[MacedonianHero]

Very interesting article by Tyll:
 
http://www.innerfidelity.com/content/testing-audibility-break-effects
 
Thanks Tyll for some great work here.
 
My experiences (although they were not double blind tests) do also match Tyll's with regards to my Ed. 8s. My newly replaced brand new drivers did sound VERY different from my older ones that had several hundred hours on them with regards to the smoothness of
the treble (or lack thereof).
 
Feel free to discuss...
 

 
 

 

[Willakan]

That IS interesting. Break in is pretty reasonable as audiophile concepts go - next logical step is to see whether one could distinguish two "burned-in" pairs to rule out, at least partially, manufacturing variations.

 

[anetode]

Nice article, though there's one conclusion that troubles me:
Quote:

That doesn't rule out that the two headphones may have just sounded different in that way from the manufacturer, but I doubt it.

 
Tyll may have good reason to doubt it, but he did not take steps to rule it out. Individual headphone drivers on high-end models are usually matched well within 1 db of each other, but the difference between pairs of transducers may be more significant. Testing should have been done on the two pairs before any burning in to get their baseline results and see whether there were initial differences that could have been recognized through DBT.
 
I think he made a very good case for burn-in on the 701s, but that last 10% of indetermination is still out there

 

[upstateguy]

 
Quote:

Very interesting article by Tyll:
 
http://www.innerfidelity.com/content/testing-audibility-break-effects
 
Thanks Tyll for some great work here.
 
My experiences (although they were not double blind tests) do also match Tyll's with regards to my Ed. 8s. My newly replaced brand new drivers did sound VERY different from my older ones that had several hundred hours on them with regards to the smoothness of
the treble (or lack thereof).
 
Feel free to discuss...
 

 
 

I want to echo an echo to Tyll for doing the work. 
 
I never doubted that mechanical things like headphones and speakers break in.  Recently, my T-1s changed rather dramatically within the first week or so.
 
That said,  I'd like to see if the measurements of the two 701s reflect what was heard.
 
I also want to mention that we should be very careful about applying the mechanical break in of headphones to solid state equipment.  It's fairly common knowledge that Tube equipment deteriorates over time but I've never seen any documentation for any SS equipment.
 
 
 
 

 

[gregorio]

I'm with anetote.

Tyll has determined a small difference between two headphones of the same model from the same manufacturer. And then, out of nowhere he makes a leap to 90% certainty that it is due to burn-in. He could just as easily have made a leap of 90% certainty that the manufacturing tolerances of the headphones resulted in a perceivable difference. If anything, this would have been the more logical conclusion.

The biased conclusions and inadequate testing made this whole exercise pointless in my opinion. What he should have done is to have measured and blind tested both cans new and then given one set of cans 1000hrs of pink noise and then measured and blind tested them again. Double blind testing would have been even more appropriate.

I could be cynical and say that Tyll has stated a few facts (such as not being able to rule out manufacturer differences) to give the impression of being objective, while deliberately manufacturing a false conclusion to convince others to share his obvious bias. But I don't know the man or his motivations, so I'll just leave it that his test was poorly conceived and provided no insight as to whether burn-in is a real phenomenon.

G

 

[Currawong]

Very likely he will try again after he has put 1000 hours on the new pair.  
 
An interesting observation is how he improved his blind test scores with practice and after he relaxed more.

 

[tranhieu]

What about pads softness? clamp force? bad solder joints inside one of the pairs? etc. all of those factors can contribute to the difference he heard, if any.
Moreover, 19 trials is a little too few for a hypothesis test. If I remember correctly, 40 is the minimum you should go for (a normal hypothesis test often includes around 200 or even thousands of samples though).
I appreciate his efforts, but as Gregorio said above, this test fails to conclude anything, I suggest (if it's possible) that the test will be redone with more trials, and more factors being taken into account.

 

[Tyll Hertsens]

Quote:

And then, out of nowhere he makes a leap to 90% certainty that it is due to burn-in. ... The biased conclusions and inadequate testing made this whole exercise pointless in my opinion.

 
Looks like your the one jumping to conclusions. 

 

[gregorio]

Looks like your the one jumping to conclusions. 

You are the one who stated you were 90% certain of burn-in without eliminating other possibilities. It is completely reasonable to conclude that your exercise was pointless if you ended up not "Testing the Audibility of Break-in Effects" when that was the entire point of the tests and the article.

G

 

[poikkeus]

The point of Tyall's study wasn't to prove the verity of headphone break-in; it was to provide evidence, based on multiple trials, whether Tyall could determine the difference between new and broken-in headphones. As Tyall states:
 
 
Quote:

Have we absolutely proven that break-in is an audible phenomenon? No. All I've proven is that I could tell one headphone from another. 

 
The point of the article wasn't to "prove" the audibility of break-in effects; it was to "test" the effects. It's an intriguing study, and I'd love to read more.

 

[gregorio]

The point of Tyall's study wasn't to prove the verity of headphone break-in; it was to provide evidence, based on multiple trials, whether Tyall could determine the difference between new and broken-in headphones. As Tyall states:
 
 
 
The point of the article wasn't to "prove" the audibility of break-in effects; it was to "test" the effects. It's an intriguing study, and I'd love to read more.

The Article was called "Testing the Audibility of Break-in Effects". If all he proved was that he could tell two headphones apart then he failed to Test the Audibility of Break-in Effects. And then he states "Be that as it may, I'm about 90% sure that audible effects from break-in exists". Sorry, am I missing something?

G

 

[Tyll Hertsens]

Yup.  
 
Occam's rasor says: the simplest answer is the likely one. Pile on more and more evidence, and slowly but surely a picture emerges. 
 
Remember, I have <i>measured</i> changes as well. 
 
Your arguments against audibility of burn-in are the longshots. AKGs manufacturing practices are far more likely to produce near-identical cans than not. 
 
Sit back, relax, take the various evidence I provide with a grain of salt, but don't ignore the meat and potatos there.
 
BTW, feel free to <i>prove</i> me wrong, but simply throwing stones is silly.
 

 

[mikeaj]

I already responded in the other thread, but here goes...
 
Likewise I'm a little disappointed that there was no listening test to compare the two prior to any break in, since I don't think it's fair to dismiss this very important variable offhand like that, without controlling it carefully.  Slight differences in overall SPL between the two headphones may be enough to distinguish them, break in and FR aside.  For some reason I can't find the measured data though.  Could someone link it please?  Anyway, I think it would be a long stretch to dismiss this all as useless data, even if it wasn't as rigorously controlled as we would like.
 
By the way, here is some data on speaker drivers breaking in over time (no shifting pads here):
http://www.vikash.info/audio/audax/

 

[Rdr. Seraphim]

I like the simple approach Tyll. Anyone could try this on their own, and come to their own conclusions, but this is a great experiment that agrees with my some of my own experiences. Other's may balk at your results, or use semantics, or prejudice (you didn't do it my way), or whatever, but then they can try it themselves if they want to. 
 
Nice little experiment. 

 

[gregorio]

Occam's rasor says: the simplest answer is the likely one.

In that case, the simplest answer is that you did the test badly and sought to mislead everyone round to your belief. I'm hoping in this instance Occam's Razor is not applicable.

Your arguments against audibility of burn-in are the longshots.

As I haven't (up to this point) made any arguments against burn-in, I would say "longshot" is an understatement!!

AKGs manufacturing practices are far more likely to produce near-identical cans than not.

I have used various AKG products professionally for nearly 20 years and while generally they are of good quality, I certainly wouldn't rule out differences near the limits of audibility. To assume otherwise, in my opinion, is what invalidates your test. My real problem though, is that even though you admitted the possibility of other factors at play, you still stated you were 90% sure that burn-in was a real phenomenon. As your evidence is unreliable, the only thing you can be basing your conclusion on is your personal bias. The audiophile world is already full of enough BS and in my opinion really could do with a lot more hard fact and a lot less fiction.

I think I should make my position clear at this point, as I haven't previously. In 20 years as a professional and having used many, many different professional audio products and a few consumer audio products, I have never personally experienced the burn-in effect. For some components, tubes for example, there is good reason to believe in burn-in. For solid state components and wires etc., there is extremely good reason not to believe in burn-in. For cans and speakers, there little reason to believe in burn-in. Certainly when I hear people say "night & day" differences, then either they have faulty equipment or faulty ears. I believe that the vast majority of the burn-in effect with cans and speakers is actually burn-in effect of the owner's ears. However, I have not seen conclusive proof that burn-in effect can never happen (with some speakers and cans) and I am certainly open minded enough to accept that for some speakers and cans it is a possibility.

So Tyll, I am not arguing against burn-in effect, even though I believe in balance that it is unlikely. Some real evidence of it's existence would have been useful, but for me, your poor methodology and your jumping to conclusion based on assumption, has done nothing to help clarify the situation.

Maybe I'm being a little harsh on you Tyll, at this point in time I'm not sure of your motivation so I'm not deliberately trying to rip you a new one. Maybe you have good intentions but even so, in this instance, the old saying comes to mind: "the road to hell ....."

G