Brilliant pieces of writing, 853guy - very articulate - the most articulate writings I've seen of this aspect of our perception of music that is so difficult to describe. I have always enjoyed your posts & we seem to share a similar mindset, based on your past posts here.
Like you, one of my interests is in the research area of auditory perception as I attempt to find explanations for what I perceive in my own listening, experiments, research & what others report (as in your excellent posts)
So at the risk of maybe distracting from your beautiful words & disturbing the tone of this great thread let me try to articulate my current thinking on our auditory perception & how it may concur with what you heard in the Aries Cerat demo room. I have posted some of these thoughts before but your descriptions stoked in me a desire to articulate them again. Like you I reserve the right to be completely wrong & to change my thinking in the future - I consider these characteristics to be necessary to the concept of life-long learning.
My journey has brought me to this simple & maybe obvious preliminary conclusion - it seems that the low level audio stuff is the most critical for portraying what you have so eloquently described - the connection to the meaning & emotion of the performance & the performers.
Now underlying this simple statement there lies a number of aspects I believe are worth investigating/analysing:
- how low a level? I suspect that this level is below what is normally considered audibly significant. It seems to me that many dismiss such quiet sounds with the throwaway remark that "masking will make this inaudible". However, what we are talking about is the release portion of a sound (or alternatively, the quiet before the sound starts or the attack portion) - in other words when there is often no surrounding louder tones that are masking these low level aspects. Remember all sounds have attack, decay, sustain, release characteristics & in this way I believe it bridges the gap between sounds & music that you mentioned above.
- I agree with your differentiating between sounds & music & particularly with your emphasis on the fact that 99% (maybe more) of what we perceive is because of the processing engine happening in the brain i.e. we derive our moment to moment auditory perception by dint of moment to moment analysis of the signal stream. Many people seem to forget this. Furthermore, because we are using analysis to hear, we can/do also anticipate what is coming next & dynamically adjust (feedback to) our hearing mechanism so that (it seems) we can perform better than is predicted from some hearing tests i.e we can hear below the noise floor, we can use modulated tones to hear tones below noise floor (the so-called co-modulated masking release phenomena, CMR), we can apparently beat the uncertainty limit that applies to FFTs - the Gabor limit (by the simple fact of pattern identification & anticipation of a sound), etc. So I completely agree with you that the only real tests of the limits of hearing (should be called limits of auditory perception) is not to be found in tests with non-dynamic, non-patterned sounds
- is it the silence from which the notes emerge that is the important factor? Probably not because vinyl & RtR which have higher noise floors than digital & yet provide more realistic auditory imagery than most digital! So is it more about maintaining a steady noise floor? I suspect that this is partly the reason - steady state noise is far easier to analyse out than fluctuating noise floors. In digital audio, fluctuating noise floors in the presence of dynamic signal processing seems to be the norm. Even if this noise floor is measurably below vinyl & tape, the fact that it fluctuates catches the attention of our analysis engine & places an unnecessary processing load on it & often results in an unnatural change in noise. Note that the Kassandra DACis based on AD1865 ladder DACs & not sigma delta DACs which tend to suffer from this noise modulation characteristic.
- is it just that when this "sound floor" is attended to (such as in the Kassandra & others) that low level distortion is also reduced? Again the same considerations might apply - at what level are these distortions audible? Do the standard hearing tests tell us this? I doubt it as we know that we are more sensitive to higher odd order distortions. And again, remember that masking decreases with signal level, not increases which means that the ear can resolve more harmonic content from distortion at lower levels than at higher ones.
An interesting measurement of ultra low level signal distortion in amplifiers was done here & seems to subjectively correlate with blind test results but a bit early to say that it correlates to audibility with such a small sample size - see
here It's also interesting to follow the discussions on some audio forums about these measurements & how the typical measurement mindset gets it so wrong.
I'm always looking for better measurements that are focussed on how our auditory perception actually works & can come close to predicting how something will sound yet I get accused of desiring some pseudo-scientific measurements that will support my thinking. It's amazing how some twist a genuine search for truth into an agenda (a self revealing debate tactic)
- is it that we can more accurately/better perceive the temporal aspects within the music because of this lack of dynamic noise?
- is it all of the above?
- there is no doubt that when we encounter playback systems of this ilk, we are fully engaged by what we are hearing - the realism illusion is captivating & often people hear this as a jump in quality, an illusion that snaps into place as a realistic vista into the performance. What I find accompanies this is that the soundstage depth takes on realism & layering making it easier to follow the musical lines of each instrument. I'm pretty sure that this perceived affect is the result of how we process the signal into audio objects in the audio scene analysis that is the goal of all this auditory processing
- one other final (for the moment) factor I wanted to mention is that I find this "believability" of the music is mainly the result of the source rather than any other aspect of the playback system given amplifier, speakers & rooms that are reasonably well behaved.
Anyway, I hope I haven't brought too much speculative analysis into what is a very enjoyable thread?