Originally Posted by BlackbeardBen
analogsurviver, quite frankly this sounds like used car salesman spiel - implying that this burn-in "method" is proven without actually showing that this is.
I understand your goal (lowering the driver fundamental resonance), but I have a few issues with that:
1. Why? Like xnor mentioned, with a low output impedance amplifier the damping factor is large and frequency response anomalies from the resonance are negligible. Also, the mechanical behavior of the transducer (its resonance frequency) is directly related to its electrical properties (the impedance of the transducer).
2. Regarding the phase shift: It should not matter in the slightest, as long as the channels are equal to each other. We are generally not very sensitive to absolute phase (See p.13 here thanks to the white hat, Dr. Floyd Toole of Harman), and without crossovers or rooms where phase alignment is not perfect, dispersion throughout the room is critical, and cancellation varies by frequency, it should not matter at all. I call B.S. on your claim about hearing a loss of imaging below the natural resonance frequency of the driver based on the phase shift - again, because that phase shift is inaudible or nearly inaudible.
3. How? You claim that exercising the driver - and in particular your "method" of doing so - will lower the resonance of the driver. First, I don't think we're talking about planar magnetic drivers because they're fundamentally similar to electrostatic drivers. So speaking about dynamic drivers, you believe that the driver's compliant portion will change its modulus of elasticity and/or damping by "exercising it - and that your method is better than others. Your method, if I read correctly, is that by exercising it at lower frequencies than its resonance will cause it to want to resonate at a lower frequency, right? And you has come up with a highly detailed process for doing so, yet again without any evidence for its efficacy?
My experience with materials (I am a mechanical engineer and I have done some graduate level studies of material behavior and failure) is that in a molded polymer object like a diaphragm will have residual stresses from the cooling process. That is, at rest the driver that will be under stress internally at varying levels at different points - and it is plausible that there are some very small points within the driver that are near yield stress - that is, the stress where the material will plastically (permanently) deform. During the first cycle of a transducer, such points near yield stress will exceed the yield stress and deform - breaking in. QC running the drivers, or the very first cycle you play of high-excursion notes (low bass, high volumes) will be enough.
As far as further cycles, fatigue failure of polymer drivers appears to be a non-issue, so I don't think that fatigue failure is a critical consideration here - though I can't prove it of course. That's a whole can of worms on its own too, as fatigue is not well understood regarding polymers and it can have very strange behavior characteristics.
I think the experimental results speak for themselves - there's little change over time. I just don't think there's even a theoretical basis for break-in beyond the first running of the driver.
Thank you for the reply - all and everybody welcome.
The reason I purposedly made everything sound like a used car salesman spiel is a very simple one, and I thought hard whether to implement it or not. I took that risk simply because I want everybody to understand what is going on and due to the fact this is not JAES but head-fi I wanted everything to be possible to be carried out by ear. I can not assume an average head-fier will have or have access to signal generator, oscilloscope, microphone, ear coupler, etc. Part 2
is at kindergarten level simply because I wanted to use signals that can be safely listened to
without causing any permanent hearing loss. Grand finale is much, much tougher on the drivers
and should be listened to for the establishment of the very maximum level for each particular transducer for the minimum time possible to establish it.
Phase in audio is the biggest and nastiest can of worms imaginable. There are many sources of it, isolating them is tough because what is assumed and usually practiced if we look at the whole chain from the microphone to the speaker/headphone is so compromised with flaws immediately appearent as well as those more hidden that they far outstrip flaws in the headphone alone. So, hearing headphone flaws regarding phase is tough due to the extremely limited availability of recordings good enough to clearly demoinstrate it.There is fierce opposition from the makers of studio equipment to change the status quo - whom would they then sell their mixing desks to if "phase correct" recording chain would be instated as standard, de rigeur, by law ? Any microphone manufacturer would prefer seeing you recording multimiked than with a stereo pair ( or single 5.1 or 7.1 mike ) - for obvious reasons - money. These are the two culprits most responsible we have to endure phase mangled recordings in the first place. Very few recording artists are aware of this in the first place, fewer still have heard better recordings ( possible at least for 6 decades ), fewer still have enough courage to decline the multimiked session for the fear of losing reording contract/job. Get the picture ?
Thank you for the link to the JAES paper by Mr. Floyd. E. Toole - I find it informative and well written.
I find it amusing but most definitely not funny that "phase naysayer" comes from Sweden - home of Opus 3, Proprius, Prophone labels that are on the spear of quality in simple low mic count recording - please listen to anything on these labels, preferably on full range speakers ( Quad 63 and its bigger succesors, full range dynamic drivers such as Lowther ,) or better headphones ( Stax, Jecklin Float, AKG K 1000, good orthodynamics ) ; you should very quickly grasp what I am talking about. Than compare these recordings to more regular ones on the same equipment - this is where I am coming from . Frequent visits to live acoustic unamplified concerts will do no harm either...
Regarding audibility of the absolute phase : IT IS AUDIBLE. Period. It requires extremely extended bandwidth electronics in the low frequencies, preferably all the way down to DC or extremely close to it. All the "blues" regarding phase mentioned applies. And then some more, and even more. I will cite the recording that opened my ears (and later eyes, as I wanted proof on an oscilloscope I am not imaging things). It is recording of the trombone on Terje Rypdal's album Odyssey, ECM LP
track called Over Birkerot.
I do not care to check for this with oscilloscope on YT, but on original LP that trombone is recorded from so close range it actually produces "pulses" in one direction only - either positive or negative (depending on absolute phase reversal(s) in one's stage(s) of amplification. ) This assymetry gets greatly ameliorated if you use normal AC coupled amplification, as it can not sustain that near DC offset produced by the trombone ( it is actually valve and can only produce overpressure, can not produce underpressure if played by blowing in to the mouthpiece - few really good jazz musicians have learned how to "suck" on their brass instruments to give them additional form of musical expression with very distinct sound, yet more formal academic institutions try to supress it as much they can ).
I was really surprised that such extended bandwidth can be recorded and played back from vynil - do not take my word for it, please check the mentioned LP by yourself.
Absolute phase is the most audible with brass:
but once you heard and understood the difference is very pronounced on piano, soprano, etc. I stress again : recordings with minimalistic miking are prerequisite for proper judgement - multimiking will generally not do.
The easiest to hear absolute phase reversal is on good phones mentioned above, but full range speakers do not lag far behind them. Speakers with crossovers ( with very few notable exceptions ) fare incomparably worse at identifying absolute phase, as they can change their phase several times across the audible range and should not be used.
I welcome your input regarding your knowledge of materials. I did mention my fear that a burnt in headphone driver that started its life with fundamental resonance of say 50 Hz, had it lowered by burn in to say 27 Hz, may return up in frequency say to 40 Hz after longe(er) period of unactivity . I do not have the knowledge necessary to understand or explain why this can happen. I do not have (yet) all equipment required to produce graphs etc, but I do have signal generator, microphone(s) and an oscilloscope and this confirm fundamental resonant frequency can be lowered by burn in. And I am making recordings frequent enough to know the sound of the real thing.
My purpose is to further the knowledge in any possible field that might lead to improvement of the reproduced sound in the end - not to engage in endless academic debates. I am willing to listen and prepared to accept help from any well meant source. Amusing but not funny Part 2 : it is academic educated people that object any progress that according to their preconceived notions should have no effect on sound quality the most. Few of these are willing to really listen, fewer still are willing to admit after sound demonstration that "there must be something in what I claim". Been there, done that before.
You said it yourselves behaviour of polymers ( most frequent used as diaphragm for headphones, dynamic and electrostatic ) is not well understood. Say that I prove fundamental resonant frequency does get lowered by burn in - are you willing to investigate further in materials matters, so that we all gain greater knowledge and better performance from our equipment in the end ?