Senn mentions hand picked +/- 1 dB matching in the promotional material for the HD650 - presumably because that was at the time a new, better than previous practice, Good driver matching for a headphone
we also have Senn's HD800 certificate graphs - again showing lots more than 1 dB variation over frequency in serial production of a flagship audiophile headphone
so why would anyone expect "exceptional" channel matching in a DAC to have any responsibility for "...sense of space, dimension and ability to discern starting and stopping edges of sounds with such clarity and precision"
when "ordinary" Audio DAC do few % ~ few times 0.1 dB already? that's still 3-10x better than "bragging rights Good" audiophile headphone L,R driver match
and for competent Audio DAC that possible few % channel mismatch is a flat gain difference - not bobbing and weaving frequency and phase over audio
while bobbing and weaving frequency and phase over audio is what L/R headphone driver tolerance is, could reasonably be expected to affect "sense of space, dimension", no?
it is a commonplace for consumer grade audio DAC to “place” “audio events” like musical notes, percussion edges to within nanoseconds – even at 16/44
DAC noise floor below recording noise from microphones, studio electronics, venue/recording room noise is also common – almost guaranteed for any recorded music source that involved mag tape is again easily found at many price level
how does "ability to discern starting and stopping edges of sounds with such clarity and precision" get better than what was recorded, is delivered in commercial recordings?
Good Schiit may be nice to own - but why does so much bad Logic/Psycoacoustics/Engineering need to flung?
Why should anything better than some generally accepted (insert specific attribute and limits here) matter? Indeed, those are good questions. I've spent a lifetime as and EE and scientist asking myself those questions. Trying to make correlations between objective and subjective. Use ABX testing where ever possible. Wish I could locate a QSC ABX tester. No longer in production and those who own them won't give them up. I may have to design and build one when I finally have more time not committed to corporate research.
Now if one believes that all equipment that meets some given attribute limits sound the same, please save yourself time and frustration. Do not waste time reading any further.
I have long used a Benchmark in the studios and labs. Didn't pay too much attention to details of space, soundstage, dimension, it all sounded clean enough and everyone was predisposed to hearing recordings as a distinct reduction in realism. Hey, ODAC would be great as it nearly matched performance of the Benchmark at a substantial reduction in cost. I have both of those too.
Hearing great musicians with well tuned instruments in good acoustic spaces, on decent cans (back then HD580) through the live feed from a really good binaural setup such as a Jecklin disc with Schoeps mics or a Neuman KU 100 and then the output of the A/D/A one could readily hear the sound drop in realism. Not everyone has the ability to have this experience unless you are a studio rat, musician, acoustic researcher, pro audio designer/mfgr or have an in to a studio or mastering facillity somewhere.
So several years ago I became more interested in headphone listening for more than voice over and dialog editing. Recreational listening hadn't been something I would do with my free time given the listening fatigue of the audio / acoustic work all day. Maybe there were some improvements. So I acquired equipment to begin A/B listening and also equipment to make measurements for myself. Initially I would A/B by myself. But we all know how Expectation / Confirmation Bias can skew those results. Eventually I asked for help, with someone else making the choice for connections to the A/B switch and draping with a towel to obscure visibility of those connections. My initial measurement equipment was RMAA and best interface I could acquire without spending a fortune. Eventually that became an RME Fireface and then RME Fireface UC. Started acquiring other measurement SW such as Room EQ Wizard and ARTA which were also useful for room acoustics. Headphone measurements too with a custom built dummy head. So far, I could detect what I thought might be improvements. But I also noticed that some setups did not result in listening fatigue as quickly as others.
Ok, fast forward a bit. I now have a PrismSound dScope, a 16 bit ADC capable Picotech Oscilloscope, several Tektronix oscilloscopes, a laser displacement measurement system, a Klippel QC, Fluke DMMs, ACO Pacific microphones, Bruel & Kjaer mic calibrator, FLIR camera, Philips frequency counter, stratum 1 time reference, NIST traceable calibration for many of those items, and a whole lot more experience in both listening evaluations and measuring audio components than several years back.
So now I present measurements on several forums and make comments of what I think might be responsible for the auditory differences I believe I can now perceive. I try to make it clear that these are my own opinions, my own perceptions and YMMV. Though until I have an ABX tester to demonstrate in a public setting, say at an AES meeting (I am an AES member), then my comments will always be subject to doubts, even by myself. If it is SW only (comparison of VST plugins as an example) I have the Fraunhofer Pro Codec VST encoder which has live ABX comparison of audio streams. While intended as an aide to encoder parameter adjustments, it works well for many other purposes.
But one thing that has become clear to me. I can return to previous equipment and work for long hours and experience listening fatigue setting in not far into a session. Then move back to what I consider better components, with listening fatigue being delayed or possibly not occurring at all in the session (in the very latest system you see me talking about.) On average, I spend 5 to 10 hrs six days a week using my ears for audio / acoustic research both for work and recreation. To this end I have also had to learn to work at 78 dB SPL instead of the more typical 83 or 85, and heaven forbid, not at the godawful 95 to 110 levels to which some listen.
There have been many hours reading the works of Leo Beranek, Harvey Fletcher, Floyd Toole, F. Alton Everest, Havelock & Kuwano & Vorlander. Studying the basilar membrane and critical bands, the bark scale, and on and on. Moving beyond matlab simulations by performing experiments in the acoustic lab and hearing for myself.
So there is a peek into my background and how I have come to the place I currently reside in my thinking on what I perceive with my human auditory system using audio components. - atomicbob
