Post by Bearcat August 2, 2003 (1 of 20)
|
|
Hey folks,
I have an SACD question. Please let me state first that it is not my intention to start a SACD v. sDVD-Audio/DSD-PCM war.
I have heard it said that DSD has an inherent flaw in that the advanced resolution deteriorates at higher frequencies. Also any EQ done has to be done in PCM, so you loose resolution when you go from DSD -> PCM -> DSD again.
Is this true?
Bearcat M. Sandor
|
|
|
Post by zeus August 2, 2003 (2 of 20)
|
|
Bearcat said:
I have heard it said that DSD has an inherent flaw in that the advanced resolution deteriorates at higher frequencies. Also any EQ done has to be done in PCM, so you loose resolution when you go from DSD -> PCM -> DSD again.
DSD (Direct Stream Digital) uses noise shaping (a term that seems to alarm some people) to push the artifacts from D/A conversion out of band and into the higher (ultrasonic) frequencies. This results in an increasingly diminished S/N with frequency ... but since the noise is decoupled from the signal, and to seemingly everyone's ears inaudible, the whole process would appear to be largely benign. Some older equipment (amplifiers and tweeters) may not tolerate this increased ultrasonic noise hence "Standard/Custom" filters are provided to set the roll-off start frequency.
As for post-processing of the recorded signal, every disc will have a varying amount of processing done and using different tools. It's best not to generalise here.
The unfortunate thing though is that topics like this generally spiral downward, with most input from people who have minimal understanding of the concepts ... some even less than me! Along the way, various papers from the 'net are trotted out to "prove" one thing or another. And those people who listen to the format on high-end gear day-in, day-out are left wondering just what the fuss is about anyway. If you really want to follow this up I suggest you consult something like Pohlmann's "Principles of Digital Audio" (McGraw-Hill).
The subject of your post is the real issue here. All I can suggest is that you do some listening for yourself and make up your own mind. With good recordings (which hopefully this site will highlight) I think you'll be suitably impressed!
|
|
|
Post by Dinko August 2, 2003 (3 of 20)
|
|
zeus said:
With good recordings (which hopefully this site will highlight) I think you'll be suitably impressed!
I think that's really what it comes down to.
DSD may be 'flawed', but when well done it sounds extraordinary, even when reduced to 16/44 for release on a regular CD.
|
|
|
Post by syrienko August 6, 2007 (4 of 20)
|
|
SACD has 4x more resolution than normal CD and it also has noise shaping techniques to move noise to unaudible frequencies. That means you get better Signal to Noise Ratio and better Dynamic Range in "hifi" frequencies 20-20000 Hz.
In my opinion it's the best audio format up to this date.
|
|
|
Post by raffells August 6, 2007 (5 of 20)
|
|
syrienko said:
SACD has 4x more resolution than normal CD and it also has noise shaping techniques to move noise to unaudible frequencies. That means you get better Signal to Noise Ratio and better Dynamic Range in "hifi" frequencies 20-20000 Hz.
In my opinion it's the best audio format up to this date.
Im not sure where you get these figures from ? please explain...
I suggest your read this webpage..
http://sound.westhost.com/cd-sacd-dvda.htm
|
|
|
Post by Karlosak August 6, 2007 (6 of 20)
|
|
I suggest your read this webpage..
http://sound.westhost.com/cd-sacd-dvda.htm
Every time I see a link to that article I scratch my head...
A lot of the points are invalid, focusing on theoretical aspects instead of real possibilities of today's hardware. Moreover, many of the hinted SACD "deficiencies" spin around noise performance, which is only one side of the coin (imho less important). What about impulse response, phase coherency, etc.?
If Mr. Öhman really thinks that his bellowed format is the best thing since sliced bread, why practically all of the listening tests many studios have made up to date, favoured DSD as pretty indistinguishable from the tape, which was often not the case with PCM based feeds?
Theory and practice are quite different things, it seems.
|
|
|
Post by Beagle August 6, 2007 (7 of 20)
|
|
Karlosak said: A lot of the points are invalid
This is a Deep Topic, and I confess my mind glazes over when I focus on the finer points. Despite my technical unease, I note that Öhman directs attention to the highest (inaudible?) ranges where he says DVD-a is superior, and dismisses the lowest (inaudible?) ranges where he confesses SACD is superior. And there are lots of 'weasel-words' such as "may..." and "possibly...".
All that having been said, Mr Ö. may still have a valid point or two, e.g. "the ultrasonic may also affect the audible sound by down mixing in the air". However he loses considerable creditability when he stoops to defending RBCD sound. This article seems to date from 2002, which now seems a long time ago in terms of higher-fidelity discs. I am left wondering: Where are all the wonderful DVD-A discs?
|
|
|
Post by raffells August 6, 2007 (8 of 20)
|
|
Karlosak said:
Every time I see a link to that article I scratch my head...
A lot of the points are invalid, focusing on theoretical aspects instead of real possibilities of today's hardware. Moreover, many of the hinted SACD "deficiencies" spin around noise performance, which is only one side of the coin (imho less important). What about impulse response, phase coherency, etc.?
If Mr. Öhman really thinks that his bellowed format is the best thing since sliced bread, why practically all of the listening tests many studios have made up to date, favoured DSD as pretty indistinguishable from the tape, which was often not the case with PCM based feeds?
Theory and practice are quite different things, it seems.
Please note this link was a response to the incorrect statement re the 4 x level? The sampling levels and frequencies are within the link..
It was not meant as the start of yet another thread re selctive discussion.
Whilst I agree with some of your comments
In general in states that sampling is higher at the important lower levels and lower at the higher levels,The exact details are not admitted by Sony/Philips but others have also run tests confirming the details.
Turning this into a discussion as to why DVDA or PCM failed is a wasted excercise.There are many well known and well discussed reasons.It may well return in the next generation super hifi.
I Suggest it be brought up as a seperate thread..
As regards impulse response,It should be obvious that above 8 k hz the PCM is superior to DSD and poorer below that but Sony decided on balance this is what they were happy with.
Are you saying that either system is better at phase? No way...though quite a few DSD items are reversed phase and nobody cares.Occasional editing in PCM also goes wrong.
|
|
|
|
|
|
I don't really know about the technicalities, but it is interesting that Blu Ray and HD-DVD use (lossless) PCM for their soundtracks. I guess this just means that PCM is rather good, and DSD is just mroe complicated to utilise at the manufacturing end. I knwo this doesn't contribute to the debate but it's an interesting (on the face of it) observation. Isn't it????
|
|
|
|
|
|
FunkyMonkey said:
I guess this just means that PCM is rather good, and DSD is just mroe complicated to utilise at the manufacturing end.
What exactly, led you to that assumption? Also what do you think was the reason for DSD in the first place (assume you know that it preceeded SACD by many years).
|
|
