Post by mdt May 17, 2004 (11 of 26)
|
|
randy said:
Try as I might, I have not been able to find specifications for any ADC or DAC that credibly claims more than 20 bits of resolution. I would be appreciative of anyone who could provide a link to the spec sheet for such a beast, as I trust it would make for a fascinating read.
The converter mentioned is the Nexus 28 bit true-match converter developed by Dr.H.Jahne. You can get the technical descripton by writing to Pauler Acoustics at info@pauleracoustics.de
Their web-site is linked to the one of Stockfisch Records, www.stockfisch-records.de
|
|
|
Post by Johnno May 17, 2004 (12 of 26)
|
|
mdt said:
About the translating of DSD to PCM. This is a mathematical consideration of the issue looking at the amount of information. But in audio not only the amount of information is important but also the quality.So the absence of certain distortions may be more important than the technical values of dynamic range or frequency range.Certain parts of the audio signal are of special importance to reproduce it faithfully, a system superior in this area may sonically outperform another one with technical specs seeming superior at first glance.
I suggest reading Ed Meitner on this, he has done much work developing PCM and DSD converters and favours the latter for the kind of reasons mentioned above.
That reminds me of another point mentioned in that article I read in "Hi-Fi News". If we assume that DSD is an extension of Bitstream, I was intrigued to read that the sampling frequency used with the latter was 11.2896MHz -- four times that used with DSD. If I remember correctly, the article stated that Bitstream involved 3rd order noise shaping whereas DSD involves 7th order -- hence the reason for the high level of ultrasonc noise with the latter. I presume that current chips couldn't handle that degree of noise shaping at the higher sampling frequency.
|
|
|
Post by Johnno May 17, 2004 (13 of 26)
|
|
randy said:
Try as I might, I have not been able to find specifications for any ADC or DAC that credibly claims more than 20 bits of resolution. I would be appreciative of anyone who could provide a link to the spec sheet for such a beast, as I trust it would make for a fascinating read.
What about a company like Chandos, which claims 24-bit for all their latest recordings -- not to mention all the highest quantized DVD-As?
|
|
|
Post by Dan Popp May 18, 2004 (14 of 26)
|
|
Johnno said:
What about a company like Chandos, which claims 24-bit for all their latest recordings -- not to mention all the highest quantized DVD-As?
Johnno,
The problem has been that, yes, we can put 24 bits of information down; but the last 2 bits are latched to bit 22. IOW, if bit 22's value is "1", then the value of bits 23 and 24 are also, "1." So you have 24 bits of data taking up bandwidth and storage space, but only 22-bits' worth of resolution.
|
|
|
Post by mdt May 18, 2004 (15 of 26)
|
|
Dan Popp said:
Johnno,
The problem has been that, yes, we can put 24 bits of information down; but the last 2 bits are latched to bit 22. IOW, if bit 22's value is "1", then the value of bits 23 and 24 are also, "1." So you have 24 bits of data taking up bandwidth and storage space, but only 22-bits' worth of resolution.
What's the idea behind that ? Makes no sense to me at all.
Is that still the case today ?
|
|
|
Post by Johnno May 19, 2004 (16 of 26)
|
|
I agree. Where is the integrity in that? Are you implying that companies who say their recordings are 24-bit are being blatantly dishonest?
|
|
|
Post by randy May 19, 2004 (17 of 26)
|
|
mdt said:
What's the idea behind that ? Makes no sense to me at all.
Is that still the case today ?
Johnno said:
I agree. Where is the integrity in that? Are you implying that companies who say their recordings are 24-bit are being blatantly dishonest?
Perhaps it is a bit dishonest, but such is the nature of marketing. When it comes to storage of PCM, the conventional choices are between 20 and 24 bits per sample. Anyone using a 22 bit ADC is not going to want to reduce the samples to 20 bits, so they are going to store 24 bits. Since everyone else is claiming their recordings are 24 bit based on storage size, anyone admitting to using 22 bit ADCs would be at a disadvantage. (Plus if you digitally reduce one of the recording's tracks by more than 6 dB, then in a way you are utilizing all 24 bits, never mind that there are no true 24 bit DACs either.)
There is a close parallel in the computer industry. For nearly every measure of storage 1 kilobyte represents 2**10 or 1024 bytes. Most hard disk drive manufacturers, on the other hand, claim storage capacity based on 1 kilobyte representing 1000 bytes. With today's drive sizes this can add up: A drive claimed to have 200 Gigabytes of capacity really has only about 188 Gigabytes. The few manufacturers that are honest about this might specify "200 billion bytes, or 188 Gigabytes," then try to explain the difference on the packaging in a way that the average customer can understand.
The claims of the manufacturer mdt references above notwithstanding, I know of no true 24 bit ADCs in common use, and Dan Popp's comments seem to corroborate this.
|
|
|
Post by Dan Popp May 19, 2004 (18 of 26)
|
|
mdt said:
What's the idea behind that ? Makes no sense to me at all.
Is that still the case today ?
m,
I don't know whether that technical issue has been solved or not, and I don't know how I'd find out with any degree of certainty. Most of the people claiming 24-bit performance probably don't know of this problem, to answer Johnno's question. Let's not go jumping to conclusions about people's motives; they're the victims of flawed technology and sloppy marketing as much as you are.
I brought this up only to illustrate my point, which was (in case the reader has forgotten) that we may be "throwing bits" at the Resolution issue and not getting much, if any return for the extra cost and bandwidth. There are certainly less expensive ways of increasing the level of quality than continually increasing system requirements for bandwidth. That's why I like DSD.
One of the main reasons Sony wanted to develop DSD was for archiving. It gave them a "future-proof", high-resolution, robust solution. Those are all good reasons to get rid of PCM altogether, IMO.
|
|
|
Post by mdt May 19, 2004 (19 of 26)
|
|
Dan Popp said :
One of the main reasons Sony wanted to develop DSD was for archiving. It gave them a "future-proof", high-resolution, robust solution.
That's what i've read as well. Also i have read that at the outset of every A/D conversion the digital data occurs as continuous bitstream like in DSD, in PCM this is then divided in to packages of defined lenght (quantization).
So the idea behind DSD as archiving format is that it can very easily be converted in to any desired format that may appear.
However i have just read today that Universal is currently converting all of their (analog) masters to 24/192 PCM to save them for the future.
At least they are keeping the originals.
|
|
|
Post by Johnno May 19, 2004 (20 of 26)
|
|
Thank you, gentlemen, for those comments. Most enlightening. I presume, Dan, that when you mention "future proofing" with regard to Sony's developement of DSD, you are referring to anti-piracy issues? If you are, I agree. May that long continue.
|
|
