Burn-in. Real or not?
Dec 25, 2011 at 11:35 PM Post #166 of 228

 
Quote:
So it boils down to "we can measure it but we don't know if we can hear it"?  Well we can forget about ever coming to a conclusion since it's impossible to perform an ABX test of an unbroken in headphone vs a broken in headphone.  Although we could come close by, say, measuring the impulse response of the unbroken in headphone, the broken in headphone, then convolve the unbroken in impulse response over the broken in impulse response (uh, I don't know the correct terms to use here) to simulate the sound of the unbroken in phones on the broken in phones then see if people can ABX between them.  Any change in frequency response is probably the most important thing to simulate.
 


So in other-words both sides (the ones who think burn in exists, and the ones who don't) are both subjective and "not" objective". No one wins here. Just stick to your belief and enjoy your music. No one will be convincing me otherwise anyways.
 
 
Dec 25, 2011 at 11:47 PM Post #167 of 228
 
Quote:
So it boils down to "we can measure it but we don't know if we can hear it"?  Well we can forget about ever coming to a conclusion since it's impossible to perform an ABX test of an unbroken in headphone vs a broken in headphone.  Although we could come close by, say, measuring the impulse response of the unbroken in headphone, the broken in headphone, then convolve the unbroken in impulse response over the broken in impulse response (uh, I don't know the correct terms to use here) to simulate the sound of the unbroken in phones on the broken in phones then see if people can ABX between them.  Any change in frequency response is probably the most important thing to simulate.
 


Now we're cooking.
happy_face1.gif

 
I can think of literally a hundred things we would need (not all physical, some methodical) to test, but still can't say if it would be conclusive after all that work. Would love to try, but I don't get paid for things like this, rather use that money for some better headphones - although I refuse to pay $5K for the Stax 009 - that makes no sense.
 
I didn't spend quite that much making my $15K 5 foot 4-way speakers & I didn't get distributor pricing for the parts at the time either. It included the insane crossover network components as well.


Quote:
Thank you for saying this.  Just like you can't throw out anything without first looking at how accurate it can be (even old tools still have good accuracy).  As long as the results are within range of accuracy, it counts (and should be counted). 

With proper data and theory that supports the data, it won't be fuzzy anymore, no joking here.  We do have ample information to create some idea of how it works, and why it behaves the way it does.  Although it's not a headphone, it's still a dynamic driver, and if it behaves a certain way in the big world, it will behave the same way in the small world (assuming we don't get atomic/sub-atomic). 
 


This may be. Honestly, I haven't played with raw headphone drivers other than some really cheap ones for fun. I doubt the physical processes that occur in speaker drivers translates 1-1 to headphone size drivers. There may be similarities (even a lot) but I I don't think it's so easy to categorically say this is true.
 
That's the beauty of science though, IMO. You sometimes find an answer that makes no sense, but is amazing anyway. Like what happened when NASA 'asked' an AI system to design a better antenna. The thing was shaped like something broken you would find in the dumpster, but it worked better than anything NASA had at the time for reasons that were not immediately apparent to the engineers involved.
 
 
Dec 26, 2011 at 12:08 AM Post #168 of 228


Quote:
 

Now we're cooking.
happy_face1.gif

 
I can think of literally a hundred things we would need (not all physical, some methodical) to test, but still can't say if it would be conclusive after all that work. Would love to try, but I don't get paid for things like this, rather use that money for some better headphones - although I refuse to pay $5K for the Stax 009 - that makes no sense.
 
I didn't spend quite that much making my $15K 5 foot 4-way speakers & I didn't get distributor pricing for the parts at the time either. It included the insane crossover network components as well.

This may be. Honestly, I haven't played with raw headphone drivers other than some really cheap ones for fun. I doubt the physical processes that occur in speaker drivers translates 1-1 to headphone size drivers. There may be similarities (even a lot) but I I don't think it's so easy to categorically say this is true.
 
That's the beauty of science though, IMO. You sometimes find an answer that makes no sense, but is amazing anyway. Like what happened when NASA 'asked' an AI system to design a better antenna. The thing was shaped like something broken you would find in the dumpster, but it worked better than anything NASA had at the time for reasons that were not immediately apparent to the engineers involved.
 


I will agree with you that it will not behave 1:1, but should still behave similarly.  Again, we don't know until we try.  According to that evidence (as of now measurements on movement do prove some sort of break in; it just doesn't prove a that the change is audible), the drivers do break in.  Remember, break in/burn in doesn't have to be audible to happen, if there still is a 1-2 dB change in sound (which is barely audible if it isn't audible at all) will still constitute to some sort of change/burn in like in the case of the K701s where they had changes varying from 0.1-1.7 dB [low] to 0.3 - 2.2 dB [high] (where low = to the error taken out and high is the error added on; I was told that the error was .3 dB).  Although it isn't audible (or not too audible) the sound technically did change.
 
Please note, I'll agree that one instance doesn't prove burn in for everything, but so far with the data we have in this thread, it is 3 for 3 where the data (although may not entirely show it) leans towards burn in.  Then again, the next 10+ tests might show otherwise.  At the moment, it does look like it's leaning towards the burn in idea, but as always 3 isn't enough yet.  Hopefully we see more information in the future. 
 
OT: can I have a link to that antenna story you have, it sounds really interesting and I'm bored and want to look into that.  It's such a cool story :)  (Please note I'm not doubting the story, I just want to know more about what this thing looked like and get more info into it :p)
 
 
Dec 26, 2011 at 12:49 AM Post #169 of 228


Quote:
I didn't spend quite that much making my $15K 5 foot 4-way speakers & I didn't get distributor pricing for the parts at the time either. It included the insane crossover network components as well.



Sounds like a spendy version of the Linkwitz Orion's
 
Dec 26, 2011 at 2:40 AM Post #171 of 228
 
Quote:
I will agree with you that it will not behave 1:1, but should still behave similarly.  Again, we don't know until we try.  According to that evidence (as of now measurements on movement do prove some sort of break in; it just doesn't prove a that the change is audible), the drivers do break in.  Remember, break in/burn in doesn't have to be audible to happen, if there still is a 1-2 dB change in sound (which is barely audible if it isn't audible at all) will still constitute to some sort of change/burn in like in the case of the K701s where they had changes varying from 0.1-1.7 dB [low] to 0.3 - 2.2 dB [high] (where low = to the error taken out and high is the error added on; I was told that the error was .3 dB).  Although it isn't audible (or not too audible) the sound technically did change.
 
Please note, I'll agree that one instance doesn't prove burn in for everything, but so far with the data we have in this thread, it is 3 for 3 where the data (although may not entirely show it) leans towards burn in.  Then again, the next 10+ tests might show otherwise.  At the moment, it does look like it's leaning towards the burn in idea, but as always 3 isn't enough yet.  Hopefully we see more information in the future. 
 
OT: can I have a link to that antenna story you have, it sounds really interesting and I'm bored and want to look into that.  It's such a cool story :)  (Please note I'm not doubting the story, I just want to know more about what this thing looked like and get more info into it :p)
 


Okay on the first two parts. But... let's think further ahead. How about a time when headphones are only in museums... When sound is projected directly into someone's head for example. How about a whole recorded concert with an accurate representation of the actual sound stage where you can virtually pick your stand/seat? What's the next step in audio tech, or the next 10? That's the really interesting thing!
 
Onto NASA AI (Artificial Intelligence). I can't really write much about it because I wasn't directly involved. I know of it through my Alma Mater. The pictures available for public consumption really don't do the program justice, but here are some links of what is on the web. You wouldn't believe some of the stuff this program dreamed up.
 
" For example, our algorithms have discovered counter-intuitive antenna designs that out-perform traditionally designed systems."
http://sites.ieee.org/scv-cs/archives/ai-evolution-of-antenna-systems-strange-yet-high-performance-designs
 
http://www.nasa.gov/mission_pages/st-5/main/04-55AR.html
 
 
 
 
Dec 26, 2011 at 12:18 PM Post #172 of 228
LOL, IDK if we'll have sound beaming into our heads in 10 years, but I doubt IEMs will be obsolete...  Tech will improve though, that's for sure. 
 
@Antenna, that thing does look messed up :p  Well, if it works, it works; don't change it.  Like Mao said, "it doesn't matter if it's a black cat, or a white cat, as long as it catches and eats mice, it's a good cat."
 
Dec 26, 2011 at 2:33 PM Post #175 of 228


Quote:
I hope something like that replaces all headphones.  I absolutely, positively hate everything connected to a wire :p



Bluetooth...  LOL, JK (that is getting better though). 
 
As for when music gets beamed into your head, well there shouldn't be moving parts then...  In return, no burn in :p


Quote:
Absolutely real. Not having very good ears but I can feel huge difference b4 and after burn-in for even my phone's default IEM.


Thanks for the testomony. 
 
 
Dec 26, 2011 at 9:09 PM Post #176 of 228


Quote:
LOL, IDK if we'll have sound beaming into our heads in 10 years, but I doubt IEMs will be obsolete...  Tech will improve though, that's for sure. 
 
@Antenna, that thing does look messed up :p  Well, if it works, it works; don't change it.  Like Mao said, "it doesn't matter if it's a black cat, or a white cat, as long as it catches and eats mice, it's a good cat."



Off topic, but that quote is actually from Deng Xiaopeng, Mao's successor, who was commenting on people's criticism of his slight departure from Mao's communist policies towards some limited form of free market.  Black cat and white cat referred to communism vs caplitalism.
 
On the topic of testimonials... for the longest time I thought my favourite IEMs the Philips SHE3580 had undergone significant break-in change, what with the accidental high-volume blasts I fed it and the long months I'd used it.  Then I bought a second copy and plugged them into a Y-splitter, then put a bud from each copy into a different ear.  It didn't really feel any different from listening to the same (new or old) copy on both ears.
ph34r.gif

 
HiBy Stay updated on HiBy at their facebook, website or email (icons below). Stay updated on HiBy at their sponsor profile on Head-Fi.
 
https://www.facebook.com/hibycom https://store.hiby.com/ service@hiby.com
Dec 26, 2011 at 9:15 PM Post #177 of 228


Quote:
Off topic, but that quote is actually from Deng Xiaopeng, Mao's successor, who was commenting on people's criticism of his slight departure from Mao's communist policies towards some limited form of free market.  Black cat and white cat referred to communism vs caplitalism.
 
On the topic of testimonials... for the longest time I thought my favourite IEMs the Philips SHE3580 had undergone significant break-in change, what with the accidental high-volume blasts I fed it and the long months I'd used it.  Then I bought a second copy and plugged them into a Y-splitter, then put a bud from each copy into a different ear.  It didn't really feel any different from listening to the same (new or old) copy on both ears.
ph34r.gif


OK, confused about what happened here... 
 
Are you saying that
A) the old one and the new one sounded the same... 
 
Or
 
B) are you saying that having one in each ear (from old and new) sounded the same as new, then you did it again, and this time switched to the old and it sounded the same? 
 
There is an explanation for B, at the moment, none for A (assuming the new model hadn't gone though any burn in/was used).  Also, how many months precisely did you use it? As well as your age (range)...
 
Dec 26, 2011 at 10:16 PM Post #178 of 228
Don't know what the difference is.  As far as I can tell I did (B) (one old and one new in each ear sounds same as old in both ears) then inferred (A) (old and new sounds the same).  This I did when the new was brand spanking new.  The old was at least 6 months old.  Then again I bought another new one recently and listening to them one after the other I thought the new and old sounded different.  But I didn't do the one-in-each-ear test with this copy before giving them away to my brother.
 
I'm 30 y.o. and can hear up to about 15kHz.
 
HiBy Stay updated on HiBy at their facebook, website or email (icons below). Stay updated on HiBy at their sponsor profile on Head-Fi.
 
https://www.facebook.com/hibycom https://store.hiby.com/ service@hiby.com
Dec 26, 2011 at 10:59 PM Post #179 of 228


Quote:
Don't know what the difference is.  As far as I can tell I did (B) (one old and one new in each ear sounds same as old in both ears) then inferred (A) (old and new sounds the same).  This I did when the new was brand spanking new.  The old was at least 6 months old.  Then again I bought another new one recently and listening to them one after the other I thought the new and old sounded different.  But I didn't do the one-in-each-ear test with this copy before giving them away to my brother.
 
I'm 30 y.o. and can hear up to about 15kHz.



OK, the reason why B can be explained is that you create a mid-way point for the coming of a new sound (or old sound).  The actual change in sound is reduced greatly since half of the sound is new, half of the sound is old.  Essentially, you cut the actual change of sound over time by a constant (S = S/A where A > 1), which is a great amount to cut it.  With this cut, it's possible that the new change in sound (from new to old; or old to new) can now be below your sensitivity (this is using my hypothesis in the thread) thus allowing you to not hear the sound (because when you created that mid-way bridge, the actual change in sound was cut by some amount; EG S = S/A where A is some number > 1). 
 
In order to really get a proper AB, you have to literally listen to one pair for about few minutes (through at minimum one song), then switch for the same amount of time.  And do a couple (at minimum) trials to ensure accuracy. 
 
I asked for your age to see if your sensitivity would have decreased (it goes down with age; you're 30 years old, still young :)  your sensitivity hasn't changed much :p) . 
 
Hope this explanation makes sense, if you don't understand parts of it or need clarification, please ask (don't make assumptions on what I'm saying if you don't understand please).
 
EDIT:  I've created a small graphic to show you what I mean.  Although T1 and T2 in the graphs below are equally spaced, the space between them is the actual time spent switching between each IEM setup (NN -> NO -> OO; N = New, O = Old).  So in actuality they are not equally spaced.  The Blue dot on the first graph is the measurement of a smaller change compared to just going straight from OO to NN or NN to OO.  It may not be placed correctly.  Blue lines show average changes over time.
 
 ​
Look at the much great change in the change of sound over time with the direct swap while the average change in change of sound is much slower.  The actual change in change of sound from point one to 2 will never equal in graphs 1 and 2.  Hopefully this helps out a bit.  Remember my hypothesis, if the sensitivity to the change of change of sound < the actual change of sound of the system over time, then you won't hear it, if it's > it, then you will.  By lowing the actual change of change of sound of the system (you played time here), you couldn't hear the differences since the change became really small. 
 

Users who are viewing this thread

Back
Top