Thread: On SA-CDs from early PCM sources

Posts: 26
Page: 1 2 3 next

Post by zeus February 11, 2013 (1 of 26)
I wanted to address a comment on the Bruckner: Symphonies 4, 6-8 - Celibidache set decrying its (presumably) lowly PCM origins and deemed a separate thread more appropriate. This has taken me a while to figure out, and will probably take you as long, or maybe never if you just focus on numbers.

Let me start off by saying that early PCM recordings (even on CD) can sound great, assuming the recording engineers have done a great job. You probably have a hard time accepting this because they don't on your system. They didn't for me until now. Given that there's losses in any conversion, the ideal is the original data on a setup that can do it justice. If you don't have this (and I didn't until recently) the next best thing is conversion to a format that your system can play well, namely DSD. It should be obvious that the path for direct PCM to analogue conversion is different to that of PCM first converted to DSD, then converted to analogue. How the differences manifest themselves (and which sounds better) depends on the strengths of the individual conversions. DSD to analogue is a pretty simple process and it makes sense to leverage this as the results are likely to be good for most people. While I would always prefer to have a more state of the art recording, the music and performance can easily outweigh the losses from conversions. The only way to judge the worth of something is evaluating it on your own system. Don't discard something just because the bit rate is lower than some imaginary minimum threshold you may have.

Post by sibelius2 February 11, 2013 (2 of 26)
I own these recordings on DVD. Years ago I owned the 6th on VHS, and it was always my favorite recording of the work on any medium (although I didn't know until you pointed it out that the EMI CD of the 6th is taken from the exact same performance.) When the DVD set was released, I was hoping for a DTS surround track, but was disappointed to find that they only include a 16/48 PCM track. If the original master was only two-channel, so be it, but my preference is always for multitrack masters being presented in a multichannel format. No matter; I'm just happy to be able to watch this concert again on a format which won't wear out or get erased easily.

The DVD set cost me $30. The 16/48 track on the DVD already has at least the potential to sound just that tiniest bit better than the RBCD. What I find myself wondering is, would conversion to DSD really sound that much different to make it worth spending an additional $275?

There are 16- and 20-bit digital recordings from various labels which were originally released as CDs with Dolby Surround encoding, because that's the best the technology of the time would allow. If any of these were released as multichannel SACDs, I would pounce on them. But this Celibidache SACD set is a case where I'm simply going to have to be happy with what I already own.

Post by lennyw February 12, 2013 (3 of 26)
zeus said:

Let me start off by saying that early PCM recordings (even on CD) can sound great, assuming the recording engineers have done a great job. You probably have a hard time accepting this because they don't on your system. They didn't for me until now. Given that there's losses in any conversion...

Well, what is the conversion process here? Are we saying that these recordings were originally made at 16/44.1 or something higher ("Sibelius2" mentions that the DVD versions had a 16/48 PCM track, implying that the originals were recorded at this rate)? If they were made at 16/44.1 then I see no reason to involve oneself in the conversion process to DSD and then onto SACD, as the only perceivable advantage is that the DSD DAC in the SACD player is better than the PCM one, either in the (SA)CD player, or external DAC.

Perhaps it might be better to do a PCM=>Analogue=>DSD conversion, in order to better maintain the best possible sound?

As the above post also points out, at $275, it's an awful lot of money to shell out just to see whether they're marginally better. I would be more tempted if they were K2SACDs or SH(a)M-SACDs, as they would obviously then be of much better SQ...

I refer the honourable gentleman to a comment by PT Barnum some years ago.

Post by zeus February 12, 2013 (4 of 26)
lennyw said:

Perhaps it might be better to do a PCM=>Analogue=>DSD conversion, in order to better maintain the best possible sound?

How does this work? Maybe you haven't understood what I said.

Also, I created this thread specifically to divorce the argument from a specific example.

Post by lennyw February 12, 2013 (5 of 26)
zeus said:

It should be obvious that the path for direct PCM to analogue conversion is different to that of PCM first converted to DSD, then converted to analogue. How the differences manifest themselves (and which sounds better) depends on the strengths of the individual conversions.

I was referring to the comment above.

It could be said that the amount of time and research spent on perfecting the various conversions has been far greater for PCM=>Analogue (30 something years) and Analogue=>DSD (since the advent of DSD in iirc the 90s) than has been spent on PCM=>DSD, hence my idea that this might be the route to take.

From what I can tell, the argument you are putting forward is that one needs to hear the discs in question (be they the Bruckner or any other) before making any judgements on their merits (there is also a secondary point about the quality of the forum members' equipment which others may find quite condescending: "because they don't on your system"). This is all very well, but as said, spending, in this case, some $250 for the privilege (read: risk) is a little steep, and one has to question the merits of doing so when, in this case, the mathematical advantages don't really add up to the price.

There has been far too much opaqueness on behalf of the various record companies when it comes to describing how their SACDs were produced, so it is hardly surprising that conversations/speculations about the SQ based upon recording dates takes place. Harmonia Mundi, for example, released a whole batch of recordings on SACD when they first dipped their toes into the SACD waters, and there was little to no information in the booklets about where they came from, and whether they were any more than an (upsampled?) RBCD put onto an SACD.

As long as the record companies are clear about exacly how they've gone about producing their SACDs (like Mr. Bissie), no one can really complain.

Post by Lute February 12, 2013 (6 of 26)
lennyw said:

There has been far too much opaqueness on behalf of the various record companies when it comes to describing how their SACDs were produced, so it is hardly surprising that conversations/speculations about the SQ based upon recording dates takes place.

This is precisely the reason why it is up to the listener to decide.

Speculation is just speculation.

Trust your own ears!!

Post by Claude February 12, 2013 (7 of 26)
16/44 or 16/48 recordings can sound good, but I don't get the point justifying putting these on SACD instead of CD or DVD.

Post by Nagraboy February 12, 2013 (8 of 26)
Claude said:

16/44 or 16/48 recordings can sound good, but I don't get the point justifying putting these on SACD instead of CD or DVD.

The point is that 16/44.1 upsampled to DSD could produce a better result than straight 16/44.1 through some DACs.

My dCS player gives the option of DSD upsampling before D/A conversion when playing a RBCD or using the digital inputs.

Post by AmonRa February 12, 2013 (9 of 26)
Sample rate converters, now usually software based, can convert PCM to DSD and back. As this is pure math, it certainly is better than DA-AD with inherent analog approximations and added noise.

Certainly PCM DAC can be worse than PCM -> DSD -> DAC, but it is a low quality DAC then. It is useless to hope that this route would be better than decent CD player.

Post by Kveld-Úlfr February 12, 2013 (10 of 26)
Let me start off by saying that early PCM recordings (even on CD) can sound great, assuming the recording engineers have done a great job. You probably have a hard time accepting this because they don't on your system. because the bit rate is lower than some imaginary minimum threshold you may have.
The Queen SHM SA-CD, though they are an obvious and easy commercial move from Universal Japan by exploiting an already existing 96/24 remaster by Bob Ludwig, do sound great.

The problem is : the early albums were recorded on tape. As from "A Kind Of Magic" (1986), the band only used new digital 44.1/16 recording equipment installed in their private Montreux studio. And you can feel a strong sound limitation because of that, even though a 96/24 upsampling was done.

These SHM SA-CDs are a proof that PCM-based SA-CDs can sound great indeed... provided the source sampling is high enough.

Page: 1 2 3 next

Closed