Is digital audio intrinsically "broken"?

[MylesBAstor]

Then there is this bizarre thing about acoustics and speaker design. It seems hardly anyone agrees with anyone else. Surely this is proof point that very hard to demonstrate what is right. Or else, we would only be buying products that follow that scheme. Magico, Wilson and Revel speakers produce three different sounds. How could they all be right?

Surely it must be a crap shoot then that we think one is more right than the other. Likely as we go from song to song, the level of fidelity to the original recording as heard by the talent is varying all over the place. Maybe that is why we itch to upgrade as this happens to us. But maybe we are imagining it all because we never know what is right, what is wrong!

Thankfully it all sounds dynamic and grand. So we feel good, thinking we are getting it all. I don't think we are. We are getting a carbon copy with colors shifted all over the place and we don't even know it. And worse yet, the color shift depends on what the source looks like!

You guys really wanted to get me started this way? I think not.

HP asked the same question, a decade or two ago, about cartridges.

 

[fas42]

A little story. The friend I've mentioned a number of times recounted this some time ago, as a joke on himself.

He'd been pretty active in the hifi game for a while, listening to quite a number of ambitious systems, so he had decent experience of how good sound was from the conventional rigs. One day, he was in an office environment, people had on some of their personal music devices as background for their working. Then, he heard down a hallway, coming from quite a distance some beautiful music, it was rich, sonorous, it was a delight to his ears; it had all the hallmarks of being "right". He followed it down the passageway, and determined it was coming from behind a certain closed door. Lucky fellow, he thought, he's got a top notch system set up in his office! He knocked, chap opened the door, and my friend looked quickly around. Nary a stereo component to be seen!

"Where's that marvellous music coming from??". "Ssh", said the chap and pulled him inside, and then pointed to a shelf on the side of the room. There, in all its glory, was a windup gramophone, straight from the the label of "His Master's Voice", still producing rich, luscious sound, eminently satisfying in every way ...

Frank

 

[Orb]

.......
BTW, the book costs only usd 43.49 - and it is great reading!

http://www.amazon.com/Sound-Reproduction-Acoustics-Psychoacoustics-Loudspeakers/dp/0240520092

Yeah but one has to read it...
Is it available in comic strip with images?
More seriously, most like to discuss this hobby but not necessarily go through the intense theory/maths/science studies/etc in the level of detail required to have such a level of knowledge.
More fun to chat than to burn cycles taking in heavy papers or knowledge books on the subject, not being critical just the way many hobbies are; quick example many into astronomy but not the relevent theory-physics-astrophysics that is the foundation of the hobby or take the time to delve deeply into such theories and maths.
Who wants to study when they can have fun chatting with other like minded people (hypothetical question so no replies thanks )

That said there definitely will be some who do enjoy the theory so thanks for the link (although most probably have the book already who are ).
Cheers
Orb

 

[Phelonious Ponk]

Amir-Does this reinfoce the point I was tyring to make in an earlier thread where I said we don't really have good measurements for audio (if we have them at all)? It's funny that we can nail down the light spectrum with science and we can't nail down the audio spectrum.

Mark

I'm not Amir, but I think his point has to do with video having better standards, not better measurements. Or, to put more fine a point on it, it's about video having standards at all, which audio does not. The standards for the calibration of the camera and the calibration of the monitor would be like having a standardized set of measurements taken from some kind of full-range test tone, in the recording studio, and measured again from the speakers, at playback, so the two could be compared. There are a million arguments to be made against the idea, a million reasons why it would be difficult, or perhaps not work at all but if, in theory, you could make it work, it could greatly improve recording (at least you'd have a standard from which to manipulate) and it could potentially eliminate the old argument that, in video would go something like, "my monitor is not to brown, it's more natural." More accurately, I think it could have avoided that argument in the first place. Now that it's here, I'm afraid there's no killing it.

Tim

 

[Orb]

Ho boy, Amir, don't get me started! The one big hole in your analogy is that the eye is intrinsically (where have I seen that before?) a thousand times less sensitive than your ears. Dynamic range of vision is 60dB, of auditory is 120dB. And guess what ? It's pretty damn easy to get things accurate to 1 part in a thousand, falling off a log easy for measuring gear in particular. But once you go past that point it starts to get dicky, you have to be fussier and fussier to make sure you're really getting a reading that measures what you're really after, and that the gear is working that precisely. That's where the headache with audio is, that the ear/brain is smarter and more picky than the components and test instruments are set up to be ...

Frank

One needs to be careful on how they state sensitivity, and needs to be put into conext of cognitive perception/decision process.
Because in terms of the eye and sensitivity, we still do not understand how cosmic rays are picked up by the eye or nerve response in the retina.
That is pretty small and is also picked up by astronauts cognitively (average if I remember about 2 to 3 a minute in space).

But trying to compare both in terms of this discussion and more relevent.
Well the most basic suggestion is real world examples of 8bit/12bit/16bit/24bit colour-resolution images.
Isnt it accepted that 24-bit Trucolor is required for the best images perceived by someone (more than 16-bit colour anyway)?
Meaning it is sensitive enough in how it relates to the discussion raised by Amir.

Cheers
Orb

 

[amirm]

Ho boy, Amir, don't get me started! The one big hole in your analogy is that the eye is intrinsically (where have I seen that before?) a thousand times less sensitive than your ears.

Yet, that less sensitive situation has standards, and audio does not!

 

[microstrip]

Yet, that less sensitive situation has standards, and audio does not!

Amir,



I find curious that you say video has standards and audio not. But every time I read a review of a video product, fundamental aspects such as movement are appreciated subjectively, and even image quality is described with terms such as vibrant and colorful (from the review of a recent Epson projector at projectorcentral.com). A good friend of mine, who has a great home theater system feels that only by viewing he can appreciate the quality of a screen or a projector - measurements can help, but the final decision must be taken by viewing. He even has a selection of movie scenes he uses for that purpose. Does not this situation look familiar to us?

Besides, IMHO, comparing audio with video is misleading and usually only adds noise to our debates - I feel the same about cars or wine metaphors. Video is essentially a creative and dynamic process. The image is so far from reality that no one would carry a life versus image session.

 

[Orb]

Yet, that less sensitive situation has standards, and audio does not!

Amir you may know about imaging and colour depth.
Is it fair to say that for best perceived images the standard is set to Trucolor (24-bit depth) rather than 16 bit or less.
This is relevent IMO because we have a situation of comparing 16-bit color depth to 24-bit color, which I am sure Trucolor is the acknowledged standard for best images.
Therefore suggesting perceived vision is as great as audio for digital (whether digital image or audio).
Not a subject I know much about so putting it out there.

Cheers
Orb

 

[amirm]

Video resolution is tricky and not directly comparable Orb. From the presentation I post already: http://www.madronadigital.com/Library/VideoForAudiophiles.html

Put another way, it is only 8 or 10 bits for the critical black and white portion of signal.

 

[Orb]

Yeah understand about video,
but if talking eye sensitivity then image quality should also be part of it I would think.
Especially as it is possible to correlate perception to bit color depth of a digital image.
Is the difference in quality requirements down to static vs moving?
Or more related to the encoding technique, which is academic for a static digital image.
Edit:
That said, anyone know why 24-bit True color and higher seems to be the standard for the very best in perceiving digital images, compared to say 16-bit color depth.
Appreciate we are not disagreeing Amir, just trying to understand from Frank's perspective how the eye is deemed to be less sensitive and yet it seems it is critical to have greater than 16-bit color depth for photographic/digital images; 24-bit color is a heck of a lot and seems to be a minimum.

Cheers
Orb

 

[Phelonious Ponk]

Appreciate we are not disagreeing Amir, just trying to understand from Frank's perspective....

Good luck with that.

Tim

 

[amirm]

Is the difference in quality requirements down to static vs moving?

There is. There are tests to measure the dynamic resolution of video where some technologies like LCD don't do as well.

That said, anyone know why 24-bit True color and higher seems to be the standard for the very best in perceiving digital images, compared to say 16-bit color depth.

I should clarify that what we use for video is not the same as what you use in your computer. Your computer operates in RGB mode. We don't use that in video. Instead, we separate the color and black and white so that we can lower the resolution of the former independently. A 16-bit video system can actually represent many of the colors in 24-bit RGB. The relationship between the two is rather complex. Here is a visualization of it:

As you can see, YUV can have colors that are "illegal" in RGB. This, despite the sample resolution being less than RGB as a whole

Appreciate we are not disagreeing Amir, just trying to understand from Frank's perspective how the eye is deemed to be less sensitive and yet it seems it is critical to have greater than 16-bit color depth for photographic/digital images; 24-bit color is a heck of a lot and seems to be a minimum.

Cheers
Orb

No analogy is perfect. One can certainly find holes in my comparison. But taken as a whole, there is more attention to correctness in video.

 

[Orb]

Thanks Amir great info.
And with your info and a bit of digging appreciate how they are different but in a way both can be applied but in subtly different context.

Thanks again.
Orb

 

[fas42]

Another way of looking at it is that the eye starts to run out of puff when more than about 256 levels of "brightness" of a certain colour have to be distinguished. The eye is typically most sensitive to black and white, and so if you do like a colour chart of greys across the screen, or printed, going from total black to the whitest white, in a total of 256 steps, then you will great difficulty picking the edges from one shade to the next: the eye has run out of resolution. 256 steps requires 8 bits, for colour you need 3 "guns", red, blue, green , RGB; 3 times 8 gives you the magic 24 bits for Trucolour ...

Frank

 

[Old Listener]

More seriously, most like to discuss this hobby but not necessarily go through the intense theory/maths/science studies/etc in the level of detail required to have such a level of knowledge.
More fun to chat than to burn cycles taking in heavy papers or knowledge books on the subject, not being critical just the way many hobbies are; quick example many into astronomy but not the relevent theory-physics-astrophysics that is the foundation of the hobby or take the time to delve deeply into such theories and maths.
Who wants to study when they can have fun chatting with other like minded people (hypothetical question so no replies thanks )

Unfortunately, your observation hits the mark. I'm one of those people who would get more out of audio forums if there was more real knowledge being exchanged.

Bill

 

[Phelonious Ponk]

Originally Posted by Orb
More seriously, most like to discuss this hobby but not necessarily go through the intense theory/maths/science studies/etc in the level of detail required to have such a level of knowledge.
More fun to chat than to burn cycles taking in heavy papers or knowledge books on the subject, not being critical just the way many hobbies are; quick example many into astronomy but not the relevent theory-physics-astrophysics that is the foundation of the hobby or take the time to delve deeply into such theories and maths.
Who wants to study when they can have fun chatting with other like minded people (hypothetical question so no replies thanks )

If it were that simple it would be rather easy to understand. But I'm afraid many in the hobby avoid even light science and run in circles, hands on ears, babbling to keep out the heresy when any science is spoken to them. A deep knowledge of electronics and testing methodology is not required to understand what audio reproductions systems are and are not capable of. The willingness to believe in it is.

Tim

 

[microstrip]

If it were that simple it would be rather easy to understand. But I'm afraid many in the hobby avoid even light science and run in circles, hands on ears, babbling to keep out the heresy when any science is spoken to them. A deep knowledge of electronics and testing methodology is not required to understand what audio reproductions systems are and are not capable of. The willingness to believe in it is.

Tim

Tim,

The big issue is that the real reason of most debates is just our desire to discuss the connection between the objective and subjective sides of the hobby. If you accept to keep them separate you can discuss them with light concepts. But if you want to make the bridge, the light science will not allow you to progress, only to present a very reduced view of the problem. And in this situation, frequently the "truth" guy is just the best spoken. But happily we have experts in WBF!

 

[Phelonious Ponk]

Tim,

The big issue is that the real reason of most debates is just our desire to discuss the connection between the objective and subjective sides of the hobby. If you accept to keep them separate you can discuss them with light concepts. But if you want to make the bridge, the light science will not allow you to progress, only to present a very reduced view of the problem. And in this situation, frequently the "truth" guy is just the best spoken. But happily we have experts in WBF!

With all respect, Micro, I disagree. I think the bridge between the science and subjectivity is right in front of us, but it requires that those of us who believe in the science accept and honor preferences that don't align with the available science, and it requires those whose preferences don't align with science to accept and honor the science, while retaining the confidence to hold onto their preferences. It also requires that all of us look deeper than the surface, while being carefull not to sweat things too small to perceive. And to accomplish any of it, all of us, every one of us, have to understand how powerful bias is, and know that we are not immune.

Most of what blocks the bridge between science and subjectivity is ego.

Or perhaps you're talking about the bridge between engineering and perception? Same block, same problems. Much, much more interesting subject.

Tim

 

[amirm]

With all respect, Micro, I disagree. I think the bridge between the science and subjectivity is right in front of us, but it requires that those of us who believe in the science accept and honor preferences that don't align with the available science, and it requires those whose preferences don't align with science to accept and honor the science, while retaining the confidence to hold onto their preferences.

Tim

Tim, that doesn't leave any room for what we don't know. Let's take speakers. Do we know and agree on the same science? Or is that what you mean by "available" science?

 

[microstrip]

With all respect, Micro, I disagree. I think the bridge between the science and subjectivity is right in front of us, but it requires that those of us who believe in the science accept and honor preferences that don't align with the available science, and it requires those whose preferences don't align with science to accept and honor the science, while retaining the confidence to hold onto their preferences. It also requires that all of us look deeper than the surface, while being carefull not to sweat things too small to perceive. And to accomplish any of it, all of us, every one of us, have to understand how powerful bias is, and know that we are not immune.

Most of what blocks the bridge between science and subjectivity is ego.

Or perhaps you're talking about the bridge between engineering and perception? Same block, same problems. Much, much more interesting subject.

Tim

Tim,

Also with all the respect, IMHO, you are putting yourself in the zone of those who do not want to look deeper than the surface and already have the answers to everything ...

Do you agree that perceptual science is part of sound reproduction?