Quote:
Originally Posted by
Soaa- 
Several people with measurement equipment (namely Tyll from InnerFidelity) have shown that burn-in exists but only to a very fine degree. That's beside the question though. Assuming burn-in does exist, is the rate of change constant, increasing or decreasing over time? In the first two cases, one could project that a headphone could go through so much burn-in that it acquires undesirable characteristics. In the third case, the burn-in process would end at some point, and the drivers would reach a final state which no longer changes. In this case, the only way to actually damage the driver would be to subject it either to current that is beyond its designed limits, or to physically damage it.
If they did measurements in dB, then a small change in dB is extremely large. Starting from zero, every ten decibels you gain is 10^x the energy (the SI measurement for loudness). For example:
10 dB = 10^1 more energy
20dB = 10^2 = 100x more energy
30dB = 10^3 = 1000x more energy
40dB = 10^4 = 10000x more energy
<continue however far you want to go>
So lets say they measured a before and after and a certain frequency went from 80dB to 60dB... This is also 10^8 - 10^6 or 80dB - 60dB. So you'd have an energy change of about 99000000, which is a lot of energy. Now lets say that it went from 20dB to 0dB. This is also 10^2 - 10^0 = 100-1 = 99 energy change. They are both 20dB drops, but one is a heck lot more energy than the other.
I do want to see if their data shows that a driver can actually fail due to overuse (normal use), it would show up in their data for sure :)