[Orpheus]

yes..... but now there's a dillema then. if you agree with what we've said earlier.... many SACD's SHOULD NOT sound better than DVD-A's. the fact that it does might actually present some evidence that some opinion's are more psychologically based than actually based on good hearing.... at least the ones made from digital masters.

 

[stuartr]

Or, as grinch so eloquently stated, perhaps there is a difference which is not easily measurable by machines? That is a very real possability as well. Sound is not simply a frequency...it has many more aspects...

 

[daniel422]

Now wait a second....sound IS just a frequency. Measurements and specs can only tell you things if you know how to interpret them (in a physical sense) and you are measuring the right things to beging with. That we don't have an exact defined set of measurements that say something sounds "good" or "bad" is beside the point. There are some things that can clearly be shown to affect the sound one way or another, and these can be measured.
As for DVD-A vs SACD -- I've got regular CDs that sound better than either of these setups in most cases. I'd argue that the care and precision taken in the recording process REGARDLESS of medium is far more important and critical than any technology used -- be it PCM. DSD, or analog.
The other problem is the near impossibility of reviewing similar titles on different formats. Even with multi-layer hybid SACDs, the redbook audio tracks are often a seperate mastering from the SACD. DVD-A is about the same, as many of the 2 channel mixes are downmixed 6 channels versions and there are pitifuly few true 24/192 recordings.
SACD, DVD-A -- they really shouldn't SOUND all that different -- they both possess audio reproduction capabilities far beyond the range of dogs' hearing, let alone humans. I'm betting 99.99% of any difference you hear between formats is due to the mastering process and not the format itself.

 

[sacd lover]

SACD is not the same process as PCM. 24/96 or 16/44 is encoded in the same manner but at different densities. SACD samples at 64x the PCM rate but is broken down to the simplest possible form; 2 characters. There is much debate and claims that to be sonically invisible you need a sampling rate of atleast 350 khz to prevent ringing. Everyone says upsampling cant add more information, so why does it sound better? I know it does firsthand; perhaps its the sampling rate which allows less radical filters that dont ring as much. SACD has a sampling rate that is well above this theoretical minimum, PCM does not. Furtermore with two characters the data stream has two options up or down; PCM has many (16/24) characters that can add in errors if they are not exact. Encoding is about transcribing the waveform and SACD seems to do this better. Just as upsampling seem to improve REGULAR 16/44 PCM, maybe converting PCM to SACDs higher sampling rate reduces distortions inherent to PCM encoding. Mastering, original production quality etc all done well figure in on producing a superior product. However, SACD, all things being equal (which they rarely are) sounds better from most peoples perspective who have heard both. SACD has an ease thats just easier to listen to in my experience. SACD vs PCM is not comparing apples with apples.

 

[NathanJM]

Quote:

Originally posted by daniel422 Now wait a second....sound IS just a frequency.

Hey sorry I just can't agree with that statement. First off, it seems to me that every wave has both an frequency and an amplitude to it. In this case of sound, this is created by pessure oscillations in the air (or other medium). Additionally, the same propagations maybe perceived differently based on the transmission/reception medium, and whether or not the observer is moving toward or away from the sound source which would yield the illusion of frequency change, but no actual such change in the waveform exists. Therefore, sound, from a basic and from a perceptual standpoint, is MORE THAN just a frequency.

 

[daniel422]

Quote:

SACD is not the same process as PCM. 24/96 or 16/44 is encoded in the same manner but at different densities. SACD samples at 64x the PCM rate but is broken down to the simplest possible form; 2 characters.

OK, I'll agree with this (although PCM starts out as 1-bit, too and higher sample rates -- 24/96 and 24/192 used by DVD-A) are much closer to this higher sample rate with additional bit depth to make up for lack of sample rate.
Let's not forget that sample rate can easily be traded for bit depth. It's BANDWIDTH that's really important.

Quote:

There is much debate and claims that to be sonically invisible you need a sampling rate of atleast 350 khz to prevent ringing. Everyone says upsampling cant add more information, so why does it sound better? I know it does firsthand; perhaps its the sampling rate which allows less radical filters that dont ring as much.

I've never heard this (350k) before. Where does this 350k number come from? Nyquist theory states you need greater than 2 times the highest desireable reproducable frequency to exactly reproduce a signal. I will agree that the higher sampling rates make it much easier to design the low-pass filter on the outputs of your DAC -- my personal theory as to why upsampling sounds "better" (or at least different). This output filter (IMHO) is one of the key stumbling blocks of most DACs.

Quote:

SACD has a sampling rate that is well above this theoretical minimum, PCM does not.

OK... well, technically I can create a PCM stream that samples at 350KHz. (I just couldn't play it on anything) I'm just whining here...

Quote:

--cut-- However, SACD, all things being equal (which they rarely are) sounds better from most peoples perspective who have heard both. SACD has an ease thats just easier to listen to in my experience. SACD vs PCM is not comparing apples with apples.

Well, I'd definitely agree with this. Almost impossible to compare directly.

 

[daniel422]

Quote:

Hey sorry I just can't agree with that statement. First off, it seems to me that every wave has both an frequency and an amplitude to it. In this case of sound, this is created by pessure oscillations in the air (or other medium). Additionally, the same propagations maybe perceived differently based on the transmission/reception medium, and whether or not the observer is moving toward or away from the sound source which would yield the illusion of frequency change, but no actual such change in the waveform exists. Therefore, sound, from a basic and from a perceptual standpoint, is MORE THAN just a frequency.

You're killing me

Yes, a GROSS simplification on my part. Of couse it has magnitude, direction, etc. I think my point was that of the prototypical engineer -- "If I can hear a difference I can measure a difference." I just mught not understand what I'm hearing vs what I'm measuring. Always the trickiest part from an engineering standpoint.

 

[markl]

Quote:

The other problem is the near impossibility of reviewing similar titles on different formats. Even with multi-layer hybid SACDs, the redbook audio tracks are often a seperate mastering from the SACD. DVD-A is about the same, as many of the 2 channel mixes are downmixed 6 channels versions and there are pitifuly few true 24/192 recordings. SACD, DVD-A -- they really shouldn't SOUND all that different -- they both possess audio reproduction capabilities far beyond the range of dogs' hearing, let alone humans. I'm betting 99.99% of any difference you hear between formats is due to the mastering process and not the format itself.

How many SACDs/DVD-As have you owned? How many SACDPs/DVD-A players have you owned? What other equipment did you use in your evaluation of the new formats?

If the difference between SACD/DVD-A and CD is negligible, where is the cut-off before sound starts to degrade? Going from 16 bits/44.1 KHz down to 14-bit, 32 KHz? 12-bit at 22 Khz? Obviously if lower bit and sampling rates don't matter. why not listen to 1-bit recordings on 1Khz sampling rate?

Mark

 

[daniel422]

I should have said after the nyquist rate sampling rate doesn't matter.

As for DVD-A and SACD players -- I own 2 DVD-A, and have had the opportunity to hack about 6 more. I've never owned SACD, but I've reviewed (and hacked) 3. I've listed to countless players at CES the last 3 years and the Stereophile hi-fi show as well.
I'm not saying that SACD and DVD-A don't have significant advantages over CD -- they do (I think I mentioned that before). What I'm saying is that a carefully produced (and reproduced -- both are key) redbook CD can sound just as good as SACD or DVD-A. I think these last two formats make it easier to achieve this level of sound (particularly on the reproduction end), but it is not.

 

[JaZZ]

Quote:

Originally posted by sacd lover There is much debate and claims that to be sonically invisible you need a sampling rate of atleast 350 khz to prevent ringing.

Could it be that there's something confused? Ringing is an issue when it happens within or near the audible frequency range, thus it is with redbook. DVD-A and SACD offer high-frequency extensions which are far beyond this danger. Even the lower DVD-A resolution (96 kHz) allows a low-pass filter at around 45 kHz, and it's not very likely that such a resonance has any audible impact on signals below 20 kHz (= ringing artifacts with fourier components within the audible freqency range). SACD is even much less critical in this regard with its sampling rate of 2822 (!) kHz (64 x 44.1 kHz). But the problem with SACD is the bit depth; it's only the equivalent of 6 bit at 22.1 kHz and about 9 bit at 15 kHz – just because there are not enough stair steps per second available given that every step is just 1 bit. It's been said that the SACD's high-frequency resolution is clearly below the redbook standard – and objectively this is true (as to the amplitude values).
Quote:

Everyone says upsampling cant add more information, so why does it sound better? I know it does firsthand; perhaps its the sampling rate which allows less radical filters that dont ring as much.

I agree with the better sound. Nevertheless, the appearing of ringing is not dependable of analog or digital (= upsampling) filters; every low pass filter produces the same resonance, independent of its nature, be it analog or an upsampling algorith or a combination of both. Upsampling DACs should measure the same ringing like non-upsampling ones.
Quote:

Just as upsampling seem to improve REGULAR 16/44 PCM, maybe converting PCM to SACDs higher sampling rate reduces distortions inherent to PCM encoding.

This is just as illogical and unlikely like any sound improvement through upsampling... so: why not!? Another theory: as mentioned, D/A conversion is a much simplier process with SACD than with PCM. Couldn't it be that market-conform DAC standards don't allow hi-rez-PCM DACs to sound at its best, whereas SACD DACs have less signal-degradation potential? So that even the format conversion, which is a pure mathematical process, BTW (at least I assume so), can't compensate for this?

JaZZ

 

[sacd lover]

Daniel 422, the 350khz sampling rate has nothing to do with the nyquist theory. If you look at digital PCM signals they are like a staircase and there is visible ringing. The 350khz sampling rate is the minimum to compleyely remove this ringing. DCS, the company has a white paper that discusses the ringing along with upsampling.In addition, Wadia digital built their designs on removing this problem at the expense of some roll off at the highest frequencies. We know upsampling alone, not resolution enhancement (which is sometimes looped together with upsampling with perpetual tech gear)does not add information. We know theoretically we cant hear above 20khz so it cant be missing overtones/ harmonics. SO WHY DOES UPSAMPLING SOUND BETTER? From all indications its the absence of this ringing. SACD avoids this now and PCM could with a higher sampling rate, however, when will we see 192khz sampling as the standard, let alone 384 khz, which would be the next logical and necessary step. As for SACD, I like it better than PCM but it has a long way to go to rival analog. The other reason I like SACD is with SACD we still have access to a PCM redbook layer we can copy and upsample (in most cases).

 

[JaZZ]

Quote:

Originally posted by sacd lover The 350khz sampling rate has nothing to do with the nyquist theory. If you look at digital PCM signals they are like a staircase and there is visible ringing. The 350khz sampling rate is the minimum to compleyely remove this ringing. DCS, the company, has a white paper that discusses the ringing along with upsampling.

Please explain what ringing has to do with 350 kHz! The common redbook ringing is the result of the 22-kHz low-pass filter, which BTW upsampling doesn't make needless – quite the opposite: it's part of it! Are you talking of 350 kHz upsampling or effective sampling rate?

 

[grinch]

Quote:

Originally posted by Orpheus if you agree with what we've said earlier.... many SACD's SHOULD NOT sound better than DVD-A's. the fact that it does might actually present some evidence that some opinion's are more psychologically based than actually based on good hearing.... at least the ones made from digital masters.

that's my whole theory, that sacd's are not made from digital masters.. hence their superiority.

actually i haven't really yet heard a dvd-a setup, but the more i read about it i really don't think i need to. i think sacd was designed to sound more like analog than cd or dvd-a. i think sacd was made to just sound better, not be easier to produce. so, because a lot of people use digital recording now, they probably won't be able to take advantage of this new technology, especially considering if they're recording digitally they probably don't care much about sound quality anyway.

 

[Wodgy]

Quote:

Originally posted by JaZZ But the problem with SACD is the bit depth; it's only the equivalent of 6 bit at 22.1 kHz and about 9 bit at 15 kHz – just because there are not enough stair steps per second available given that every step is just 1 bit. It's been said that the SACD's high-frequency resolution is clearly below the redbook standard – and objectively this is true (as to the amplitude values).

Yes, yes, yes! You've hit the nail on the head. The argument is a little simplistic, but accurate. There are many academic papers which do a more accurate analysis of the deficiencies of SACD, but the cores of the arguments boil down to this.

As I've said before, the only technical motivation for DSD is to make building an excellent DAC easier/cheaper. When cost is not an issue, PCM is inherently superior from both an engineering and theoretical standpoint.

 

[sacd lover]

It supposedly takes a 350khz sampling rate to make the low pass filter gentle enough to completely eliminate the ringing seen as a consequence of the decimation filter, and best illustrated by the 22khz brickwall filter with 16/44 pcm digital you noted. The nyquist frequency for 350khz would 175 khz. The audioband only needs to extend to a minimum of 20khz. They can then use a very gentle filter roll off and prevent the ringing of the steep slope filters.
350khz refers to the actual minimum sampling rate claimed to be needed for whatever digital encoding system, be it sacd, pcm or something yet to come, to remove the ringing. I read this article in stereophile but I havent found it yet to referance.