Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: More SACD gibberish (Read 15822 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

More SACD gibberish

An interesting article on SACD that only an "audiophile" could love:

http://www.chicagotribune.com/technology/c...echnology%2Dhed

And I thought Chicago Tribune writers were just misinformed on local politics, etc

"But CDs didn't offer perfect sound. Digital compression and equalization gave many CDs a thin, plastic, tinny sound. Music sounded artificial and dynamically dead"

Here we go again. . . .. .

:D:D

More SACD gibberish

Reply #1 –
I can't read the article (not registered).  But if it's referring to CD's in the first 5 years after they came out, they're right.  Digital mastering techniques were truly awful and the first few generations of CD's sounded BAD.  If you ever get your hand on a CD from about 1982-1988 (maybe not easy, some of them are no longer viable) give it a listen on some good quality gear & maybe give your opinion.

Recently, with over-use of dynamic compression, some CD's have started to sound shit again (to these ears)... overall tho, digital mastering has improved greatly.  It took maybe 50 years to get analog mastering perfected, what makes you think digital is so different?

Edit P.S... "Here we go again" applies to this thread, which you started (not the Chicago Tribune).    Also a P.S. question, have you ever ABX'ed a properly mastered SACD against the same recording on regular CD?

More SACD gibberish

Reply #2 –
Quote
"But CDs didn't offer perfect sound. Digital compression and equalization gave many CDs a thin, plastic, tinny sound. Music sounded artificial and dynamically dead"
Yeah, that is stupid.

When a format is fucked up because of the industry's greed polluting the delicate mastering technique, how does a new format in any way help this?


Edit : fixed quote

More SACD gibberish

Reply #3 –
Quote
Originally posted by fewtch I can't read the article (not registered).  But if it's referring to CD's in the first 5 years after they came out, they're right.  Digital mastering techniques were truly awful and the first few generations of CD's sounded BAD.  If you ever get your hand on a CD from about 1982-1988 (maybe not easy, some of them are no longer viable) give it a listen on some good quality gear & maybe give your opinion.

Recently, with over-use of dynamic compression, some CD's have started to sound shit again (to these ears)... overall tho, digital mastering has improved greatly.  It took maybe 50 years to get analog mastering perfected, what makes you think digital is so different?

Edit P.S... "Here we go again" applies to this thread, which you started (not the Chicago Tribune).    Also a P.S. question, have you ever ABX'ed a properly mastered SACD against the same recording on regular CD?


Oopps, I forgot about the Trib's stupid (and greedy) new requirement. The article talks about the Rolling Stones releasing SACD's and gives several incorrect bitrate figures on CD/SACD audio, among other things. I thought it was interesting and timely given the discussion on the MPC boards regarding SACD.

Also, I do have a few very early CD's, and they do sound terrible. But that was the early-mid 1980's; this is the early 2000's. Most audiophiles made thier peace with CD in the late 80's when they started to sound better, or they stuck to vinyl (I have both). CD's are starting to sound worse, I will agree with you there. . . .look at many classical albums, or rock albums like Rush's new "Vapor Trails" that are full of digital distortion. Yet a well-done CD sounds great.

Which brings up my final point: in the other thread, I discussed my own experience with SACD and how I could not distiunguish one SACD title from the remastered CD but could distinguish the badly-mastered CD from the SACD version. SACD is not necessary, IMO. Red Book audio can sound great if done correctly, and lately, for whatever reason, some CD's don't sound up to thier capacity. But that doesn't mean I'm going to go out and buy another format just because it has enough margin of error to enable better sound. . .. if the record companies would expend more effort into pressing quality CD's, there would be no reason to think about SACD. To me, SACD is like a 5-foot wide basketball hoop: no matter how bad you are, the ball would go in, whereas a CD is like a regulation hoop, in that you would need to be more accurate, but would achieve the same end results (i.e. the ball goes in the hoop). In otehr words, it may be easier to make good-sounding SACD (or on purpose as some as suggested) than CD, but that doesn't mean that a quality CD release can't sound just as good.



Edit : fixed quote

More SACD gibberish

Reply #4 –
Quote
Originally posted by Cygnus X1 In otehr words, it may be easier to make good-sounding SACD (or on purpose as some as suggested) than CD, but that doesn't mean that a quality CD release can't sound just as good.


Well, I don't think so. Once you have a good master, it's the same to put it on CD that on SACD. I think SACD sounds better because most SACD discs have been carefully mastered (and/or mixed?), but that same master put on CD would sound as good as on SACD.

A good test would be to record a SACD output in 24/96 format with a quality card, then resample it to 16/44.1, and see if it's possible to ABX both files. I think it wouldn't.


Edit : fixed quote

More SACD gibberish

Reply #5 –
Quote
Originally posted by KikeG A good test would be to record a SACD output in 24/96 format with a quality card, then resample it to 16/44.1, and see if it's possible to ABX both files. I think it wouldn't.

It would be an interesting test, where I think quality of equipment would be paramount (unlike listening for artifacts with psychoacoustic encoders).  You'd need everything in the chain capable of reproducing up to ~46KHz without significant rolloff (most headphones & many speakers, even some amp circuitry are limited to 20hz-20KHz).  Also, music that takes advantage of a wide dynamic range (one of Beethoven's symphonies, maybe?).  Finally, a "golden ear" for musical detail (an unquantifiable perception) would be helpful.


Edit : fixed quote

More SACD gibberish

Reply #6 –
Do such animals exist? I've only seen a few amps claim to go up that high, and the only tweeter I've seen rated past 40Khz were those cheap "super-tweeters" from Rat Shack. So would a person with "average" equipment be able to benefit from SACD/DVD-A's extra headroom?

More SACD gibberish

Reply #7 –
Quote
Originally posted by Cygnus X1
Do such animals exist? I've only seen a few amps claim to go up that high, and the only tweeter I've seen rated past 40Khz were those cheap "super-tweeters" from Rat Shack. So would a person with "average" equipment be able to benefit from SACD/DVD-A's extra headroom?


Not likely. The average person is a horrible eq abuser anyway.

More SACD gibberish

Reply #8 –
Quote
Originally posted by fewtch I can't read the article (not registered).

Once you register (just a few seconds) it gives you a link to the story without waiting. 
Quote
Originally posted by Cygnus X1...if the record companies would expend more effort into pressing quality CD's, there would be no reason to think about SACD.).

I would lay a couple bucks on that but I have yet to hear a SACD recording.  Also, they would have to ensure that quality is maintained in the reproduction process.
Quote
Originally posted by KikeG
I think SACD sounds better because most SACD discs have been carefully mastered (and/or mixed?), but that same master put on CD would sound as good as on SACD.

I'll bet that's exactly right - surely there'll be many more articles like this one.  But the 'breathless' descriptions of fantastic sound didn't describe anything that I don't hear most every day on (mostly remastered, I'll admit) CD.
Quote
Originally posted by KikeG
A good test would be to record a SACD output in 24/96 format with a quality card, then resample it to 16/44.1, and see if it's possible to ABX both files. I think it wouldn't.

Quote
Originally posted by fewtch
....quality of equipment would be paramount (unlike listening for artifacts with psychoacoustic encoders). You'd need everything in the chain capable of reproducing up to ~46KHz without significant rolloff.  Also, music that takes advantage of a wide dynamic range (one of Beethoven's symphonies, maybe?). Finally, a "golden ear" for musical detail (an unquantifiable perception) would be helpful..

Not sure about equipment, but wouldn't a good sound card and a good pair of headphones be good enough?  My main concern is having a foolproof way to make certain that both the CD sample and the SACD sample are transferred onto the respective discs from the exact same Wave file.  Know for sure that the exact same recording is not good enough, even if it's a duplicate CD - there's almost always (if not always) a slight to great difference.  BTW, I was shocked (no kidding) when I discovered this, though I shouldn't have been, thinking about it.  No doubt the guys on the MPC thread are right - the mfg./dist. people can't be trusted.  For music - not a problem - Barber's Adagio For Strings would probably work and there are many others.  I think it would be good to have some different stuff, i.e, 5 or 6 ballads and uptunes - maybe an oldie, classical, pop, big band, and a couple of representative modern tunes.  Not because it would be absolutely required, but because more people would have an opportunity to listen to something they're familiar with.  I may be wrong, but I think 'ears' are often overrated - sure, it helps, but it would seem like anyone who likes music enough to be on this board could probably hear enough to form a good opinion. 


Edit : fixed quotes

More SACD gibberish

Reply #9 –
Quote
Originally posted by KikeG A good test would be to record a SACD output in 24/96 format with a quality card, then resample it to 16/44.1, and see if it's possible to ABX both files. I think it wouldn't.


That wouldn't prove anything.
If we assume that there is no improvement with SACD, and the recording and downsampling are perfect, there will be no audible difference.
Now if we assume that SACD is superior, the above process reduce the signal to CD audio quality anyway, and they will sound the same again.

It would have to be tested the other way : to have a CD recorded and pressed on SACD.
If the original SACD sounds better than the copy of the CD, then SACD is worth, if there is no difference, then SACD is not worth.

Quote
Originally posted by IveyLeaguer to make certain that both the CD sample and the SACD sample are transferred onto the respective discs from the exact same Wave file


A wave file is CD, or, at best, DVD-A quality. You won't get SACD quality anyway starting from a wav file, since SACD is supposed to be an improvement over PCM, that is wav file.
You must stay in DSD format, avoiding wav, from the beginning to the end. And that's another matter for SACD mastering, since most professional gear run in PCM.
It seems "wide-DSD" gear, that is supposed to keep SACD quality, is quite still uncommon, if not at the prototype stage.

More SACD gibberish

Reply #10 –
Quote
Originally posted by Pio2001
A wave file is CD, or, at best, DVD-A quality. You won't get SACD quality anyway starting from a wav file, since SACD is supposed to be an improvement over PCM, that is wav file.
You must stay in DSD format, avoiding wav, from the beginning to the end. And that's another matter for SACD mastering, since most professional gear run in PCM.
It seems "wide-DSD" gear, that is supposed to keep SACD quality, is quite still uncommon, if not at the prototype stage.

Thanks for the explanation.

Edit : fixed quote

More SACD gibberish

Reply #11 –
is there any paper online "DSD for the dummies" ? 'cause i'd read it...

More SACD gibberish

Reply #13 –
big big thanx!!!  - now i see (at least in general). would never look at the wiki by myself

More SACD gibberish

Reply #14 –
Thanks for the info! I remember reading about DSD and if I do recall, the exact numbers for DSD are 1-bit wordlength, 2.83Mhz sampling rate (correct me if I'm wrong). Given this difference in how DSD samples a waveform, I would think trying to copy SACD audio as 24/96 PCM  for comparison with 16/44 PCM would be an inaccurate test. Do any DSD software implementations exist, or is the source proprietary?? Something tells me that like ATRAC, Sony is purposely keeping the lock on this format

More SACD gibberish

Reply #15 –
Quote
Originally posted by Cygnus X1
Something tells me that like ATRAC, Sony is purposely keeping the lock on this format


I would believe they don't release specs for security reasons. (I.E: to avoid piracy)

Anyway, DVD-A format (Meridian Lossless Packing) is closed as well.

More SACD gibberish

Reply #16 –
Quote
Originally posted by rjamorim


I would believe they don't release specs for security reasons. (I.E: to avoid piracy)

Anyway, DVD-A format (Meridian Lossless Packing) is closed as well.


Many have said that this is the major "advantage" of DVD-A and SACD, that is DRM. Of course, this gives the record companies another reason to push the new formats.

More SACD gibberish

Reply #17 –
Quote
Originally posted by Cygnus X1

Many have said that this is the major "advantage" of DVD-A and SACD, that is DRM. Of course, this gives the record companies another reason to push the new formats.

Convenient, isn't it? 

More SACD gibberish

Reply #18 –
Quote
Originally posted by Pio2001

That wouldn't prove anything.
If we assume that there is no improvement with SACD, and the recording and downsampling are perfect, there will be no audible difference.


Even if we asume there is no improvement, this test would be a good way to test it.

Quote
Now if we assume that SACD is superior, the above process reduce the signal to CD audio quality anyway, and they will sound the same again.


But the point would be to compare the recorded 24/96 clip with the 16/44.1 clip. The signal with CD audio quality would be only the 16/44.1 one, then we could compare it with the 24/96 signal. To make things better, it could be possible first to try to ABX the SACD original clip with the 24/96 version. I don't think there would be any difference, since 24/96 PCM is superior to DSD (SACD audio format). Also, you could instead directly ABX the SACD disc with the 16/44.1 recorded and downsampled file. For this comparations you would need an hardware ABX box, self made, or a commercial one, I think there are still hardware ABX comparators made and sold by QSC.

Quote
A wave file is CD, or, at best, DVD-A quality. You won't get SACD quality anyway starting from a wav file, since SACD is supposed to be an improvement over PCM, that is wav file.


A wave file can be 24/96, which is superior to DSD. DSD has a max. dynamic range of 120 dB at the audible range of frequencies, and gets quite worse at ultrasonic frequencies. At this frequencies, there are lots of ultrasonic noise, the SNR is quite bad and get worse as you go to higher frequencies. A 24 bit PCM file has a max. dynamic range of 146 dB for the whole range of frequencies sampled, if you sample at 96 KHz you will have a max. dynamic range of 146 dB for the whole bandwidth, from 0 to 48 KHz.

This is in theory, in practice it is really very difficult to achieve a dynamic range over 120 dB in any case. Still, DSD outputs lots of ultrasonic frequency noise, compared with linear PCM. DSD is no advantage to PCM, is, again, just marketing.

More SACD gibberish

Reply #19 –
Quote
Originally posted by KikeG
This is in theory, in practice it is really very difficult to achieve a dynamic range over 120 dB in any case. Still, DSD outputs lots of ultrasonic frequency noise, compared with linear PCM. DSD is no advantage to PCM, is, again, just marketing.


In practice it's damned hard to get anything over 70dB except on something like a solo instrument using an extremely quiet mic.  Mic noise, and equipment noise introduced along the signal chain during recording usually gets you to about 70dB.  This is fine though because 70dB down at a normal volume is practically dead silent.  At that point only listening through a sealed pair of headphones with the volume up would you hear much of anything, and then only during passages of silence.

Considering that 14-16dB average peak to peak is considered "good" dynamic range on modern music (the overcompressed crap people complain about is generally 6dB or less as I recall) 120dB of dynamic range is useless.  The 96dB available on a CD is in fact perfectly fine.  Most of the demos supposedly showing off the dynamic range possible have just been mixed better.  In practice, most people don't want lots of dynamic range in their everyday music because of the background noise in the environments they listen in.  Try listening to a dynamic piece of classical music while driving in anything but the most isolating of vehicles at a "normal" volume for the loud parts.  In my Jeep Cherokee I can't hear the quieter parts unless I crank it to the point where loud passages hurt my ears.

The only advantage that DVD-Audio and SACD have over CD's is that it's easier for a DVD-A or SACD player to output unfiltered sound since the inherent noise of 96Khz sampled music is well beyond human hearing.  I'm willing to bet that a piece of music recorded and processed using superb equipment, and played back on superb CD equipment, would sound just as good as the same material recorded for DVD-A or SACD.  In fact, recording at 96Khz, downsampling to 44.1 using the best available format converter, then upsampling to 96Khz before playback would probably be indistinguishable from the original 96Khz recording.

My point really is that if you can divorce the signal from the required eq to remove the noise audible during 44.1Khz playback, 96Khz gives you nothing over 44.1 except the ability to record information which nobody can percieve.  However, due to that potentially poor quality EQ needed for 44.1Khz playback, the higher bitrate recording will often sound better.

G

More SACD gibberish

Reply #20 –
Quote
Originally posted by KikeG

A wave file can be 24/96, which is superior to DSD. DSD has a max. dynamic range of 120 dB at the audible range of frequencies, and gets quite worse at ultrasonic frequencies.


If I follow you, you're just saying that 24/96 PCM is superior to DSD (SACD) because it has more dynamics.
But if we look at frequency response, DSD is superior to 24/96 PCM. So what ?

Yes, these DSD frequencies are inaudible, but the PCM dynamics too : if you put the lowest sound at 10 db, certainly the quietest sound audible in a perfectly silent room (0 db is below the noise floor of a recording studio. Acoustic laboratories can acheive it), the loudest would be 156 db, that would lead to deafness, and maybe death.

Anyway, the purpose is not to get higher numbers, but to design better and simplier digital-to-analog converters in order to improve their sound.

More SACD gibberish

Reply #21 –
Quote
Originally posted by Pio2001


If I follow you, you're just saying that 24/96 PCM is superior to DSD (SACD) because it has more dynamics. 
But if we look at frequency response, DSD is superior to 24/96 PCM. So what ?

Yes, these DSD frequencies are inaudible, but the PCM dynamics too : if you put the lowest sound at 10 db, certainly the quietest sound audible in a perfectly silent room (0 db is below the noise floor of a recording studio. Acoustic laboratories can acheive it), the loudest would be 156 db, that would lead to deafness, and maybe death.

Anyway, the purpose is not to get higher numbers, but to design better and simplier digital-to-analog converters in order to improve their sound.


DSD usable dynamics:
Code: [Select]
frequency DSD PCM24/96 without noise shaping

<10 kHz    120 dB    146 dB

 20 kHz    114 dB    146 dB

 40 kHz     90 dB    146 dB

 80 kHz     66 dB    -

100 kHz     58 dB    -

160 kHz     42 dB


When you use the frequency response up to 100 kHz you have a noise level
of -58 dB at 100 kHz.  An intermodulation of 0.1% at 100 kHz (note thta amplifiers
have higher distortion at 100 kHz than at 1 kHz) will reduce your SNR.

DSD solves nonexisting problems (for the customer) and introduces some new ones.
The main reason for introducing DSD is DRM and to force that people must buy new equipment.
--  Frank Klemm

More SACD gibberish

Reply #22 –
Quote
Originally posted by Pio2001
Anyway, the purpose is not to get higher numbers, but to design better and simplier digital-to-analog converters in order to improve their sound.


Aha. I studied DAC's and ADC's, this year in the University
The higher you go in bits, the more it increases the difficulty of not adding noise in the less significant bits, due to errors in hardware not being precise enough. (Scalar DAC's, etc.)

So yes, it can be an improvement in fidelity as in reproducing the stored signal more accuratedly, and clipping does not occur, which are some nice features.
(Except if they are only used in order to increase still more the volume of the recordings)

On the other hand, it will need two bits per sample (one for turn up/down, and the other for don't modify), or either it creates a fluctuating signal at 1Mhz, which might be the case.
And the changing dinamic range is a bit scary. A  10Khz signal, if that 2.83Mhz value is correct, is stored in 8bits!!! (a 20Khz in 6!)
If we assume that the max bit depth of the signal is 20bits, this means that the 10Khz signal will sound way 12bits lower.

In the end, the stored signal might get reproduced better, but the recorded signal might be way different than the original. Hope I only have missinformation.



Edit: Huh? Well.. seems that Frank has some better numbers than me


Edit : fixed quote

More SACD gibberish

Reply #23 –
I've been wondering about exactly the same thing.
I found no technical papers.
All i can do is assuming that this 1 MHz fluctuation acts as a dither and allows a good definition in audible frequencies.

More SACD gibberish

Reply #24 –
here comes the audio n00b with his stupid questions...

What exactly is an SACD  ???

Is DRM that same encryption crap in wma that only allows a file to be played on the computer where it was ripped??? and also windows installation...

Sorry if i bother you audio pros..