'HD Download debacle' investigation published on HFN&RR June 2011

AT&T did have the best amplifiers and speakers in the world, at least in the 1930s and 1940s, so I doubt that was a problem. And obviously, they would have focused on speech before music.

But why do you even bring it up? Even if Harry Nyquist had been brought in as a special consultant on PCM in the early 1970s and heard it through giant Magneplanars, and pronounced everything A-OK, you still wouldn't grant it any credence b/c all that matters to you is what you hear, and what others here if and only if they concur with you.

And by the way, the "Nyquist" theorem was actually developed independently by five or six researchers and as such officially carries a very long name, making it even more ironic you would personalize this to what he did or did not hear.

If there is a problem with digital audio, it most certainly is not the fallibility of sampling theory.

 

perhaps you can elaborate a little more ? Are you suggesting that given the same analog master tape...that the A/D 16 Bit will be as good as the A/D 24 bit ? I think not !

 

I think some people are hazy on the distinction between a theory (science) and a theorem (math). Nyquist's proof can be verified line by line. That doesn't necessarily mean it's applicable in all real-world circumstances, but his listening experience is as irrelevant as it's possible to get.

 

Why do you think not?

 

basic theory backed up by science. Given the same master tape....

16 Bit. 16 pulses to approximate the sampled analog wave form.
24 Bit. 24 pulses to approximate the sampled analog wave form.

Analog is what we hear correct ? The analog wave form is smooth, not chunky.

What do you think will make a smoother approximation of the analog wave form sample...16 or 24 pulse chunks ? This is not a trick question.

 

But, as has been written countless times, the D-->A process isn't about "connecting" chunks on a line.

I could ask a similar question of you. What better defines a straight line: two points or three points?

 

a line is not applicable. Sound waves are not flat lines... they are lines with varying degree's of amplitude over a given sample of the signal....

 

Just to throw another wrench in the mix now that we've gone back to the transparency of digital audio:

"Is Vinyl Really Better?"
http://www.itrax.com/Pages/ArticleDetails.php?aID=15&x=12&y=6

Edit: for those who haven't followed from the top of the thread, I submitted this because it's a follow-up by the same author who kicked off this debate, and not just re-energize Hi-Res/Digital vs Vinyl debates.

 

Yes it is, from my experience...

 

Sure. But the argument is that more information necessarily defines something better. That isn't necessarily the case. Two points are adequate to define a line. Similarly, with regard to digital to analog conversions, the question is whether the information in the system can, after conversion, recreate the input. While 24 bit implementations have real value with regard to things like digital transformations, it is unclear that they can always produce audible differences, especially with regard to rock music with limited dynamic range to begin with. Throwing more information at a limited system will not always improve the system.

 

Points well taken. From a rock music standpoint I do have DVD Audio hi-rez versions that I also have in redbook CD. To my ears the hi-rez does sound significantly better.

exmaples:
DVD Audio Fleetwood Mac Rumours
HDTracks Hi-rez version of Steely Dan Gaucho
DVD Audio - Yes Fragile
DVD Audio Miles Davis- TuTu
DVD Audio Donald fagen Kamakiriad
DVD Audio Grateful Dead - Workingman's Dead
DVD Audio - Chicago S/T (Note, the DVD says 24/192, when in fact it is really 24/96 as I ripp DVD Audio MLP files to hi rez wavs for playback on my system and can tell the actual bit depth, sampling frequency)


But as you know, it all comes down to mastering decisions whether one format sounds better than the other. All things equal 24 bit > 16 bit. I don't think you are arguing that point.

 

Well, take a 24 bit musical file, dither it down to 16 and listen!
Revealing system and discerning ears are certainly pre-requisite to hear the difference.
But I would be very surprised that there is absolutely no discernable difference.

(We are getting pretty close to this subjective vs objective thing, isn't it?)

 

Don't give 'em ideas...

Is it possible that D/A (not A/D) chips perform better at higher resolutions/bit rates even if the original data is lo(wer)-res?

 

Great article. If I had a really good DAC and a hard drive full of true 24-bit 96khz stereo recordings, my turntable would indeed collect dust.

Buddies who come over to watch the game see me putting records away and kid me about it - but when I play something then they ask me "where they can get that cd" or "would you rip it for me" and I tell 'em back "Ever seen Lee Oskar on cd? I haven't!!"

Is that not the theory behind making a cd player that decodes at 192khz (and costs +$1500)?

 

Well, rockitman, let's start with this:
You can't compare a cd to an SACD or DVD/A of the same title and learn anything about the medium unless you know the mastering is the same. So, if you know the cd contains the downsampled version of the hi-rez, ok, but usually the hi-rez mastering is different.

Second, while a 24-bit recording might be more accurate, the question is whether the difference is audible.

Third, no room is silent including an orchestra hall so the relevant dynamic range is from background noise to peak and that is hardly ever more than the 96 dB of 16 bit digital. Besides which, few audio systems will play the 125 to 135 dB necessary to exceed the background noise in the listening room by more than 96 dB so it's irrelevant.

 

Thought provoking stuff. Thanks. I will admit my digital stuff have been collecting dust over the past fews months since jumping into hi-end analog. It's nice to hear the music as it's meant to be and not a digital approximation...

 

I use 24bit and 16bit audio every day in my work. To my ears, the difference exists and is usually not so subtle.

I'm always puzzled when I read that the difference is not audible or doesn't benefit this or that musical genre. Even a 70's Rock album recorded in analog can sound better in 24bit resolution than in 16bit resolution.
Really, I don't understand why some of you guys are so hung up on proving the difference is imaginary and that 16/44.1 is already utterly perfect.

 

The music biz is full of charletons who pass off inferior products , separating the wheat from the chaff is the rightful preoccupation of folks on this forum.
Vinyl hucksters gave you Dexterised Beatles and deranged fake stereo
They thinned the records and reprocessed the vinyl, including even the label in the plastic mix!
They tried to purvey a hopeless 4 channel system on an inadequate carrier.
Cassette was promoted and then declared evil because it was killing the music business.
They developed a new digital carrier and then tried to copy protect it.
There is no end to their greed and vulgarity it seems.
However we all have fabulous music collections in varying formats of our own choice, something previous generations would have loved.
We got them by word of mouth and listening experience.
Hucksters offering indifferent High Resolution Down loads will not survive imho.
However it could be a long and painful process.
Bring on the digital threshing machine please

 

Have you paid attention to HDTracks's pricing structure? Note how the files with the "better" numbers cost a significant amount more? See, e.g., the Stones tracks. If there was no difference between them, it would kind of kill the incentive to charge people an arm and a leg for a product that offers absolutely no improvement.

 

edb15,

Feeling a little vindictive today? Please don't tell me what matters to me since you haven't the vaguest idea. The first thing that matters is manners.

Seems to me you've reached an awful lot of conclusions based on what you believe to be the case and won't let yourself be inconvenienced by facts, not to mention civility.

Just where did you see me question the validity of sampling theory? Don't you see this is something you ascribed to my question, with no logical basis for doing so?

I'm here to have fun and share experiences with folks. Your reading of my question has nothing to do with what I said or what I meant. But then, we've been here before, haven't we?

With that in mind, I wish you the best. I will ignore your posts and hope you have the confidence to return the favor.

Best regards,
Barry
www.soundkeeperrecordings.com
www.barrydiamentaudio.com

 

There's oversampling and upsampling. Similar operations but different in concept. Upsampling is a true sample rate conversion. Oversampling can get done internally inside the DAC chip as part of the DAC design. And of course you can upsample the audio and send that upsampled audio to a DAC that does some crazy high oversampling rates internally.

So the theory behind a CD player that decodes at 192khz or even higher depends on whether the CD player is upsampling or oversampling.

 

It might be due to what is called 'cognitive dissonance'. See here: http://en.wikipedia.org/wiki/Cognitive_dissonance

And on the second half of this article: http://healthland.time.com/2011/05/...will-grow-stronger-when-the-world-doesnt-end/

 

Barry, I didn't mean to transgress the boundaries of politeness and grace.

However, I know at least three other readers read your statement the way I did. Feel free to explain what you meant to me or them...

As for proving 16/44 is perfect--I am not aiming for that. I'm just saying if it is not, it cannot be due to the "invalidity" of the Nyquist theorem or the "jaggedness" of a 16-bit waveform nor due to dynamic "compression" as has been advanced in this very thread.

 

I agree that in this case, charging more for 24/176.4 files containing no more information than 24/88.2 is a con that should be avoided.
Nobody's forcing you to buy them. If you're quite content with 16/44.1, leave it at that. Eventually this pricing structure will dry up.