SACD Technology

From Edison cylinders to pre-amps to ProTools: talk about it here.
thomh
Posts: 124
Joined: Wed Sep 17, 2003 3:16 am
Location: Norway

Postby thomh » Thu Jan 22, 2004 4:13 pm

krabapple wrote:
Conclusion: LeeS will NEVER understand what you are saying, Thom. But kudos to you for trying , and trying, and trying to get the fundamentals across to him. I'm also quite interested to see how Mr. Hoffman answers your msot recent questions too...especially since he has already noted that his CD and SACD layers used *different mastering chains*.


You know, Krab, I have a feeling Mr. Hoffman is going to remain silent. He's probably happy now that he managed to twist it around and find that the fault of not hearing this "higher resolution" in the lower octaves is due to my crappy equipment and loss of hearing.

BTW, like your new avatar. You make this Norwegian proud.
Thom

User avatar
lukpac
Top Dog and Sellout
Posts: 4591
Joined: Wed Apr 02, 2003 11:51 pm
Location: Madison, WI
Contact:

Postby lukpac » Thu Jan 22, 2004 7:23 pm

BTW, while it seems fairly clear that increasing the sampling rate with PCM won't really do much good, what about the sampling *depth*? Wouldn't a higher depth result in a more accurate representation of the sampled frequencies?

Along these lines, does anyone have any idea how HDCD works? Read: how they get "20 bits" on a 16 bit CD.
"I know because it is impossible for a tape to hold the compression levels of these treble boosted MFSL's like Something/Anything. The metal particulate on the tape would shatter and all you'd hear is distortion if even that." - VD

User avatar
Rspaight
Posts: 4386
Joined: Wed Apr 30, 2003 10:48 am
Location: The Reality-Based Community
Contact:

Postby Rspaight » Thu Jan 22, 2004 10:59 pm

My understanding is that increasing the sampling depth (or word size or whatever) gives you potentially greater dynamic range/ lower noise floor. So yeah, things like decays and low-level detail probably benefit from more bits, but the *frequencies* themselves won't be any more "accurate," at least once you have enough bits to capture a reasonable dynamic range (I think 16-bit gives you 96dB).

For PCM, at least, the way I think of it is, sound vibrations (your friendly sine waves) have basically two properties -- speed (frequency) and amplitude (volume). The sampling rate (speed of data capture) gives you your frequency, and the sampling depth (size of sample) gives you your volume. This is probably a gross oversimplification, but it seems to be close enough to right to be useful.

I'm probably all wrong, and Thom will patiently correct me. He's very patient, you know.

As far as HDCD goes, other than vague generalities like "clever bit-shifting hocus-pocus," I haven't a clue. I'm kinda curious myself. (They probably don't go out of their way to explain it in detail, said he who is too lazy to go look it up.)

Ryan
RQOTW: "I'll make sure that our future is defined not by the letters ACLU, but by the letters USA." -- Mitt Romney

User avatar
lukpac
Top Dog and Sellout
Posts: 4591
Joined: Wed Apr 02, 2003 11:51 pm
Location: Madison, WI
Contact:

Postby lukpac » Thu Jan 22, 2004 11:42 pm

I think you're right in regards to "speed and volume", but my sticking point is exactly how accurate that "volume" is. How much does that matter?
"I know because it is impossible for a tape to hold the compression levels of these treble boosted MFSL's like Something/Anything. The metal particulate on the tape would shatter and all you'd hear is distortion if even that." - VD

User avatar
Rspaight
Posts: 4386
Joined: Wed Apr 30, 2003 10:48 am
Location: The Reality-Based Community
Contact:

Postby Rspaight » Fri Jan 23, 2004 9:50 am

Not sure on that myself. In the Steve/Thom discussion quoted above, Steve uses the sound of a snare echo on one of his CCR SACDs as evidence of SACD's superiority -- the SACD captures more detail. That seems to me to be a clear case of more accurate "volume" (though of course DSD does "bits" very differently than PCM), since it isn't the tonal accuracy that's in question, but he ability to resolve subtle amplitude attributes. Detail, not tonal accuracy, is where SACD shines for me.

(I'm tempted to chalk up the "smoother highs" so many rave about on SACD to less processing during mastering and better electronics in the converters -- the stuff Sony totes around to the different labels is primo gear -- not necessarily something in DSD itself.)

I do know that some on SH.tv say that they hear more of a benefit from more bits than higher sampling frequencies, which does make some sense...

Ryan
RQOTW: "I'll make sure that our future is defined not by the letters ACLU, but by the letters USA." -- Mitt Romney

User avatar
krabapple
Posts: 1615
Joined: Wed Apr 16, 2003 4:19 pm

Postby krabapple » Fri Jan 23, 2004 12:42 pm

For an intersting, if sometimes quite technical, discussion of the silliniess of ultrahigh sampling rates,
see this thread on Usenet, and pay particualr attention to the contributions of noted converter designer
Dan Lavry, who comes right out and says, '192 kHz is a crock'.

Google thread
"I recommend that you delete the Rancid Snakepit" - Grant

thomh
Posts: 124
Joined: Wed Sep 17, 2003 3:16 am
Location: Norway

Postby thomh » Fri Jan 23, 2004 4:21 pm

krabapple wrote:For an intersting, if sometimes quite technical, discussion of the silliniess of ultrahigh sampling rates,
see this thread on Usenet, and pay particualr attention to the contributions of noted converter designer
Dan Lavry, who comes right out and says, '192 kHz is a crock'.

Google thread


Krab, this is the same thread I posted a link to in Numbers Game thread. It is an excellent read, though.

I have had some great email correspondence with Dan and he is extremely knowledgeable and knows his craft. And he makes perfect engineering sense.

Defintely one of the good guys.
Thom

thomh
Posts: 124
Joined: Wed Sep 17, 2003 3:16 am
Location: Norway

Postby thomh » Fri Jan 23, 2004 5:28 pm

Let us talk a bit about bits.

Bitdepth (word size) is directly related to dynamic range, so the more bits there is available the lower the broadband noise floor and hence the potential for more information.

It is very important to keep in mind that when we talk about dynamic range we are also talking about resolution. The two are unbreakable. You increase one and you also increase the other. Think about it, it makes perfect sense.

What you also need to factor in here is that digital processing bits is different from AD and DA bits. When processing a lot of data one looses bit accuracy because of rounding errors. A lot of DSP causes some error and that is where more bits come in. This is why more bits during the recording stage is a *very* good thing. And why when you go from 16 to let's say 20 bits, you will probably hear an improvement.

Equally important is that once we have a signal with a certain dynamic range, as is the case with an analog master tape, we CANNOT IMPROVE IT. Sampling can only capture what is there.

So given a master tape where there is nothing above 15 kHz then 16 AD bits is plenty and so is 44.1 kHz sampling rate. Increasing the bitdepth or the sampling rate will do nothing.

Ryan is correct in saying that the lower the noise floor the more you are able to reveal things like echo decay and other low level details. But remember, *once a dynamic range is set then it cannot be improved upon*.

What SH wants you to believe is that a 30+ year old analog master tape which, according to him, has nothing above 15 kHz somehow exceeds the bandwidth set for Redbook. A *good* analog tape deck probably has dynamic range of around 70 - 80 dB. That is equal to about 14 bits (2^14 gives a DR of 84 dB). A 16/44.1 kHz sampling should be able to capture all the information contained on this tape including the echo.

So there is something else at work here. As noted by Krab, his CD and SACD layers used *different mastering chains*. Can we assume a less than perfect PCM conversion is culprit here? Or has he proven the sampling theorem wrong?
Thom

User avatar
lukpac
Top Dog and Sellout
Posts: 4591
Joined: Wed Apr 02, 2003 11:51 pm
Location: Madison, WI
Contact:

Postby lukpac » Fri Jan 23, 2004 5:59 pm

thomh wrote:So given a master tape where there is nothing above 15 kHz then 16 AD bits is plenty and so is 44.1 kHz sampling rate. Increasing the bitdepth or the sampling rate will do nothing.


While it's agreed that a higher sampling rate won't help, I don't see how the dynamic range is tied to the frequency range. You can certainly have something that's very dynamic but limited to a small band of frequencies, can you not?

Ryan is correct in saying that the lower the noise floor the more you are able to reveal things like echo decay and other low level details. But remember, *once a dynamic range is set then it cannot be improved upon*.

What SH wants you to believe is that a 30+ year old analog master tape which, according to him, has nothing above 15 kHz somehow exceeds the bandwidth set for Redbook. A *good* analog tape deck probably has dynamic range of around 70 - 80 dB. That is equal to about 14 bits (2^14 gives a DR of 84 dB). A 16/44.1 kHz sampling should be able to capture all the information contained on this tape including the echo.


That's assuming the formula you're using has more than enough resolution. Heck, you could have "84 dB of dynamic range" with *one* bit, but the resolution would be terrible (you'd either have -84 dB or 0 dB). You could represent that same 84 dB with 32 bits - each bit would just count for half of the volume difference of a 16 bit representation.

Make sense?
"I know because it is impossible for a tape to hold the compression levels of these treble boosted MFSL's like Something/Anything. The metal particulate on the tape would shatter and all you'd hear is distortion if even that." - VD

thomh
Posts: 124
Joined: Wed Sep 17, 2003 3:16 am
Location: Norway

Postby thomh » Sat Jan 24, 2004 8:06 am

lukpac wrote:While it's agreed that a higher sampling rate won't help, I don't see how the dynamic range is tied to the frequency range. You can certainly have something that's very dynamic but limited to a small band of frequencies, can you not?


It's the source that sets the dynamic range. And in this case it's a 30+ year old analog master tape.

That's assuming the formula you're using has more than enough resolution. Heck, you could have "84 dB of dynamic range" with *one* bit, but the resolution would be terrible (you'd either have -84 dB or 0 dB). You could represent that same 84 dB with 32 bits - each bit would just count for half of the volume difference of a 16 bit representation.


These are non linear systems. DSD is such a system where you have 1bit which, through noise shaping technology, manages 120 dB DR up to 20 kHz. Problem is that it nose dives from there. And while Redbook manages 90 dB within the same bandwidth DSD does have the edge here. But that presumes that the source in question goes above 90 dB and there is no analog master tape in existence which does that.

Oh, and 24/96 PCM has a DR of *144,5 dB* with a smooth and even noise floor. So I do not know what the big deal is with DSD.

In a *linear* system (like PCM or analog) dynamic range and resolution are unbreakable. You cannot have an increase in one without an increase in the other.
Thom

User avatar
krabapple
Posts: 1615
Joined: Wed Apr 16, 2003 4:19 pm

Postby krabapple » Tue Jan 27, 2004 1:35 pm

thomh wrote:Let us talk a bit about bits.


So given a master tape where there is nothing above 15 kHz then 16 AD bits is plenty and so is 44.1 kHz sampling rate. Increasing the bitdepth or the sampling rate will do nothing.

Ryan is correct in saying that the lower the noise floor the more you are able to reveal things like echo decay and other low level details. But remember, *once a dynamic range is set then it cannot be improved upon*.


This is something I admit to being unclear on -- even though I bet it's been explained to me before (stupid *brain*) -- if the dynamic range is the diference between the loudest and quietest bits, and you applly NR that makes quiet bits audible that were previously masked, haven't you at least increased the *perceived* dynamic range?
"I recommend that you delete the Rancid Snakepit" - Grant

thomh
Posts: 124
Joined: Wed Sep 17, 2003 3:16 am
Location: Norway

Postby thomh » Wed Jan 28, 2004 4:49 pm

krabapple wrote:
thomh wrote:Let us talk a bit about bits.


So given a master tape where there is nothing above 15 kHz then 16 AD bits is plenty and so is 44.1 kHz sampling rate. Increasing the bitdepth or the sampling rate will do nothing.

Ryan is correct in saying that the lower the noise floor the more you are able to reveal things like echo decay and other low level details. But remember, *once a dynamic range is set then it cannot be improved upon*.


This is something I admit to being unclear on -- even though I bet it's been explained to me before (stupid *brain*) -- if the dynamic range is the diference between the loudest and quietest bits, and you applly NR that makes quiet bits audible that were previously masked, haven't you at least increased the *perceived* dynamic range?


AFAIK, there is a trade-off here. With noise reduction we introduce non-linearity into a linear system. When we talk about the resolution of a system we are actually talking about resolution in amplitude domain (dynamic range) and resolution in the time domain (bandwidth). Noise reduction only affects the amplitude domain. You get a bigger dynamic range, but the overall system resolution does not get any better. In fact, it makes it worse. Witness how during crescendo passages you can hear the noise floor rising. So with a combination of smaller and louder signals, those smaller ones will get buried in the noise floor. So the payoff is more simple dynamic range at the expense of linear resolution.
Thom

thomh
Posts: 124
Joined: Wed Sep 17, 2003 3:16 am
Location: Norway

Postby thomh » Wed Jan 28, 2004 4:55 pm

Rspaight wrote:Not sure on that myself. In the Steve/Thom discussion quoted above, Steve uses the sound of a snare echo on one of his CCR SACDs as evidence of SACD's superiority -- the SACD captures more detail. That seems to me to be a clear case of more accurate "volume" (though of course DSD does "bits" very differently than PCM), since it isn't the tonal accuracy that's in question, but he ability to resolve subtle amplitude attributes. Detail, not tonal accuracy, is where SACD shines for me.

(I'm tempted to chalk up the "smoother highs" so many rave about on SACD to less processing during mastering and better electronics in the converters -- the stuff Sony totes around to the different labels is primo gear -- not necessarily something in DSD itself.)

I do know that some on SH.tv say that they hear more of a benefit from more bits than higher sampling frequencies, which does make some sense...

Ryan


You know, Ryan, there is no guessing at all here. Audio is just physics and when we come down this level, certain rules apply. I know that the audiophools on the SH forum would want everyone to believe that SACD and analog is above science, but, alas, it just ain't so. They cling to analog (and now SACD) like it was some kind of religion and will resort to any kind of techno-babble inorder to prove its superiority.

A waveform has 2 dimensions, amplitude and time. If you sample a 20 kHz signal at 44.1 kHz, 96 kHz or higher, it is within Nyquist of all and is captured *100%*. No amount of handwaving from the audiophools will change that. This is a *fundamental* principle of the sampling theorem. What this means is that you get a wave of the *exact* amplitude you put in and positioned in time *exactly* as it went in. In other words, the phase information is preserved. To claim that DSD does this more accurately is to claim that it somehow is able to do better than 100%.

The system's ability to resolve a signal is totally dependent on its broadband noise floor. The noise floor will always be there whether there's no signal or not and the presence of *any* noise *anywhere* in the system limits the resolution and hence the ability to represent signals unambiguously. Any time you lower this noise floor you increase the potential for more real information to be communicated through the system.

Now, as I said in the previous post, through noise shaping technology, Sony claims that DSD delivers about 120 dB over the entire audible bandwidth. That is equal to around 20 bits in the PCM world. No big deal, really. Redbook, properly dithered, has about 93 dB. So again, DSD does have better *potential* at "resolving the "subtle amplitude attributes" than does Redbook. But that presumes that the *source* exhibits greater than a 90 dB dynamic range. If it does *not* then both systems should have the *same* potential.

So higher bit depth is good, but higher sampling rate is not so good and going higher than what is necessary to get the job done is a waste. Dan Lavry makes excellent technical sense when he says that higher sampling rates yields less accuracy. Think about it, if you have more time to charge capacitors, to settle opamps, f.ex., you *are* more accurate. Science and engineering back this up.

As I have said before, if a PCM conversion is *not* able to extract all information from an analog tape with nothing above 15 kHz and DR probably limited to 60 - 70 dB then I would defintely start shopping around for a new converter.

If SH, LeeS or Joe Blow is so sure of SACDs superiority, then present their case to the AES for peer review. Sony has had ample chance to do this. But as of today, they have not. I wonder why?
Thom

User avatar
lukpac
Top Dog and Sellout
Posts: 4591
Joined: Wed Apr 02, 2003 11:51 pm
Location: Madison, WI
Contact:

Postby lukpac » Wed Jan 28, 2004 5:31 pm

thomh wrote:A waveform has 2 dimensions, amplitude and time. If you sample a 20 kHz signal at 44.1 kHz, 96 kHz or higher, it is within Nyquist of all and is captured *100%*. No amount of handwaving from the audiophools will change that. This is a *fundamental* principle of the sampling theorem. What this means is that you get a wave of the *exact* amplitude you put in and positioned in time *exactly* as it went in.


Not exactly. The *exact* amplitude falls somewhere in an infinite range, while digital is inherently finite. On a CD, for example, you can have ~65k values at any point in time, while in reality what you're sampling will almost always fall somewhere in between two of those values. That isn't to say that 65k values aren't enough (I'll leave that open to debate), but to say that you get the *exact* amplitude is untrue.
"I know because it is impossible for a tape to hold the compression levels of these treble boosted MFSL's like Something/Anything. The metal particulate on the tape would shatter and all you'd hear is distortion if even that." - VD

User avatar
Rspaight
Posts: 4386
Joined: Wed Apr 30, 2003 10:48 am
Location: The Reality-Based Community
Contact:

Postby Rspaight » Wed Jan 28, 2004 7:30 pm

Thanks for the ever-so-patient elaboration, Thom.

Not exactly. The *exact* amplitude falls somewhere in an infinite range, while digital is inherently finite. On a CD, for example, you can have ~65k values at any point in time, while in reality what you're sampling will almost always fall somewhere in between two of those values. That isn't to say that 65k values aren't enough (I'll leave that open to debate), but to say that you get the *exact* amplitude is untrue.


Not nearly as exact as dragging a stylus through a plastic groove, right?

Ryan
RQOTW: "I'll make sure that our future is defined not by the letters ACLU, but by the letters USA." -- Mitt Romney