DSD 256 file support

Enjoying my Bartok HDAC. I have several DSD256 files from NativeDSD. I would like to add my voice to those requesting that dCS add DSD256 file support to the totally awesome RingDAC platform. The momentum is there at this point. Thank you to the geniuses at dCS for giving me the best sound I have ever heard from my large digital library (except for those pesky DSD256 files).


It’s known I support this feature request. But I will ask you this, as I ask myself: because you can still play the files through Bartok—I use Roon for downsampling to my Vivaldi stack—do you have any sense you are missing something from the music?

I convert my DSD256 files to DXD (352.8kHz PCM) using DSD Master. That’s essentially the same software nativeDSD use to create DXD versions of their DSD files.

Quote from nativeDSD website:

These FLAC files, available in 24 bit 96 kHz and 24 bit 192 kHz FLAC as well as CD Quality 16 bit, 44.1 kHz FLAC are created from DSD and DXD editions of the albums by BitPerfect, using their unique “Zephiir” conversion process which produces exceptionally high quality FLAC files

I find the result outstanding and don’t miss DSD256 at all.

1 Like

I’m very slowly rolling into DSD audio now that I have a DAC - the Bartok - that can play them. And learning about DSD audio. I have 5 albums so far; I tend not to listen to classical much, but am very much into acoustic guitar. Solo piano to some degree too.

I noticed last night that Native DSD has a new pure/native DSD256 album. In looking at what DACs play DSD256 (in their spreadsheet) there are far fewer DACs that play 256 - most top out at 128. I’m curious why that is? Is it the processing power required to parse a 256 file, or something else?

I’m learning about DSD in general and also about how different devices process/convert DSD files. What is a good resource for this; I’m not technically inclined, would love a more consumer oriented resource(s).

Interesting. Rudi; I have a couple of questions:

  1. Why not just purchase the files in DXD? Maybe you normally do.

  2. What are you using for playback software? The folks at DSDMaster acknowledge that converting DSD to PCM is not a lossless process. I have a fair number of 256 files (and even some 512), that are all playable on my MSB, but not on the Vivaldi, I let Roon do the downsampling to 128. I do not know if this downsampling is lossless, but I would be surprised if it was. I have not found an answer yet at the Roon community forum. Have you compared downsampled 128 vs. DXD? Does your playback software perform downsampling?

1 Like

I am a purest I guess. I don’t like upsampling and downsampling. Something is likely lost in the process. Personally, I prefer the sound of DSD at all native resolutions. DSD64 SACD’s which I rip sound great. I have listened to music over the years in numerous formats. When I say to myself “gee, that’s sounds awesome…” and glance at the display it usually says “DSD”. That’s my benchmark. It’s annoying that dCS says “of course the RingDAC can handle DSD
256” but then say they have not gotten around to it in the same statement. I would rather listen to any resolution on dCS than the equivalent or higher on a lesser chip based DAC. But for the money we invest in dCS, they should “get around” to supporting DSD256 in my opinion. I love my Bartok. It’s my first dCS product. I am thinking of a Vivaldi (dac, upsampler and clock) for my main listening room in the future.

1 Like

I’m a big fan of DSD, but with the Vivaldi, I have become more sensitive to the file provenance. Many titles at Sound Liaison, for example, are recorded in DXD, and I buy them that way. Same at NativeDSD and Blue Coast; I try to go for wherever format and sample rate in which the music was recorded. Blue Coast is releasing a new live-studio album by Fiona Joy Hawkins and Rebecca Daniels next week. It’s recorded in 256DSD, and that’s what I’ll buy. Hopefully, dCS will close this gap. It does not prevent me from enjoying the music, but it does seem that one of the top tier digital playback companies ought to have this capability. I understand the company’s position here and at this point, I’ve come to expect no DSD256 until Vivaldi’s successor arrives.

1 Like

Yes I think that you may well be correct. A new flagship model will need to have certain unambiguous “boasts” to differentiate it from its predecessor. The same consideration may apply in respect of any headphone capability.

In any case there are more requests for improved features than can be provided from dCS’ resources. DSD256 is certainly now slightly more prevalent than it was 24 months ago. But it still is of relevance to only a tiny fraction of available repertoire and is thus still of relatively minority interest.

Like you I like to purchase music in format most closely to the original. This is not as simple as it sounds. “Recorded in DSD256” means just that. It does not mean the audio has not been processed on its way to the final audio file. For example Merging’s Pyramix platform converts DSD to DXD for processing and re-modulates to DSD at the very end. So a “Recorded in DSD256” file may have spent part of its birth process as DXD. Does it matter? Merging claim that DXD is transparent to DSD and that it does not. Others such as Andreas Koch may disagree. A short summary of the conundrum can be found here.
Having said that, I usually purchase both DSD256 and DSD128 files at nativeDSD. You can get an additional sample rate for a “few Dollars more”. The DXD I usually produce myself using DSDMaster. I buy the DSD256 because I like to have the “Original” (with all caveats attached) and other DACs that I use can do DSD512 (the Neukomm and Chord Hugo 2). I have no idea how exactly nativeDSD convert the sample rate of DSD, but for sure the process is not lossless and involves demodulation and remodulation.
I have asked the question about converting DSD sample rates on several forums, but never got a real answer. I suspect there is a demodulation and remodulation somewhere, as no-one wants to be explicit about how they do it. Interestingly the Roon signal path used to explicitly show a DSD to PCM conversion and a subsequent re-modulation. It only shows one step now. Whether the math behind changed or whether this is just cosmetics, I don’t know.
Regarding what sounds best, I am afraid there is no simple answer. Adding the mysterious DSD sample rate conversion to the up sampling possibilities in dCS equipment provides almost infinite possibilities (even more for the poor Vivaldi owners). To me, @acousticsguru is the golden ear on this subject. Maybe he can weigh in with some of his experiences.
I try to follow the “minimal loss concept” that I have picked up from Hansruedi Neukomm (the father of the CDA126S). He is of the view that any component or processing has the potential of losing quality. You cannot improve the quality of the original, you can only decrease it. So his approach is to ensure that minimal losses occur. So I go for minimum number of conversions. Any processing must have a reason and yield a benefit that is greater than the losses created by it.


Agreed. I oversimplified with respect to buying the original provenance. But it’s still my goal. Sometimes I also buy more than one version as well “just to check.” I can often perceive a difference, but rarely anything that would detract from my enjoyment had I not known the difference. In this realm, I try to remain mindful of Teddy Roosevelt’s admonition that “comparison is the thief of joy.” My ultimate goal here is to enjoy music, not equipment or specs.

1 Like