[Omega]

Quote:

Originally Posted by Pio2001 /img/forum/go_quote.gif I disagree. You are comparing reproduction of sound with production of color. The right analogy would be either comparing production of both, thus asking "is this sound perfect", listening to a musician, or comparing reproduction of both, thus asking "are these two blue the same ?" The problem of sound reproduction is completely technical. There are just two major difficulties : first, the means required to perform accurate sound reproduction are far beyond the normal use of a home stereo system, and include room acoustics and must take into account the listener's position. 2-channels stereo implies that the listener is sitting still. If the listener moves, we need multitrack recording, one speaker per original instrument, that move exactly like the original (e.g. the speaker playing an opera singer will have to walk and rotate exactly like the singer did during the recording) plus extra speakers to simulate exactly the reverberation of the recording location. Second, since the late 70s, a new paradigm has shifted the home high fidelity from the professionnal one (used in theatres, for example) : measurments have no correlation with listening experience. This paradigm have vastly hindered the development of consumer hifi. Now, we can see that much more effort are put into power cables than into, say, equalisation.

You're free to disagree of course, but you haven't addressed the meat of my argument...that the issue is philosophical rather than technical. Positioning, level matching, room interaction, timbre--these are all technical phenomena.

I am asserting that, given the perfect sound reproduction device that fully satisfies all your technical challenges, different people will still judge "good sound" differently. Taste is subjective. I can describe, in a very well-specified way, what tastes good to me....but I would be arrogant to assume that it must "taste" good to you as well.

I used color perception as an example because it is a well-known philosophical dilemma, that cuts straight to the issues. For example, see here (Univ. of Waterloo Philosophy). Perception of color differs from perception of sound only in biological implementation.

That said, I was hoping to submit an orthogonal (and hopefully useful) thought to a thread that was beating about more of the same. But I don't enjoy philosophy when I cannot see my opponent and share a good glass of something intoxicating

 

[jvgig]

The discussion is not one of which subjectively "sounds better", but how close are the physical quantities to those of the original. Accuracy can be measured, "better" cannot.

I think that high end recording studios would have to assume that the listener would have the ideal listening environment for the source tracks that they are providing. Once the user has the most accurate source, then they have to determine how to optimize their system/room/listening habits to achieve the desired sound (may or may not be accurate). In theory, a computerized program and meters could measure sound pressures and apply a curve/delay/etc to the source to counteract any environmental factors.

On the color example, whether you perceive blue#1 (475nm) as 470nm, 475nm, or 480nm is a different question than what we are addressing. Either way, the wavelength hitting your eye has a specific wavelength associated with it, not a range. This is environmental accuracy, which I think would be similar to audio accuracy. Thoughts? The philosophical question is whether "better" is defined by saying that the reproduction of that 475nm light is best if the new source is 475nm as was originally presented, or if it should be "colored" to make it a deeper blue of 460nm. This obviously has the potential to change the overall impact of the piece which would not be accurate, but may be preferred by some viewers.

 

[jcx]

I think an interesting question is why binaural or "virtualized” mixes aren't generally available given the huge fraction of mp3's being listened to on headphones

the crossfeed that can be done in a player or amp after the fact with stereo source that was mixed for room loudspeakers should be way inferior to mixing for headphone listening to start

 

[b0dhi]

Quote:

Originally Posted by Omega /img/forum/go_quote.gif You're free to disagree of course, but you haven't addressed the meat of my argument...that the issue is philosophical rather than technical. Positioning, level matching, room interaction, timbre--these are all technical phenomena.

I think it's fair to assume that what's required is achieving a presentation of audio that equals whatever the person would hear of the world naturally. You know - if you're listening to a recording of a concert, it sounds to you exactly as though you were physically there.

Delving into qualia and deeper issues of taste is unnecessary here IMO because issues of "taste" only exist because we're so far from "real" reproduction today. If headphones could "take us there", issues of taste would no longer exist. I've never heard anyone complain that their own ears are too bright or too bassy or the mids are too far back when they're sitting in a room having a conversation

 

[Kees]

The answer to the original question is easy:
Since you don't have the original to compare the recording with, it is impossible to check/measure if the recording is good.
That's why the only reference is what people remember from the original, and those memories are subjective and personal.

 

[Pio2001]

Quote:

Originally Posted by Keithpgdrb /img/forum/go_quote.gif I forgot which company right now, but they developed a mic that reproduced the voice exactly, with no coloration at all, and people hated it, it was a flop.

In this article about PSB speakers, Paul Barton says

Quote:

“Take the Stratus Gold, which has been one of our most successful speakers. We designed it for a flat frequency response [in which no particular frequency range – bass, midrange, or treble – is emphasized] in the anechoic chamber. But thanks to blind testing, we realized that a perfectly flat response sounds a little too bright in an actual room. So we toned down the response a little in the 5 kilohertz [treble] range, and right away it became a winner. If you don’t do this kind of testing, you’ll probably never uncover those kinds of flaws.”

 

[milkweg]

It's subjective because no two humans hear exactly the same so what one persons says something sounds like is not how it sounds to you. Some like chocolate and others prefer vanilla.

 

[gregorio]

If what was being said here were true or accurate, we wouldn't need producers.

Let's take the example of an orchestral recording. As we sit in a concert hall 100 musicians are playing from 100 different positions. All these sound sources are then reflecting off all the walls, ceiling, floor, seats and anything else in the concert venue. What reaches our ears is comming from every direction (vertically and horizontally) and the brain is then very clever at looking at the phase and frequency of all these signals and discarding many of them as irrelevant. To reproduce this perfectly would require thousands of perfect microphones and thousands of perfect speakers and systems capable of storing and outputting thousands of channels of audio. I'm sure you'll all agree that this is not exactly a practical solution!! So with a limited number of mics and just two speakers, engineers and producers are required to create the illusion of sitting in a concert hall rather than being able to recreate the reality. This is how it has always been and will be for at least the forseeable future.

Also bare in mind that most top class recording control rooms have spent a great deal on acoustic treatment to give flat frequency responses, to randomise reflections and produce an RT60 of around 0.4secs. The point four secs (approx) is chosen to mimick the average home listening environment. Obviously this means that listening to a production on cans is never going to represent what the producer created (it's going to sound much drier).

G