Btw, just by measuring the change of Thiele Small parameters before and after the speaker drivers are burned-in, you can see (even if you don't trust your ears) that there will a change in sound.


Regards,
Goran Tomas

Has anybody ABX'd headphone/speaker burnin?

Reply #20
It's pretty obvious, so I don't think anybody made such test. Why don't you make one?


Because I don't have access to brand new speakers. However, "obvious" is a word that doesn't mean much in high-fidelity. That's why in this forum, we always analyze audio with controlled tests. The term of service number 8 of the boards says it all :

Terms of Service point 8:
Quote
All members that put forth a statement concerning subjective sound quality, must -- to the best of their ability -- provide objective support for their claims. Acceptable means of support are double blind listening tests (ABX or ABC/HR) demonstrating that the member can discern a difference perceptually, together with a test sample to allow others to reproduce their findings. Graphs, non-blind listening tests, waveform difference comparisons, and so on, are not acceptable means of providing support.

Hydrogenaudio is supposed to be an objectively minded community that relies on double-blind testing and relevant methods of comparison in discussion about sound quality. The usual "audiophile" speak of non-audio related terms which are completely subjective and open to redefinition on a whim, are useless for any sort of progression in discussion.

This rule is the very core of Hydrogenaudio, so it is very important that you follow it.

Here is a discussion explaining why
http://www.hydrogenaudio.org/forums/index....showtopic=11442

You can read how to easily perform double blind listening tests here :
http://www.hydrogenaudio.org/forums/index....howtopic=16295




Btw, just by measuring the change of Thiele Small parameters before and after the speaker drivers are burned-in, you can see (even if you don't trust your ears) that there will a change in sound.


Yes, but the question is wether this change is audible or not. It is also possible to measure a change between two interconnect cables. 0.01 dB at 20 kHz, for example. This is completely inaudible.

A simulation of speaker sound before and after break-in was made, but doesn't allow to reach a conclusion about its possible audibility : http://www.hydrogenaudio.org/forums/index....ndpost&p=210126

By the way, this simulation shows an improvement after break-in. Not a degradation.

Has anybody ABX'd headphone/speaker burnin?

Reply #21
Has anything changed about this in the past 7 years?
I only heard about speakers/headphones burn-in recently and am pretty skeptical about it actually doing anything. I don't remember the sound changing significantly over time after buying a pair of headphones.


Has anybody ABX'd headphone/speaker burnin?

Reply #23
I guess I should have made myself clearer. Has some evidence there is an audible difference between freshly bought and burned-in headphones/speakers shown up? Maybe some ABX results?