When the forum discusses specific requests for functionality then I’m afraid that neither I nor James can comment on whether functionality of any type (not just DSD256 support in this case) may or may not be added to products until such time as an announcement of any specific functionality is officially made by dCS.
Nor can we comment upon the processes implemented by other companies in their creation of (in this case) DSD media.
I hope that the forum can understand that this has to be the case.
Both PCM and DSD are capable of quite stunning levels of replay quality (and without necessarily ‘chasing numbers’) when the content is well mixed and sympathetically post processed - the analogy that I normally use is of the format simply describing the shape and size of a bucket which is used to carry the contents and the contents can be anywhere between vinegar or fine wine (or in my case grapefruit juice and Diet Coke).
I am consistently stunned by the quality of playback from my (soon to be Apex-ed) Rossini, including when streaming.
At present I am listening to a devine recording of Stamitz’s Clarinet Concert in B Sharp via stream, courtesy of Radio Swiss Classic : )
I also agree that we should not be “chasing numbers,” but the catalogue of DSD256 is growing, and other leading manufacturers support it, and speaking personally, it’s on the top of my feature request list.
In my early years (my late twenties and early thirties) I was in charge of the vocal studio where I was working at the time and we mostly did voice recordings for games.
I was always amazed at the qualities of the human voice when it is well recorded and isolated - I would often spend hours just listening through the daily sessions in the studio in the evenings just enjoying the sound of the recordings that we’d made that day and listening to the characteristics of how those people would deliver their scripts (for example - we had Tim Curry in and I remember just flipping through his recordings - absolutely lovely) - and that was all recorded at 44.1kHz 24bit.
I originally viewed possible limitations of supporting DSD256, not so much with the actually processing capabilities of the platform itself - especially considering dCS having doubled the Ring DAC frequency with just the 2.0 software upgrade - but with the I/O of the various dCS components.
For example, with Dual-AES being the only interconnect between the Vivaldi Upsampler and DAC, any AES rate above 192kS/s would put it outside the AES3-2008 specifications. However, with the 2019 revision to the standard, I believe even 384kS/s is a supported rate. So, 705.6kS/s to carry DSD256 in DoP could be quite possible.
So, I wouldn’t put DSD256 beyond the realm of possible.
@keiserrg From an engineering standpoint, there is no reason DSD should be better than PCM (at comparable sample rates). Quite the contrary. With DSD you need to go to higher sample rates to get rid of ultrasonic noise than with PCM. Simply using PCM would suffice completely from an SQ point of view.
Whether someone will record in PCM or DSD is purely a subjective and marketing decision. People tend to think DSD sounds better / smoother (add your favorite adjective). “Pure DSD” sounds high quality for sure. But there is no mathematical reason for that.
Did you mean on the Ring DAC? It’s running at 6MHz, but it’s 5-bit. So, no, it’s not DSD128, and it used to run at 3MHz/5-bit before it was doubled in frequency with the 2.0 firmware upgrade.
Presumably in order to support DSD256 that rate has to double again to 12MHz, albeit, clearly it’s not a 1:1 relationship between the incoming DSD rate to Ring DAC noise-shaper/requantizer rate since the latter is multibit.
Simplistically, whether that’s possible in addition to the I/O issue (in the case of the Vivaldi components) is really the speculative question.
If I’ve understood that correctly, Anup, does that I/O issue mean that the Vivaldi components might be the only ones in the range that may NOT be able to do it?
I’m not chasing the high-rate DSD stuff, by the way — I’m interested mostly in the awkward situation that could put dCS into. A flagship that couldn’t play material the entry-level gear could would be a tricky thing indeed.
Of note is as of Pyramix 12.0, it supports DSD256; it used to be you had to record in DSD128 and upscale to DSD256.
There is no substitute for recording at the highest possible resolution. Even if the final delivery format is at a different rate, having a pristine recording for posterity should be the goal. Pyramix has been pushing the boundaries since the beginning. DSD has been one of the options since the earliest days of SACD and now the adoption of RAVENNA has made DSD256 a reality. DXD was jointly developed by Merging and Philips so we understand the technology. The best Sample Rate Converter in the business ensures that your recording will sound magnificent at whatever rate you want.
Ha! That’s not what I meant. (Which I think you realise )
I mean that whoever manages the dCS product range would have it hard if Lina, Bartok and Rossini all could, but Vivaldi couldn’t. I think we can already see dCS aren’t spec chasers by nature, but if only some of the range could do it that would make things tricky.
Yes, I mean the ~ 6MHZ speed. No I don’t mean 1bit obviously, but you cannot exchange bits for speed in hardware, the clock runs at the speed it runs. My understanding is the platform always run at that speed internally anyway - the v2.0 firmware just allowed the front end to run at that same speed rather than 1/2 of it.
For example, Wadia’s DACs maxed out at 96/24 when 192/24 recordings were becoming available.
No matter how good 96/24 sounded on their DACs, basic specsmanship meant their product was not considered by some because it wouldn’t do 192/24.
The fact that a $60 Amazon DAC can natively do DSD256 where dCS cannot has I’m sure already discounted their products in the minds of some, given many of dCS competitors (MSB Reference, Wadax Reference, most anything with an ESS ES9038PRO or ES9039PRO) can handle native DSD256.
Not expecting an audible improvement between DSD256 and DSD128 or between DSD and PCM just because of the format per-se.
In my case the inquiry about DSD256 is mostly about practicality. As I have more than a few DSD256 albums which I have collected over the last couple of years, it would have been nice that I could play those with the Rossini, without the need of decimation or similar processing.
I imagine I am not the only one in this situation.
Furthermore, in my case as I don’t use Roon or Mosaic, the only option is to make an offline conversion, which is mildly annoying.
I am not aware about the implications of supporting DSD256 on the dCS Ring APEX platform, of course if the platform hardware is not compatible or a huge development work would be necessary (would appreciate dCS could comment on this - i.e. how big of an endeavour is this), I would not suggest DSD256 support as a top priority, unless - as others have commented - purely as a marketing asset.
P.S.: I was listening to this yesterday (DSD128), and what a glorious recording of a glorious performance it is!
My point really being that Phil’s comment in this thread means that the request has been noticed by dCS. As their policy is not to comment unless and until the feature is implemented unfortunately there is now not a lot further that this thread can add.
Thanks for mentioning the HDTT transfer . I have to say that Sir Clifford was a pianist I greatly enjoy. May I recommend his Mozart piano concerto 27 ( UK Decca). A performance that he blocked from release during his lifetime.
Well, I totally understand that dCS cannot commit on the Feature Requests threads on whether a certain feature will or will not be added, or about the timing of such potential implementation.
A technical assessment of the challenges and implications would be interesting to know nonetheless.
Anyway, this matter is far from being a deal breaker on the Rossini or Vivaldi, I am certainly not holding my breath on it. I am not a believer in the superiority of latest and greatest resolutions / formats.
In fact most of my best sounding material is 16/44, often originally recorded with analogue equipment (the good-old Mercury Living Presence, RCA Living Stereo, Three Blind Mice - to name a few - still have an immanence that very few modern recordings, regardless of format, achieve).
In my opinion, there are two things that happens when an artist and studio choose to record in DSD and other high res formats:
(a) the first is obvious: the initial information capture is higher;
(b) the second more subtle: the entire quality of the process is higher. By choosing HD, both artist and studio are saying that they care about the process and the output
I think we all know some amazing Redbook recordings that are outstanding. But when artists and studios put the same effort in w DSD and DXD, the results are, again IMHO, much better. And also, IMHO, better than what dCS’s impressive upsampling approaches can achieve with Redbook. There is no substitute for more input material.
I continue to argue that commercial titles in pure DSD 256 are very, very limited. Apart from tape transfers and a handful of titles from boutique labels, commercial DSD titles are transcoded to DXD PCM for editing and then back to DSD for resale. IMHO this situation will remain indefinitely, as it is just too expensive to work with DSD in a pure state or convert back and forth from analog for editing.
Personally I couldn’t survive on a diet of recordings from Octave, Just Listen, Eudora and tape transfers. Titles from these labels sound glorious even when sampled from their websites - so there is something in the engineering process!
)
