1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.

    Dismiss Notice

Can the burn-in skeptics leave us alone?

Discussion in 'Sound Science' started by kingpage, Nov 12, 2011.
First
 
Back
1 2 3 4 5 6 7 8 9 10
12 13
Next
 
Last
  1. Blue Boat


    Quote:

    Recently read a review about an amplifier. Reviewer said that on his first listen, it sounded dark and muddy. Apparently the sound opened up after a couple of hours of burn-in. I think it was an Ibasso D-Zero review. Placebo effect or bad components? Hmm. 
     
     
  2. liamstrain


    Quote:


    Or psycho-acoustical acclimation, or any number of testable functions - none of which are "burn-in"
     
  3. LizardKing1
     
    Quote:


    Thanks a lot, now people are going to read this and all of a sudden people will start hearing "night and day" differences in their amps from when they bought them. Get ready, the subjecti-fi posts are coming.
     
  4. Argo Duck
    Interesting info Uncle Erik.
     
    Maybe people do hear this happening - hence upgraditis! 2000 hours might be - what? - anywhere from 1 to 3 years of dedicated (every day; 5.5 to 2 hours) use.
     
    I'm only half serious on this point - there are several other obvious reasons why people change gear. The key point is that as new explanations or overlooked aspects - such as this below - are brought up, the 'data' has to be reinterpreted.
     
    More particularly, the fact "no one" thinks to attribute something they do/see/hear to explanation X doesn't mean X is not the explanation. This is a frequent scenario in science.
     
    Quote:


     
     
  5. Head Injury
    Quote:

    Won't change much, people already suggest burn-in for amps, tubes, opamps, capacitors, and wires.
     
    Slow and gradual changes aren't going to be noticed. We adjust to changes in sound too quickly, and we forget the details of what we heard too soon. If burn-in on a noticeable magnitude exists, it has to be a rapid change early in the product's life. That's why people rarely recommend 1,000 hours of burn-in, only 100. Even 100 hours is silly, aural memory just isn't that good.
     
    Tyll heard a difference between two AKG K701s. For the sake of argument, ignore product variation and assume the differences were all caused by burn-in. Would he have heard the difference if he listened to one headphone, then listened to the other four days later? That's what burn-in believers suggest when they wax poetic about the changes they heard 100 hours later.
     
  6. LizardKing1
    I think it's more on the basis of "only after X time does the sound stop changing". I'm not saying the sound is better that way, I'm not even saying anyone will notice a difference. But it makes sense to let it burn-in if there's a chance that the sound will indeed change noticeably. That's why people let them playing through the night instead of listening during the entire burn-in process. I prefer the later one at stages, to see if I can pick up any differences.
     
  7. Head Injury
    Every hour left burning in off the head is an hour for you to forget what you've already heard.
     
    If we assume that burn-in is relatively constant, then you either:
    1. Spend enough time burning in that differences are significant, but you've forgotten what it sounded like
    2. Check frequently enough that aural memory is still intact, but differences are too small to be audible
     
    As small as the differences Tyll measured are for how long they took, you're not likely to get both significant changes and aural memory at the same time.
     
    The alternative is if burn-in happens in stages, sudden changes in sound quality. I don't know if physics agrees.
     
  8. LizardKing1
    No, sudden changes wouldn't make sense, since that would mean chunks of the driver would be coming off... I think.
     
    Ok, so maybe the best way to make a subjective appreciation would be to hear 20 minutes of very familiar music of the same kind. Maybe something that is revealing in the same aspect (like very bass-dependent, since that's where the sound is said to change the most). Then leave them on for a whole night and listen again the next day. That way no new sound memories are created and there's a decent chance you still remember what it sounded like.
     
  9. liamstrain
    1. I'm pretty sure even less than one hour away would be enough to be unable to say with any certainty if a certain sound signatures was subtly different than one you remember. Though, I admit I do not know of any studies on this. 
     
    2. Did you know that when you dream about seeing or hearing or doing something, your brain acts as though you actually see or hear or do that thing (same sections of the brain, and memories, are created, accessed and activated). Pretty crazy thing our brain, but it's going to wreak havoc on your proposal just by the way it works. 
     
  10. Head Injury
    I still think you're overestimating aural memory. One important aspect of a blind test is how quickly the stimuli are switched. And this is on the order of minutes at most between switches, but even that might be too slow. I don't think it's possible to make an accurate judgment of differences of sound after hours of burn-in, even ignoring bias.
     
    I wonder if echoic memory is relevant here? It only lasts a few seconds.
     
  11. Argo Duck
    I don't think echoic memory is relevant, at least not to support a thesis like 'aural memory is poor'. Echoic memory is a "holding tank" for sensory data with high capacity and high accuracy - for a few seconds - which our brains scan looking for meaningful information. For example, a speech stream in a language we don't know is likely to remain 'meaningless' sound. Given the high rate of decay, we encode nothing into long-term memory (LTM), and cannot give an accurate reproduction of what we just heard.
     
    However, a speech stream in a language in which we are fluent is a different matter. We rapidly chunk the stream and encode what was said into LTM. We are then able to reproduce this more-or-less accurately for some considerable time afterward - dependent on factors such as motivation to listen and freedom from cognitive distractors.
     
    Of course it is the result of the chunking - the 'meaning' - we reproduce, not the exact sensory input itself (unless we are something like a skilled mimic - potential interesting track but let's leave this aside for the moment).
     
    The question now is does this ability to encode with some degree of accuracy apply in the context of audiophile listening? Possibly it does. We have a definite vocabulary to assist encoding, to wit terms such as "bright", "warm" etc etc. Unfortunately, audiophiles who have not trained together - by which I mean have not participated in shared listening and mutual correction - may have different experiences associated with these terms. This would decrease the precision and meaning of our encoding - and ability to communicate meaningfully with each other - but wouldn't negate the fact of whether we do this or not.
     
    The next question is whether this hypothesized encoding is of any value to detecting subtle (I assume it is subtle) change, such as in burn-in?
     
    What follows is speculative, but anyway here goes. One approach is to hypothesize that this depends on (a) how finely we are able to differentiate different aural sense data and (b) our then being motivated to meaningfully associate other memory information (e.g. semantically - with words) with these differentiated chunks. Like any form of learning, the motivation to do this depends most likely on there being a community of others to make these associations meaningful in the first place, and with whom we can exchange mutual feedback that "you're right/wrong".
     
    Let us suppose we are motivated in this way, i.e. to satisfy the demand of some salient verbal community e.g. our professional/occupational community. We don't use 500 words if the vocabulary of our profession only demands 50, even if we could do it more finely. In this example, someone who develops 10 x more (finer) shades of meaning than the norm - 500 instead of 50 - would almost certainly not be reinforced to continue this ("you're talking gibberish"), and the behavior extinguishes.
     
    Looking at the vocabulary we use on head-fi - it's not large - and the fact we're mostly using different equipment, not listening under controlled conditions, and don't meet together often or at all - I'm just not sure 'we' (most of us) have a sufficiently precise instrument here to settle questions of burn-in or whatever. Either this small vocabulary means we can't differentiate the aural data well, or it's the product of this suboptimal functioning as a community.
     
    But there may be more to it than this. Memory has (or had, when I did my post-grad stuff 11 years ago) a lot of unexplored territory. In particular, not all 'meaning' is semantic. And unfortunately, most studies still seem to be at the point of permitting only limited generalizability outside their specified context, if the Glass, Sachse and von Suchodoletz (2008) study in the echoic-memory wiki link is anything to do by. I couldn't find anything much in it relevant to normal adults, let alone audiophiles.
     
  12. Clarkmc2
     
    Quote:
    From post #1. It is likely, if not certain, that large voice coil dynamic speakers would break in at least as slowly as small headphone transducers. Years ago this was studied at length by JBL. (I think it was, link long lost by me. There is a White Paper, but I can't find it anymore. Do this hobby for fifty years and you will have trouble finding things too.[​IMG])
     
    The conclusion was that there could be changes, but it would be over in a fraction of a second. So yes, burn in is possible in the motor of a headphone (as opposed to the earpads, arc, etc.) but it is over in a flash. If they sound different over time, it is not the transducer.
     
    For the sake of argument, let us speculate that it would take a thousand times longer in a headphone than in a large speaker. That would be a few minutes at most.
     
    I should add that if the magnetic structure of the motor is ferrite, there will be no change in that over any time period with any amount of current. (Greg Timbers, chief designer and engineer of JBL) Alnico can change with drive (current induced magnetic fields), but in headphones enough current to do that would burn out the voice coil. This effect is cumulative only in that the degree of change is determined by the largest magnetic flux encountered. It is not additive.
     
    Here is some info from Mr. Timbers on the effects of demag and heat: http://www.audioheritage.org/vbulletin/showthread.php?10280-Recharging-Alnico-Ferrite-magnets&p=110522&viewfull=1#post110522
     
  13. yjt93
    So is burn in not detectable on FR graphs? if it is why can't we just get 2 of the same headphones one burned in and one not and see the graphs?
     
  14. ianmedium
    Not sure if anyone has seen this link. Russ has been around for years and is one of the most respected individuals in the British HiFi scene. He has had a special relationship with Kimber for years.

    This is what he says about burn in
    http://www.russandrews.com/product.asp?lookup=1&region=UK&currency=GBP&pf_id=BURNIN&customer_id=PAA1877127111008TYPYPHNGZKCTZYYR

    must admit when I lived in the UK I have been a happy customer of his and have always found his advice, not only on cables but supports and mains conditioning to have made beneficial improvements to the sound of my then set up.

    If it is good enough for someone like Russ it's good enough for me!
     
  15. Uncle Erik Contributor
    That's sales literature.
     
    Nothing else.
     
    Of course it'll "work" if there's a dollar - or pound - to be made.
     
First
 
Back
1 2 3 4 5 6 7 8 9 10
12 13
Next
 
Last

Share This Page