How Science Got Sound Wrong

I think we're surfing on religion here!  At the end, the listener can prefer possibly a more distorted version of the original sound coming out of the instruments played by the musicians!  A more "colorful" sound liked by this user, which for another may be completely different.

I'm an electrical engineer with 27 years of experience... and i'm an audiophile too.  I've seen the arrival of digital sound.  The first CD Player i listened to sounded awful, thanks to brickwall analog filters that destroyed the sound.  Then technology refined and then came oversampling + noise shaping en refinement in electronic.  It took time to get DACs that could do the job properly.  It reminds me the first big Wadia DAC.
As we speak, now the electronic in audio is something we master much better and it translates into high quality for decent price.  Engineering is like that: You can do many things but at the end, you're selling a product for profit.  You have many constraints: Price, Technology, Production, Quality, Components Tolerance, etc.  And you target an audience, a customer.  There is of course competition: How to sell your product?  What can you offer that will translate into sale?

Over all those years, i listened to many many very high end audio kits assembled for shows, in listening rooms, etc.  The system you like might not be the most expensive.  And my friend, which is an audiophile too, doesn't necessarily like the same system.  In our case though, it's clear that we can't go back to vinyl for many reason.  We liked back then the great Linn-Sondek LP12 playing through Mark-Levinson (No 23.5 + No 26) electronic and the Kef R107 speakers.  But what we have now, because of convenience and excellent audio without surface noise / clicks, superior separation and dynamics, is very pleasing to listen (for us!):  Ayre QB9, Krell (KSA-100s + KRC) + Dynaudio Confidence 20.  We just use a little Atom Mini-ITX system running Debian + MPD + front end, hooked to a file server containing FLAC file, for us it is the perfect system.  Other may like to handling / playing vinyls over the great looking and sounding turntables (VPI HR-X, wow) !!

Last year, when we went to our nearby HiFi shop, we listened to an expensive system (150K $ +) that featured the big Mac amplifiers / preamp and i don't remember the rest: Didn't like it that much... But it may surely please someone else!  And you have the "placebo" effect too: Your brain will "contribute" to confirm what you believe!  That's really not a problem:  Listen to music and have fun (which is the ultimate goal) !

So if you consider sound coming from instruments and finally reaching your ears, you have a "transmission line" that modify the sound as it travels: Microphones, ADCs, mixing, transfer to medium, playback by your record player, preamp ( DAC / RIAA ) , amp, wires, speaker, listening room and finally your ears / brain / taste!!!  So here we are with this topic.

You know what, it doesn't bother me at all the someone prefers vinyl!  I prefer digital but i agree that listening low quality streaming music played back with a mediocre sound system sounds awful!  I could say the same with low quality vinyls, badly recorded played back on a bad system!
Saying that science failed with music is just a bit exaggerated...  Saying that engineers "don't get it" is a bit of a lack of respect IMHO...  Sure, bad engineering exists...as bad musician and "believers" listeners...  But as an engineer, your job is to design things with constraints and come up with a solution: Sometimes the best solution is the one that translate into more profit, best looking, most convenient, most portable, etc.

Not only did I take the time to read what you wrote, I researched your uncited quotes to find they were just musings on a forum by someone with fundamental holes in their knowledge, not an actual expert, and not remotely reviewed piece of work. The papers I posted do not all support that forum posters hypothesis or yours.


Shannon Nyquist theorem is highly understood and not doubted. The links I posted are highly relevant. I specifically picked a paper with wickedly bad SNR to illustrate that for audio where SNR is high, that it is pretty much non issue wrt providing well beyond subsample resolution. That you think they are not relevant either shows you don’t understand the underlying concepts or your argument is a red herring. Which is it?


Your Fermi quote is a red herring. When Fermi says contrary to hypothesis, he means he reran it 100 times, reviewed all his equipment, talked to trusted colleagues, and then published a paper fully detailed so others could replicate the experiment. He did not mean a pop science paper by someone who does not understand the science.


I provided actual scientific papers. You quoted a post on a forum. That I think says it all.

blueranger
So are we hearing distortions in the electronics? Digital filters not removing all distortions and disrupting electronics downstream. There is something going on that people hear.

>>>>What you’re hearing is jitter caused by scattered laser light and by vibration from various sources. It’s a failure of the Reed Solomon Error Detection/Correction Codes as well as the CD laser servo feedback mechanism to keep up with external vibration, vibration induced by the transport and/or transformer and fluttering and vibration of the CD itself. The CD flutters because it’s out of round and/or because the disc is frequently not absolutely level whilst spinning 

"....and not stuff about rice prices in China...."

It is a science of a magnitude many cannot imagine.

You quoted a post on a forum

Don’t mean to point out the obvious but for all intents and purposes you are just a series of posts on a forum.Though he has the advantage of saying things supported by the articles you posted.

Truth be known we once sold a noise canceling system for all manner of piping systems. The math and science behind it was amazing and screamed all knowing and complete, The problem was the digital controller which was supposed to do what you claim digital audio can do, would , when confronted with high frequency anomalies ( you know where the timing thingee plays ) lose the plot and create horrific noise. Later versions worked but they concentrated on just the low frequencies ( not where the timing thingee is ) which are more regular and easy to map. Broadband attenuation is still pretty difficult.