Announcement. AES Presentation on “What is Accuracy” by our very own member @j_j_ or James D. (jj) Johnston - Chief Scientist - Immersion Networks

[MattHooper]

And, yes, the industry itself has been in a rut since the (cough) astoundingly brilliant (cough) <that's sarcasm markers> of using the rather much worse 4 channel setup of L/R/RL/RR for "quad" (as opposed to L/C/R/Back> despite the evidence from Steinburg and Snow's 1933 work. Yes, really. The move from mono to stereo ignored complaints about distance perception that are well understood in the present day. The move to 4 channel maintaining the "phantom center" ignored that, and the understanding of spatial perception (which was rather primitive at that time due to adherence to the thoughts of the day) was both completely missing AND the mixes, which removed a very, very important cue from the end result, created the misbegotten idea that "there is only one listening position" which is demonstrably false given a proper mixing algorithm. (even for quad, but we need to dump that idea forever, please!).

JJ, would you mind elaborating on that paragraph?

I'm not totally sure I understood everything, but if I have it right you are decrying the use of phantom center vs an actual center channel. Is that correct?

I'm particularly interested in the sentence I bolded: what are the "distance perception" elements that were ignored in moving from mono to stereo?

 

[DDF]

To which I also say "Amen."

Its not preference.

 

[MattHooper]

Its not preference.

I was agreeing with the proposition that it is egregious to proclaim that euphonic distortions are "The Right Way" to do sound reproduction. In fact, I would disagree with pretty much any claim to a singular "right way" to do sound reproduction, since I think at bottom this changes with individual goals.

 

[Blumlein 88]

JJ, would you mind elaborating on that paragraph?

I'm not totally sure I understood everything, but if I have it right you are decrying the use of phantom center vs an actual center channel. Is that correct?

I'm particularly interested in the sentence I bolded: what are the "distance perception" elements that were ignored in moving from mono to stereo?

I think he is referring to the work behind the 2nd paper here in this collection of 1934 articles. They had done the work in 1933.

https://www.aes.org/aeshc/docs/bell.labs/auditoryperspective.pdf

Auditory perspective- Physical factors starting on page 6 of this pdf.

Short summary is they used tests where people listened to microphones linked to speakers with 2 channels or 3 channels (a center). It found good correspondence with reality using 3 channels, but distortion of spatial positioning with 2 channels or with 2 channels having a mono'd center speaker added.
They had started with many microphone across a soundstage fed to many speakers across a listening room. Obviously not terribly practical. They found most of that benefit was preserved with 3 channel recording/playback, but not with 2 channels.

 

[Blumlein 88]

Also interesting they investigated binaural sound then too. You might be interested in this paper on the famous demo from 1933 with 3 channel sound. It has a picture of the mannequin they used for binaural recording.

https://acousticstoday.org/wp-content/uploads/2018/06/The-New-Age-of-Sound-Gabrielle-E.-O%E2%80%99Brien.pdf

 

[j_j]

A lot of unnecessary argumentation and seemingly sham nonsense removed. Having someone apparently familiar with USENET has to know that it evolved from "netnews" and based on MOST of the content in all audio groups, "nutnews" is most certainly a reasonable description, and certainly not a rare one. I don't necessarily intend to refer to DDF's commentary, whatever that may have been, unless it was to the fallacy of impossibility.

It's not a claim, you did, for Heysers specific words I shared. We'll have to agree to disagree on that. I think its prescient because your talk linked here still frames the paradigms as accuracy vs preference, missing out on third option of the illusion of reality. That's something different. Its the core point I'm trying to get across.

I can't tell if this is a language issue, or if you just want to have an argument. There's no difference between an "illusion" of something and "what somebody perceives" in this particular context.

This is aside from preference, yes.

From your follow up, your position seems that adding perceptual clues is the best we can do to enhance the illusion of reality. I disagree. I say its not necessarily sufficient.

This makes very little sense to me. There no "enhance" here, the goal is to ENABLE the listener to use his or her own cognitive apparatus however they do it. It ***IS*** absolutely necessary to give them the necessary cues to do this.

My argument is that how we perceive reality has as much personal variance as preference does, and there should be room to accept personal tailoring for one's own condition towards achieving perceived "realism" (whether it be emotional state, memory recall of what 'live" sounds like etc) and that this pursuit & the methods used in it should not be demeaned as merely "preference".

That's all part of perception. That's how your brain works. That's how any human's brain works, and who said "mere preference"? I didn't. How many times did I say that "preference is personal" and "you can't critique preference", and then you DIMINISH PREFERENCE by using the term "mere". That's another straw man fallacy to say the least.

A comparison. I had a professor claim that religious experiences were real because it was real to the person having them. I asked his opinion about paranoid schizophrenics and then got an A and end of conversation. Its not OK to expect others to agree something is real because another perceives it to be. That's not my claim. But it is more than Ok in audio playback for your own personal needs to lead you to trash the signal any way you want if the result sounds more like the illusion of reality to you, personally. Currently, there is no room for this way of thinking, and that's a shame. It's wrong. And the "illusion of reality " is not preference and should not be downgraded to that term.

And? Either you're trying to play semantics or your PERCEPTION of what I said is extremely flawed. Did you even watch the entire talk? What happens inside your head is what you perceive. All anyone outside of it can provide is necessary cues (of any sort) to enable what you do. That's true for anyone, including your paranoid schizophrenic example. The perception may not match reality at all. That is obvious, from basic understanding and hard evidence.

I'm forced to consider very seriously that you just want to make an argument out of thin air here.

 

[j_j]

I think he is referring to the work behind the 2nd paper here in this collection of 1934 articles. They had done the work in 1933.

https://www.aes.org/aeshc/docs/bell.labs/auditoryperspective.pdf

Auditory perspective- Physical factors starting on page 6 of this pdf.

Short summary is they used tests where people listened to microphones linked to speakers with 2 channels or 3 channels (a center). It found good correspondence with reality using 3 channels, but distortion of spatial positioning with 2 channels or with 2 channels having a mono'd center speaker added.
They had started with many microphone across a soundstage fed to many speakers across a listening room. Obviously not terribly practical. They found most of that benefit was preserved with 3 channel recording/playback, but not with 2 channels.

A modern understanding is that the interference between two signals, l and r, with coherent signal, interact with the HRTF's of the listener to produce a midrange dip as well as some time domain confusion, which masks the time domain distance cues for moderate (not near field or far field) distance cues rather effectively, creating a collision of cues and confusion in the distance domain.

In Mono, the only cue you had was "distance cues" and some people who very likely listened to them liked the distance (or depth) cues. So some people undoubtedly disliked the cue they had previous learned to hear. With mono single speaker, the cues are not messed up unless there are pretty disastrous early reflections (in the under 1 ms range).

 

[j_j]

Oh, my opinion on Quad failing was very simple, it didn't provide enough gain in actual experience to be worth it, on top of various recording and playback issues. The "one seat only" was also a problem that limited the market.

The "one seat only" was due to bad production methods, but I'm not sure the ability to do that right existed at the time.

The problem now with immersive content is that the same production methods are still being used. If "immersive" content is going to be a thing, it needs to have value for the experience to matter.

 

[DDF]

Truly I'm not trying to be argumentative, I think this is a critical point but its being wholly missed. And its not just language.
The gap is too large for keyboard, perhaps it requires a bottle of scotch and 2 glasses.
Sorry for the detour.

 

[Sal1950]

The problem now with immersive content is that the same production methods are still being used.

Such as? There wasn't even Atmos coding back then, nor even 5.1 .

If "immersive" content is going to be a thing, it needs to have value for the experience to matter.

There's a large and rapidly growing market that find tremendous value in it.

 

[j_j]

Truly I'm not trying to be argumentative, I think this is a critical point but its being wholly missed. And its not just language.
The gap is too large for keyboard, perhaps it requires a bottle of scotch and 2 glasses.
Sorry for the detour.

I think the question is purely semantic. I'm too old for scotch except in tiny quantities, but a good CMS would do. But, frankly, my argument is that whatever gestalt you want to create must have the necessary cues present to create it. Preference can be involved, or not, but there must be sufficient and reasonable cues provided to point the 'right way' if nothing else.

 

[j_j]

Such as? There wasn't even Atmos coding back then, nor even 5.1 .

There's a large and rapidly growing market that find tremendous value in it.

ATMOS is primarily a loudspeaker format, I'm disturbingly well aware of it, and of how it creates some very unhappy results when used in the "most heavily coded" variety.

Atmos (the actual physical format) can be created in any one of a large number of ways. It's HOW it's created, and how well the coding can handle the resulting PCM streams, that is the issue there.

HOW it's created is disturbingly important. Do you want properly decorrelated "base" or whatever they call the reverberant content, or do you want something that kinda-sorta-mostly-somewhat unsuccessfully sounds that way?

What's more, it seems entirely inappropriate to me to limit playback to certain speaker layouts and certain specific angles. It's not a whole lot harder to just make it FIT your actual layout. Yes, this is all done and waiting, but it's not going to work in the hopelessly constrained Atmos coding format. This whole game of "using old stuff and calling it new stuff" needs to stop.

 

[Sal1950]

ATMOS is primarily a loudspeaker format, I'm disturbingly well aware of it, and of how it creates some very unhappy results when used in the "most heavily coded" variety.

Sadly the streamed codec uses a fairly lossy version of Atmos.
But it's just fine for Joe Sixpack or for previewing a release before a BluRay purchase of the lossless Dolby TrueHD mix.
But this situation is really no different than we suffered thru for a couple decades with streaming 2ch.
I'm fairly confident that if the multich enthusiast base continues to grow at its current rate, at some point in time
we'll be able to leverage the streamers into doing a lossless stream. But I know I don't have to tell you about the huge size
of lossless Atmos and the necessary bandwidth the streamer would need to appropriate. Time will tell.

HOW it's created is disturbingly important. Do you want properly decorrelated "base" or whatever they call the reverberant content, or do you want something that kinda-sorta-mostly-somewhat unsuccessfully sounds that way?

How well all music mastering has been created has always been disturbingly important.
Today we have more crap 2ch masters being created in pop than ever before, thinking loudness wars here but there's much more.
The good news is this is something you almost never see in the multich world, be it Atmos or 5.1.

I find the good surround masters being created by modern engineers such as Steven Wilson, Bruce Soord, Bob Clearmountain, Martin Giles, Alan Parsons, and so many more, extremely engaging and fun to listen to.
At the bottom line isn't that what's it's all about?

 

[Justdafactsmaam]

Sadly the streamed codec uses a fairly lossy version of Atmos.
But it's just fine for Joe Sixpack or for previewing a release before a BluRay purchase of the lossless Dolby TrueHD mix.
But this situation is really no different than we suffered thru for a couple decades with streaming 2ch.
I'm fairly confident that if the multich enthusiast base continues to grow at its current rate, at some point in time
we'll be able to leverage the streamers into doing a lossless stream. But I know I don't have to tell you about the huge size
of lossless Atmos and the necessary bandwidth the streamer would need to appropriate. Time will tell.


How well all music mastering has been created has always been disturbingly important.
Today we have more crap 2ch masters being created in pop than ever before, thinking loudness wars here but there's much more.
The good news is this is something you almost never see in the multich world, be it Atmos or 5.1.

I find the good surround masters being created by modern engineers such as Steven Wilson, Bruce Soord, Bob Clearmountain, Martin Giles, Alan Parsons, and so many more, extremely engaging and fun to listen to.
At the bottom line isn't that what's it's all about?

You mean it’s not about “high fidelity” accuracy to the original recording and the artists’ original intent?

It’s about being engaged and having fun?

Cool! Can I quote you on that?

 

[Sal1950]

You mean it’s not about “high fidelity” accuracy to the original recording and the artists’ original intent?

It’s about being engaged and having fun?

Cool! Can I quote you on that?

Maybe you don't understand multich and Atmos.
Dolby TrueHD encoded recordings are lossless 24/48k, the 5.1 base layers are 24/96k and both highly accurate digital recordings.
That's why they sound so good.

 

[Justdafactsmaam]

Maybe you don't understand multich and Atmos.
Dolby TrueHD encoded recordings are lossless 24/48k, the 5.1 base layers are 24/96k and both highly accurate digital recordings.
That's why they sound so good.

I understand it just fine. I just don’t limit my choice of recordings to such a small catalog.

But if you know where I can get original Dolby TrueHD encoded recordings of John Coltrane, Jimi Hendrix, Pink Floyd, Bob Dylan, Yes, Genesis, Ivan Moravec, Yuja Wang, etc etc etc that are truly “high fidelity” accurate to the original recordings as the artists intended them please please point me in their direction.

 

[Sal1950]

I understand it just fine. I just don’t limit my choice of recordings to such a small catalog.

But if you know where I can get original Dolby TrueHD encoded recordings of John Coltrane, Jimi Hendrix, Pink Floyd, Bob Dylan, Yes, Genesis, Ivan Moravec, Yuja Wang, etc etc etc that are truly “high fidelity” accurate to the original recordings as the artists intended them please please point me in their direction.

I got most of that on discrete quad or better, no problem.

 

[j_j]

I find the good surround masters being created by modern engineers such as Steven Wilson, Bruce Soord, Bob Clearmountain, Martin Giles, Alan Parsons, and so many more, extremely engaging and fun to listen to.
At the bottom line isn't that what's it's all about?

I have to say there's a few more using good methods that are not generally (yet) available, but one ***MUST*** stay away from the "encoded versions". There is a host of disasters lurking there. It would be (yes, we have one) a codec that can do a lot better job, of course with more bits, but again OF COURSE you want PCM if you can get it.

There are things I could say but shouldn't, starting with how panning is done in any kind of system, 2 channel upwards.

 

[j_j]

Also interesting they investigated binaural sound then too. You might be interested in this paper on the famous demo from 1933 with 3 channel sound. It has a picture of the mannequin they used for binaural recording.

https://acousticstoday.org/wp-content/uploads/2018/06/The-New-Age-of-Sound-Gabrielle-E.-O%E2%80%99Brien.pdf

View attachment 353840

I wonder if I have any photos of "Harvey the Head" from BTL, it used to sit on a shelf in my lab. I'm pretty sure Lucent has disappeared it.

 

[Justdafactsmaam]

I got most of that on discrete quad or better, no problem.

Most? No you don’t. Most never existed on Discrete Quad. And most quad mixes were an after thought.

“original Dolby TrueHD encoded recordings of John Coltrane, Jimi Hendrix, Pink Floyd, Bob Dylan, Yes, Genesis, Ivan Moravec, Yuja Wang, etc etc etc “ Are unicorns. Mythical creatures. Sorry you can’t have it both ways.

You can still be engaged and have fun with Dolby Atmos. But you can’t get high fidelity accuracy to the original recordings and the artists’ original intent with the vast majority of commercial recordings throughout the years in TrueHD encoded Dolby Atmos.