Skip to main content

Topic: Help! Sacd Good Or Bad? (Read 110507 times) previous topic - next topic

0 Members and 1 Guest are viewing this topic.
  • bryant
  • [*][*][*][*][*]
  • Developer (Donating)
Help! Sacd Good Or Bad?
Reply #25
Quote
Huh? Plain DVD does 24/96? What's DVD-A for then?

Yup. Here are 3 sources for 24/96 music recordings on standard DVDs:

www.chesky.com
www.classicrecs.com
www.hiresmusic.com

Another advantage of these is that the digital data is available on the S/PDIF outputs of many DVD players so you can use high quality DACs (I have a 24/96 MSB DAC II).

DVD-audio goes to 24/192 (with optional lossless packing) and also has better multi-channel support. But, mostly, it has DRM and won't allow the high resolution audio out of the player in digital format.

  • Joe Bloggs
  • [*][*][*][*]
Help! Sacd Good Or Bad?
Reply #26
Any DVD-A records with DRM?

  • bryant
  • [*][*][*][*][*]
  • Developer (Donating)
Help! Sacd Good Or Bad?
Reply #27
Quote
Any DVD-A records with DRM?

All DVD-A discs are [currently] impossible to rip and force the digital outputs of the players to be downsampled. To me this is generally called DRM, but perhaps there is a specific meaning to DRM that I am not aware of.

  • Joe Bloggs
  • [*][*][*][*]
Help! Sacd Good Or Bad?
Reply #28
Oh, I mean, any DVD-As without audible watermarks?

  • n68
  • [*][*][*][*][*]
  • Banned
Help! Sacd Good Or Bad?
Reply #29
Quote
All DVD-A discs are [currently] impossible to rip and force the digital outputs of the players to be downsampled. To me this is generally called DRM, but perhaps there is a specific meaning to DRM that I am not aware of.

yup...


hence... dvd-a is just as possible to ripp.
but will.. in most cases be downmixed.. in other words..
it is possible to ripp. a dds track.

ex: the x2cd (ac3/vob/ifo/pcm) i think will downmix it.. i not shure.
there are for shure.. several apps. that >can< do it.

in theory... a multi track sequenser.. should be able to do it..
as long as you go thrue a dds. soundcard.




  • bryant
  • [*][*][*][*][*]
  • Developer (Donating)
Help! Sacd Good Or Bad?
Reply #30
Quote
Oh, I mean, any DVD-As without audible watermarks?

I don't know if DVD-A discs (or SACD discs for that matter) have audible (or inaudible) watermarks on them. They obviously claimed to want to do this but don't get a lot of favorable press about it. Does anyone know the latest?

The SACD has a sort of physical watermark to prevent playback in an unsecure device (i.e. ripping).

  • Pio2001
  • [*][*][*][*][*]
  • Global Moderator
Help! Sacd Good Or Bad?
Reply #31
Quote
Audio Note Dacs are most famous of them: http://www.audionote.co.uk/dacs/dac_index.htm

For what it's worth, Audio Note is the most expensive hifi manufacturer in the world.
Their top amplifier (100W vacuum tubes) is more than 150,000 $/€

  • Joe Bloggs
  • [*][*][*][*]
Help! Sacd Good Or Bad?
Reply #32
So would you agree with this kind of design?

What kind of output does it give? The original big- and small- staircase waveform?

e.g.

  • ChristianHJW
  • [*][*][*][*][*]
Help! Sacd Good Or Bad?
Reply #33
Sorry for stepping is so late .... DVD-A vs. SACD is one of my favourite subjects .... of course i prefer the DVD-A.

In fact, standard DVD-Video specs have a 24/96 mode and it was again David Chesky and his team who were launching the first recordings ( Sara K. amongst them, one of my favourite ) as DVD-V with 24/96 audio ( no video ) more than 5 years ago.

I remember to had the chance to play them on a hi-quality chain on one of the first players with a real 24/96 DAC, a Kenwood, but the results were lousy :-) , especially when being compared to a fully modificated HK or better CD Player !

With some positive thinking one could estimate to hear advantages in some respect, but the overall quality of the CD was much better .. thanks to all the mods done on the CD player, like battery powered power supply, meachanical damping of the drive, etc.

Since then i was waiting for some good mods based on DVD-A players in my former 'scene' , but with no results i hate to say. Many of the freaks were waiting for the first DVD-A to appear, but the stupid format wars between SONY with their crappy SACD thingies and PANASONIC leading the DVD-A side was dooming both formats to being unssuccessful in the end, at least it seems to be this way.

Today PCs are my hobby and i guess i wasnt able to hear the differences between good CD Players and well modified DVD-As, as my hearing isnt trained anymore.

About SACDs inability to playback full scale high frequency signals :

If i am not completely mistaken the same is valid for CDs also, as red book forbids 0 dB 20 KHz signals AFAIK .. else no modern CD player could fulfil red book with a normal 1 bit DAC, even at high internal clock.
Support matroska - the bestest vapourware project ! http://www.matroska.org

  • Frank Klemm
  • [*][*][*][*][*]
  • Developer
Help! Sacd Good Or Bad?
Reply #34
Quote
If i am not completely mistaken the same is valid for CDs also, as red book forbids 0 dB 20 KHz signals AFAIK .. else no modern CD player could fulfil red book with a normal 1 bit DAC, even at high internal clock.

I don't understand any word.
--  Frank Klemm

  • user
  • [*][*][*][*][*]
Help! Sacd Good Or Bad?
Reply #35
DVD-Video:


Theoretical PCM possibilities for DVD-Video:

bit/kHz

16/48 - up to 8 channels
20/48 - up to 6 channels
24/48 - up to 5 channels
16/96 - up to  4 channels
20/96 - up to  3 channels
24/96 - up to 2 channels


These are specs of DVD-V , not DVD-A !
Conclusion: DVD-A is (same with sacd) a kind of copy prohibition.
The industry could provide us with excellent stereo or multichannel music on DVD-V.....
www.High-Quality.ch.vu -- High Quality Audio Archiving Tutorials

  • ChristianHJW
  • [*][*][*][*][*]
Help! Sacd Good Or Bad?
Reply #36
Quote
I don't understand any word.

Never mind Frank,

i sometimes dont understand my own shit  !

I was convinced ( but havent read red book to investigate, sorry for that unprecise information policy  )  that CD red book does not allow to put a full scale 0 dB 20 Khz signal onto a CD when mastering them .... and vice versa that no CD player complying with red book needs to be able to play such a signal. I also estimated that most 1 bit DACs as used in cheap ( and our days even expensive ) CD players wouldnt be able to reconstruct such a signal, given the fact that they had to use a very high internal clock ( 65534 x 44100 = 2.9 GHz ? ) to do that .....

Sorry if the info about red book is not correct or if i didnt understand the basic idea of a 1 bit DAC with noise shaping . My understanding of these DACs was to have a high speed 1 bit DAC feed a capacitor with small portions of current at a high frequency ( about 1 - 2 MHz ?? ) to reconstruct teh original signal, in oder to spare the (expensive ) analog anitaliasing filtering after the DAC and also to overcome linearity problems of the DACs in mass production ?

If this is wrong, how is it ever possible that a 1 bit DAC working at 1 - 2 Mhz can output a full 16 bit signal at 20 Khz ?
Support matroska - the bestest vapourware project ! http://www.matroska.org

  • Joe Bloggs
  • [*][*][*][*]
Help! Sacd Good Or Bad?
Reply #37
OK, so non-oversampling DACs use analog filtering exclusively...

I took a look at this review of an AudioNote DAC 1.1...
http://www.tnt-audio.com/sorgenti/audionot...ote11kit_e.html

And they say it was able to produce this analog filtering:

frequency relative level (dB)
1kHz 0
10kHz +0.25
20kHz -2
30kHz -10
40kHz -17
50kHz -22
100kHz -39
200kHz -57
500kHz -80
1MHz -99

And the phase remains 'essentially linear' up to 20kHz. 

Sounds good... the analog filtering should not lead to pre-ringing, unlike digital...

Can somebody tell me once and for all whether pre-ringing is actually an artifact or it is just the natural product of the calculations?

E.g. if you made a 2-way IIR brickwall filter (not 1-way, that would of course give no pre-ringing but wouldn't have correct phase response) that processes the whole track offline, would it still have pre-ringing?

I mean it's funny how with phase correct filtering you inevitably get a symmetrical output from an impulse input... which by definition has added sound content BEFORE the impulse...

  • ChristianHJW
  • [*][*][*][*][*]
Help! Sacd Good Or Bad?
Reply #38
.. arent most modern DACs using a combination of digital and analog filtering to overcome antialiasing probs ?

All the 'oversampling' stuff coming up in the early 90is had nothing to do with what oversampling really is ( during recording ), but the attempt to replace (expensive ) analog filtering with ( cheaper ) digital filters, thus making the analog filters more or less redundant .....
Support matroska - the bestest vapourware project ! http://www.matroska.org

  • Kim_C
  • [*][*]
Help! Sacd Good Or Bad?
Reply #39
Quote
For what it's worth, Audio Note is the most expensive hifi manufacturer in the world.
Their top amplifier (100W vacuum tubes) is more than 150,000 $/€


Yes, their stuff is very expensive, but they have also very reasonably priced  lower range available. Mainly Zero and One series (which has non oversampling 1.1 DAC).

Quote
Sounds good... the analog filtering should not lead to pre-ringing, unlike digital...


Here is more 0-oversampling DAC's which use analog filtering if you are interested:

Zanden Audio Model 5000
http://www.zandenaudio.com/english/con-5000.html

Morgan Audio Deva Cd-player
http://www.morgan-audio.co.uk/deva.htm

DAC100 HIBARI (which is PCM56 based dac by Kondo)
http://www.audionote.co.jp/digital/index.htm

Other related information:

For Do-It-Yourself people, articles about building 0-oversampling DAC. Very interesting and worth a read.
http://www.tnt-audio.com/clinica/solidstate.html

Audio Asylum discussion thread "Brickwall filter vs No filter or Analog filters"
http://www.audioasylum.com/scripts/t.pl?f=...digital&m=42318

Audio Asylum discussion thread "Building DAC, Filter of Filterless?"
http://www.audioasylum.com/scripts/t.pl?f=...digital&m=24610


Personally i'm interested on 47 Laboratory DAC's. They don't oversample at all and they have not any filtering. Despite absence of digital or analog filtering, reviews of them are positive. Model 4705 Progression is part of my "dream system" which i'm going to buy sometime in far, far future. 

47 Laboratory Model 4705 Progression
http://www.sakurasystems.com/products/47dac.html

47 Laboratory Model 4715 Shigaraki DAC
http://www.sakurasystems.com/products/shigadac.html

Quote
.. arent most modern DACs using a combination of digital and analog filtering to overcome antialiasing probs ?

All the 'oversampling' stuff coming up in the early 90is had nothing to do with what oversampling really is ( during recording ), but the attempt to replace (expensive ) analog filtering with ( cheaper ) digital filters, thus making the analog filters more or less redundant .....


Yes, this is correct.

  • Joe Bloggs
  • [*][*][*][*]
Help! Sacd Good Or Bad?
Reply #40
What troubles me is that I don't even know what's wrong with oversampling and digital filtering sonically to give non-oversampling an edge in sound quality (if indeed this is true!~  ) I suspect it has something to do with the pre- and post- ringing introduced by digital filters. Do contribute to the new thread I started about this

  • Pio2001
  • [*][*][*][*][*]
  • Global Moderator
Help! Sacd Good Or Bad?
Reply #41
Quote
Can somebody tell me once and for all whether pre-ringing is actually an artifact or it is just the natural product of the calculations?


It is a natural product of the calculation.
It is the calculation itself that is an "artifact"  A brickwall filter is an artificial DSP.
Natural (analog) filters are not phase linear...

Quote
E.g. if you made a 2-way IIR brickwall filter (not 1-way, that would of course give no pre-ringing but wouldn't have correct phase response) that processes the whole track offline, would it still have pre-ringing?


yes

  • Frank Klemm
  • [*][*][*][*][*]
  • Developer
Help! Sacd Good Or Bad?
Reply #42
Quote
Huh? Plain DVD does 24/96? What's DVD-A for then?

DVD-D can only uncompresed LPCM up to 6 MBit/s.

96 kHz x 24 bit x 2        (4,6 MB/s)
48 kHz x 24 bit x 5        (5,8 MB/s)
48 kHz x 16 bit x 5.1    (4,6 MB/s)


DVD-A can also compressed LPCM up to 10.2 MBits/s.

192 kHz x 16 bit x 5.1  (4...6 MBit/s)
96 kHz x 24 bit x 5.1    (4...6 MBit/s)
--  Frank Klemm

  • KikeG
  • [*][*][*][*][*]
  • Developer
Help! Sacd Good Or Bad?
Reply #43
About non-oversampling filters, I have to say that any DAC is supposed to filter as much as possible over half the sampling rate, otherwise you get lots of aliasing above this frequency. This is ultrasonic sound not audible by ear, but if at high levels, may intermodulate (because some equipment is more non-linear above audible range) and cause products that fall into audible range.

So, proper DAC MUST filter over half the sampling frequency (over 22.050 KHz for CD), otherwise the DAC is lacking the necessary reconstruction filter.

About superiority of analog filters over digital filters, etc, any filter response, including analog, can be realized using digital filters, in a more easy and less problematic manner. That means that you can use digital filters at a DAC that show no pre-ringing the same way that an analog filter woud do, but with the consecuence that it won't be linear phase anymore, as analog filters.

And about problems with brickwall filters on CD, time smearing, etc, I think this is not problematic at all for the reasons I explained here: http://www.audio-illumination.org/forums/i...=1&t=2957&st=13

  • Kim_C
  • [*][*]
Help! Sacd Good Or Bad?
Reply #44
Quote
About non-oversampling filters, I have to say that any DAC is supposed to filter as much as possible over half the sampling rate, otherwise you get lots of aliasing above this frequency. This is ultrasonic sound not audible by ear, but if at high levels, may intermodulate (because some equipment is more non-linear above audible range) and cause products that fall into audible range.

So, proper DAC MUST filter over half the sampling frequency (over 22.050 KHz for CD), otherwise the DAC is lacking the necessary reconstruction filter.


Ryohei Kusunoki says this on the interview:
Quote
KUSUNOKI SAN:
There is a slight possibility that a digital filter-less DAC's intrinsic quantizing noise, existing beyond the audible range, can badly influence the sound. In my experiments, however, the noise is effectively eliminated with a first-order low-pass filter.

The original Compact Disc format was based on the assumption that a "human can hear up to 20kHz" in essence. So why bother oversampling and cutting off the "inaudible sounds" generated by oversampling? I hope my readers to be skeptical on this methodological inconsistency.
So, what is the sampling frequency in essence? Sampling the sound with 44.1kHz means that the CD can "differentiate the sound up to 25 microseconds." Raising the sampling frequency to 96kHz, for example, should not be considered as an extended frequency range up to 48kHz; it should be regarded as an "enhanced precision - over time domain," instead.

TNT-AUDIO:
There are studies showing the human ear sensitivity is extended to frequencies higher than 20kHz, at least in dynamic situations. This seems to contradict your theory. Our ears, anyway, tell us that you cannot be far from being right. What do you think about this?

KUSUNOKI SAN:
My theory is based on the assumption that our audible range is limited to 20kHz, as I have explained in the Audio Amigo interview. Therefore, if we can hear the sound beyond 20kHz and be influenced by it, this would be inconsistent to my theory.

TNT-AUDIO:
On the other side there is an audible loss in high frequencies, a few dBs at around 20KHz. According to you, is the drop in high frequency response an advantage or a shortcoming?
If you consider this a limiting factor, have you ever tried to solve it?

KUSUNOKI SAN:
Certainly there is such loss when you measure the frequency response. However, the loss can be detected only by those people who are very sensitive to high-frequencies, and most listeners cannot differentiate the attenuation of sound. The loss is not favorable, but I think it is not that important.


Quote
About superiority of analog filters over digital filters, etc, any filter response, including analog, can be realized using digital filters, in a more easy and less problematic manner. That means that you can use digital filters at a DAC that show no pre-ringing the same way that an analog filter woud do, but with the consecuence that it won't be linear phase anymore, as analog filters.


Quote
KUSUNOKI SAN:
I have been paying attention to the digital filters these days. I have described my DAC design as a "non-oversampling" in the MJ articles, and the appellation got out of control thereafter. Among those DAC components, the digital filters should be more important -- that's what I think at this moment.


So, yes digital filters are more practical but CD-player and DAC manufacturers should concentrate on making better quality digital filters.


I again recommend TNT-Audio article about building a 0-oversampling DAC anyone who's interested about the concept. It discusses the theory and technical charactistics of 0-oversampling DAC and mentions Kusonukis DAC filterless concept as "purist":
http://www.tnt-audio.com/clinica/convertus1_e.html

Index of related TNT-audio articles:
http://www.tnt-audio.com/clinica/solidstate.html

  • 2Bdecided
  • [*][*][*][*][*]
  • Developer
Help! Sacd Good Or Bad?
Reply #45
I've heard David Chesky's 6 channel 24/96 stuff (that he intended to release on DVD-A) on demo at the AES, and it is amazing. Much better than crappy 5.1 which was never designed for music anyway.


As Frank said, DVD-A allows for lossless compression of the audio data, which enables 6 full bandwidth channels at 24-bit 96kHz. The DVD-V format did not include this lossless compression, so is restricted as described by user, because the maximum data rate allowed on a DVD is about 9-10Mbps (someone will jump in now with the exact figure!). The lossless compression is required to bring the 24/96 6-channel data down below this limit. search: Meridian Lossless Packing, or MLP


Cheers,
David.

P.S. maybe we could say that an oversampling DAC gives "tonal purity" while one without a filter gives " transient accuracy". All the theory of digital sampling goes out the window without a reconstructing filter, but since the ear low passes (typically) somewhere below 20kHz, you could say that is the reconstruction filter.

There is still theory to say that a filter should be included (because of intermodulation in the audio system causing the ultra-sonics to get back into the audio band) - but in addition to the quotes in this thread, I've heard a mastering engineer say that nyquist low pass filtering is a BAD THING when converting higher sample rates to 44.1kHz for CD release.

  • KikeG
  • [*][*][*][*][*]
  • Developer
Help! Sacd Good Or Bad?
Reply #46
In reply to Kim_C:

It is true that human ear cannot hear much above 20 KHz, and this is specially true for dynamic situations. Most people can't hear anything over 19 KHz with steady signals, and maybe over 17-18 KHz on dynamic signals.

This doesn't mean that if you let pass lots of high level ultrasonic garbage (that was not in the original analog signal at all!), due to the lack of a reconstruction filter, there is not going to be problems over the audible range. Most amplifiers have a passband that goes up to 100 KHz. If you let happen that ultrasonic garbage to pass from 22 KHz to 100 KHz, there is a good probability that with such high level ultrasonic signals, the usually greater nonlinearity of the amp at such high frequencies causes intermodulation products that fall into the audible range.

Analog reconstruction filters that filter properly this ultrasonics are very difficult to build, and its properties (stability, phase response, passband ripple) are suboptimal in comparison, that's why they were dropped many years ago. On the other hand, using a simpler low order filter that has good characteristics, it is not possible to filter much of this ultrasonic garbage.

Again, I don't see any problem with today's cd oversampiling digital brickwall filters, in theory. In practice, if some phenomena not taken into account happens, well, show me some proper ABX or double blind tests that prove there is really a problem.

In my opinion, Kusunoki is another of those "illuminate" people who sell expensive esoteric solutions to solve unexisting problems. Same happens with things such as upsampling external DACs, SACD, even things like green pens, cd demagnetizers, cable holders, silver cables, etc.

  • Kim_C
  • [*][*]
Help! Sacd Good Or Bad?
Reply #47
Quote
This doesn't mean that if you let pass lots of high level ultrasonic garbage, due to the lack of a reconstruction filter, there is not going to be problems over the audible range. Most amplifiers have a passband that goes up to 100 KHz. If you let happen that ultrasonic garbage to pass from 22 KHz to 100 KHz, there is a good probability that with such high level ultrasonic signals, the usually greater nonlinearity of the amp at such high frequencies causes intermodulation products that fall into the audible range.

I am aware of this. In this case amplifier needs to be able to cope with this.

Quote
Again, I don't see any problem with today's cd oversampiling digital brickwall filters, in theory. In practice, if some phenomena not taken into account happens, well, show me some proper ABX or double blind tests that prove there is really a problem.


No disagreement here, IMHO modern cd-technology is very very good and many ways "good enough" that there is not a need for DVD/SACD audio on customer side, except for the multichannel audio.

Quote
In my opinion, Kusunoki is another of those "illuminate" people who sell expensive esoteric solutions to solve unexisting problems. Same happens with things such as upsampling external DACs, SACD, even things like green pens, cd demagnetizers, cable holders, silver cables, etc.


Might be, might be not. It's another point of view for sure... 

OTOH, i'm not a engineer and don't know much about these things.. i'm just interested on different technologies and how well they work.

  • Kim_C
  • [*][*]
Help! Sacd Good Or Bad?
Reply #48
Quote
P.S. maybe we could say that an oversampling DAC gives "tonal purity" while one without a filter gives " transient accuracy".


Yes, here is comment from Audio Asylum thread "brickwall filter vs no digital or analog filters" which is on same lines:

Quote
1. Brickwall filter
Implemented as a digital filter. Usually used with synchronous oversampling. You cut off all information from the the digital signal after c. half the (over)sampling frequency. Pitfalls: ringing - with a simple single pulse the pulse will start to sound before it is played, because the filter "sees the signal before it is played". Most implementations also ring after the pulse is played. Some call this smearing in the time domain. One could say that this implementation is more correct in the frequency domain.

2. Filterless DAC
You don't cut off (or even attenuate) the alias images predicted by Nyquist theorem by filtering after c. half the sampling frequency. You get alias images right after half the sampling frequency and multiples of that. With CDs these mean ultrasonic noise in the analog signal. If implemented properly, you don't have ringing like with digital brickwall filters, but you do get ultrasonic noise. One could say this implementation is more correct in the time domain.

So it's either frequency or time artifacts. Pick your poison.

  • Joe Bloggs
  • [*][*][*][*]
Help! Sacd Good Or Bad?
Reply #49
Quote
P.S. maybe we could say that an oversampling DAC gives "tonal purity" while one without a filter gives " transient accuracy".

Where can I find the :insane: smiley from old Hydrogenaudio?