Why DSD isn’t everywhere

August 19, 2022
 by Paul McGowan

As diehard advocates of DSD as the ultimate capture technology, we are in the infinitesimal minority.

We find ourselves here on the wild frontier for one reason and one reason alone. DSD sounds remarkably better than any other capture method (including analog).

So the question continues to pop up. If DSD is so darn good why wouldn’t major studios like Abbey Road use it? They claim to be state of the art.

I’ll venture forward with a couple of thoughts on the matter.

First, DSD is a pain from a workflow standpoint made worse by the one program to work with, Pyramix: an ultra-sophisticated network-capable DAW used by many of the major studios and orchestras. It is used not because it is easy (it is anything but that) but because of its incredible network capabilities. Without worry of latency or loss, hundreds of channels of high-sample rate PCM or DSD audio can be moved around a network connected only by CAT6 cable.

But moving hundreds of channels of audio data around isn’t something Abbey Road needs, or for that matter, any of the famous recording venues. And of the studios, live venues, production houses, and orchestras where Pyramix is used it is almost never DSD. (In fact, in speaking with the engineers and owners of Pyramix it turns out almost no one uses it for DSD)

Big and famous studios have to accommodate workflows and knowledge chains of visiting engineers and producers—none of whom have any experience with DSD or Pyramix. They are in the business of being the best there is within the bounds of what those who might use their services know and understand.

And, that is not DSD nor Pyramix.

That our quest for the best sound regardless of the difficulty and limitations in achieving it is what drives us in a certain direction—one not shared by the mainstream—should come as no surprise.

It’s why they call us Audiophiles.

Subscribe to Paul's Posts

55 comments on “Why DSD isn’t everywhere”

  1. I understood, Pyramix is in fact not the only one program to work with when processing DSD, but just one of the options, of which you mentioned three here already:


    I also heard about the Mytek and Merging/Horus converter, but I don’t yet understand enough for what they apply, just that some prefer the Mytek (not sure if it’s a DSD to PCM converter at all) to the Pyramix, just as you said in a previous post somewhere (and many seem to support this), the Pyramix is practically the worst solution and you just use it after Bitperfect developed the Zeephir process for conversion for you. Again others like Cookie prefer the analog conversion without going to PCM at all.

    My most relevant findings were, that it seems to make sense to start in DSD vs. PCM, no matter which way of conversion to PCM or analog you go inbetween. That every one of the necessary steps inbetween seems to be a kind of compromise (strong different opinions between DSD professionals). And to get back to todays topic: that the whole thing is at least complicated enough so that it doesn’t seem to get commonly used technology unfortunately.

    1. Well, to be clear, when working in any rates higher than DSD64 there is to my knowledge only one program that can be used for recording and editing. Pyramix. Sonoma is limited to DSD64.

      After having worked with Sonoma (which is great) I would have to say the minimum I would be willing to record in is DSD128. There the noise is out of the audio band. DSD256 is maybe better, I don’t know, but since we can use it, why not?

      1. I for sure know much less than you, just Jamie Howard commented on the Horus in this YouTube interview and from a google search it supports DXD and DSD256, but I really have no clue.

  2. I would imagine that at 74 years of age compromise is no longer a concern.
    Good for you Paul for pushing recording technology to the limit!
    Hopefully your legacy will be an important one.

  3. So DSD = USP for Octave. It’s like with digital radio, broadcasters don’t use the maximum available bandwidth or capacity they could so we end up with lower quality broadcasts. I believe it was the same with FM. I guess it’s a mass market thing. Why provide top quality when you can get away with less. So for audiophiles it’s a good thing that Octave are going for it. Problem is most of the music we like and listen to isn’t DSD. That’s what we’re stuck with and so we have to make the most of it, which is what we’ve been doing for years.

    1. See below. The problem with DSD/SACD was that it never even caught on with the vast majority of audiophiles. It is a massive error to suggest that all audiophiles like DSD (quite the opposite is the case) or that they feel they are being palmed off with an inferior PCM product.

      1. Yes, I didn’t really mean to infer PCM was inferior but I can see that’s what I did, I was just couching it in Paul’s terms. I like SACD but in my last upgrade of transport and DAC I returned to pure red book as there was such an uplift in performance from my previous set up and it encompassed the majority of the music I listen to. I have read that some believe by combining red book and SACD in one machine detracts from the performance, ie. the purity of one concept over the other, but I suppose, like just about everything, it’s a compromise.

        1. This whole DSD thing reminds me of a client about 20 years ago who invented what was arguably the best electro-plating system in the world. MARGINALLY better than anything else. The problem was that the process took many times longer than conventional systems, the machine itself was about the size of a tennis court, with a price to match.

          It was of course a complete commercial failure and deserved to be because the marginal improvement in quality came at such a huge cost.

          The good news is he invented a great product that cost $5 to make, sold for $20 and he made a fortune. Even when he had lots of money, he remained bitter and consumed about the commercial failure of his electroplating system.

      2. Sometimes I am concerned! My concern is whether or not the human auditory mechanism can discern the variety of platforms from one another.
        Stated simply, can one discriminate, beyond chance, the difference between PCM, etc etc? It would be nice for a study to be carried out to resolve this issue, rather than it being put to “trial” in the online media by people who have been “appointed” as the sages of audio!

    2. I think the thing is, DSD isn’t really an effective USP, as the understanding of the technology is difficult, the simplified claims around it don’t reflect the complicated processes and conversions involved and most don’t even have equipment to play the final format.

      The USP can only be the addiction to good sound and the quality resulting.

      If nobody perceives a disturbance factor in a gorgeous PCM recording or another DSD recording or a noticeable difference between them (or even has no chance to make a meaningful comparison), then (even I’d it factually is more) it’s not more than a marketing topic.

      Then Octave or Blue Coast are two of the few labels caring about great sound quality with whatever technology they use. Which is good enough.

  4. If you want to do DSD recording using Pyramix at Abbey Road you can.
    See https://www.simonkilnmusic.com/skmeditingandmixing
    He’s a top guy, was at Abbey Road for 25 years, I know quite a few of his recordings and he records for the Royal Opera and the Royal Ballet.

    Abbey Road has DSD capability. It is described here, by Simon Gibson, in an article about remastering their best historic classical recordings. He’s one of their top guys, he was entrusted with the Beatles mono tapes to produce the recent mono vinyl/CD remaster. The did DSD when it had its day and used SACD (DSD converted from 24/96 PCM) when he remastered this classic collection.

    He says:

    But why SACDs now, when many audiophiles see the format as something that never caught on? Can the trained ears of Gibson and his Abbey Road colleagues hear the difference between DSD Super Audio Compact Disc and standard PCM (“Red Book”) CDs?

    “There’s a whole debate there. If you make a new recording in the studio and you record it going straight to DSD, the SACD format, it just sounds fantastic. The bottom line, though, is if your engineer puts up your microphones and knows how to record something well, it doesn’t really make a lot of difference as to which format it ends up in or which format it was recorded on, as long as the engineer is making a decent recording in the first place.

    “You’re frankly not going to hear that much difference between the two, because some of the recordings are pretty vintage. In that sense, the improvement we’ve got in these versions, as opposed to previous CD versions, is in making the initial analog-to-digital conversion. And we’ve chosen to do that in the PCM domain. We did it at 96kHz/24-bits, clearly because it’s almost impossible for us to remaster and use our remastering processes in the DSD format. We can only work in the PCM format. So we produce a finished, remastered 96kHz/24-bit file, and then that gets transferred into DSD format to author the SACD. Some people will say that’s pointless, that’s not true SACD. Well, I’m not going to argue. It’s the way we do it, and we do it for valid reasons.”

    1. Well they basically say the format difference just matters secondary, which is why they use PCM for convenience reasons. It’s what Paul says. I wouldn’t say they are no audiophiles, but they are no perfectionists at least, which Paul is in this regard for the sake of the best audiophile result.

      1. The difference is that in music mixing and mastering Paul is almost a complete novice whereas the likes of Simon Gibson are world leaders and what’s the point of a recording if it can’t be produced properly?

        For example, he explains that when he remastered the famous du Pre / Boult recording of the Elgar Cello concerto, one of the most famous classical recordings ever, there are loads of things he could do to restore and improve it using tools in PCM. A straight transfer to DSD would have been far inferior.

        1. As you can imagine, of this release I have the audiophile EMI Testament AAA release and wouldn’t switch it for a digital version, but I agree to disagree for the rest.

          I also have most of the digital Warner Parlaphone digital versions of Du Pre’s EMI albums, mastered by Abbey Road studios, which are really excellent, too.

          But after all I know so far, I’m sure a DSD transfer of the analog tape (or if the original recording would have been done in DSD) and then whatever one likes to apply on that in the PCM domain as Abbey Road does (or in the analog domain as others do), would be a better result than transferring the analog tape directly to PCM and do all that.

          The important part in my understanding is, that the step from analog (live recording or tape) to digital is done in DSD, not PCM. The following processing of the DSD in PCM or analog is compromised more or less, but still better than an analog transfer to PCM and staying there would have been (all possibly on a level many wouldn’t be able to hear anyway).

          Furthermore I don’t see the need to do too many things with analog recordings in the digital domain as e.g. also the Plangent Process folks do. I think fixing things like drop outs and pitch variations in the digital domain tends to do more harm than good (although through a YouTube interview with Jamie Howard I also heard what it can positively do).

          Finally as I mentioned, the Abbey Road article to me confirms they prefer DSD, too, they just don’t see it as relevant in this case due to the vintage character of the recordings and they also choose pure PCM for other convenience reasons.

          Regarding Paul being a novice in recording business vs. long established practices of studios like Abbey Road, I think it depends on the approach. A novice with the right approach can be better in short time when he questions the right things and improves upon.

          1. All producers of classical music know that the studio engineer’s set-up is the most vital component, which is why labels and studios keep hold of their best ones for decades. A great example I’ve mentioned before is Philip Hobbs, Linn’s chief engineer since they started some 40 years ago:

            Linn did more DSD than anyone else, but reverted to PCM. It was not due to mass-market pressure as they do not stream, they are still download-only. Their sound quality is as good as anything, with brilliant production.

            At the end of the day EMI/Abbey Road (now UMG) has to remain commercially viable. 10 years ago EMI were about to close it and sell the property for conversion into apartments, until the UK government intervened with a preservation order on the building. It is now UMG’s global centre for Audio & Sound, run by Giles Martin (George Martin’s son).

            Octave is a mini studio with Gus Skinas who built a DSD system for Sony and maybe a few others, apparently recording mostly unsigned local performers. The financial commitment and profitability is entirely at the whim of one man (and his wife, I suspect).

            Paul correctly highlights that Abbey Road is a studio for hire like most others and trying to run a viable business, which is not easy, and it would be impossible to justify using something like Pyramix for DSD.

            The elephants in the room are that Octave, so far as I know, does not represent any artists that have wide appeal and require wide distribution, and we don’t know if Octave is profitable. There are only so many people who will spend $50 on a recording, and I’m not one of them.

            I have a couple of du Pre/Elgar vinyl originals (one very clean) and the EMI 17-CD du Pre box set (now ripped).

            1. So if your point is, that labels like Octave and Blue Coast are small and that they usually can’t provide world reknowed artists (in case Zuill Bailey/Octave or Kai Eckhardt/Blue Coast don’t count as two of them for you), then this doesn’t make the technology nor their ambition any worse. The big labels usually chose convenience in mass production and profit before audio perfection for their own reasons.

              1. Most of the studios near me were set up by artists or producers (Air – George Martin), Livingston (World Circuit Records), British Grove (Mark Knopfler), Metropolis (a bunch of producers), The Church Studios (Dave Stewart and Annie Lennox), besides Abbey Road and others. Don’t tell me any of these can’t make a first class recording and why they attract the world’s top recording artists.

                If perfection is a format that you can’t edit and next to no one has equipment to listen to, then it’s not perfect. Far from it.

                I’d never heard of Zuill Bailey. Where does he rank in the USA? Does he have a permanent label? My cellist of choice is Stephen Isserlis, usually put in the top box with Ma and Maisky. Always a lot of fun on stage and encourages young performers, currently Mishka Rushdie Momen. If I want to listen to the Bach Cello Suites I will listen to Isserlis on Hyperion (24/96 digital) or Fournier on Archiv (vinyl). In the only review that counts, Bailey fell into the usual trap with the cello suites (too much Bailey, too little Bach), but commented on nice sound.

                  1. I have Starker (Mercury 1965 reissue) on vinyl, but have only listed to it three or four times.

                    The question aimed at these is whether dance suites should be played more as dance. Having heard Jean-Guihen Queyras play all 6 suites with dance accompaniment provided by Anna Teresa de Keersmaeker, I like that approach, the same with the Goldberg Variations (also recorded by Kolesnikov for performance with de Keersmaeker), but also heard live in a piano-only recital. That’s the think about Bach, there’s nothing definitive.

                    Call me old school, but I mainly prefer to buy or listen to recordings based on what I’ve heard live, or the Editor’s Choice in Gramophone. As I’ve never heard Zuill Bailey play and his Bach had a very mediocre review, I’m not going to spend 2 hours listening to it.

          2. Incidentally, the Royal Ballet recently produced a balled called The Cellist based on the life of du Pre. Score by Phillip Feeney, recorded and produced by Simon Kiln, the Pyramix guy. I can’t say I was hugely impressed by the ballet, it retained my attention. The other piece on the double bill (Dances at a Gathering) in great, seen it about 3 times already. Also saw American Ballet Theater do it, that was amazing.

    2. This Stereophile article was published in 2012, when EMI launched many excellent classical music records as SACD and as FLAC 24/96. I own some of them and generally the sound is really good.
      I remember Paul saying in some place and time: “it seems that 24/96 is the sweet spot”
      Now I find that most of the classical music remastered after 2015 / 2016 is sounding better and that for example Warner is launching very good records made in 50’s, 60’s and 70’s in 24/192
      I would like to know if there is some technical reason for that or it’s just that engineers manage better and better digitalization.
      I also find that most of the recordings made after 2015 / 2016 sound better: again: why?
      Finally, I see that there are more and more small new recording companies using DSD to capture the sound as Paul is doing and really quality is clearly above the average.
      Pity for me, there is no recording from these companies with any of the great orchestras in the world.

      1. There are about 250 orchestral recordings on NativeDSD recorded in DSD, mostly DSD64. Channel Classics, Pentatone and quite a few live recordings of the London Symphony and Concertgebouw.

        I agree, 24/96 does seem to have become the norm for classical, some labels do 24/192, but for downloads why pay the extra $5?

    3. Over the past few months I have been “auditioning” integrated amps in my home since there are no shops at the level of “audiophile” where I reside. Well, there are two within 25 miles of my home, but they carry a very small sample of what is available. Nevertheless, I have gone through 5 units and each one had its good and each had its bad performances. I just gave up and settled for an integrated amp that had the “tools” I wanted and stopped since it was becoming an exhausting task-especially the ship and return process-
      One preamp/amp combo I did buy, I sold because of poor service! It got to the point that the email conversations were misleading and untruthful and finally the manufacturer admitted that their product was at fault- it took 6 months for the confession! I had to sell the gear out of disgust!

      What I did discover, much to my chagrin, that the variety of amplifiers I tried were probably very similar to each other, but it was the inconsistency of the internet that serves my router/streamer. Yesterday I was listening to a broadcast via the streamer from a station from South America (Batuta) that has great programming and great quality of sound, and I was amazed
      at the clarity of the performance. My speakers never sounded so crystal clear and In about one hour into listening, the quality of the transmission was awful resulting in poor music quality.

      So what is all this about? What I have realized is that if my major source of listening is going to be via the internet, the class of recording will matter very little.

      BTW I’m not at all unhappy since my gear and my room are in harmony and it does a great job masking the persistent bilateral tinnitus. I just wait for that moment in time when the internet behaves!

  5. Awesome Paul, now I am Googling DSD to learn about this technology. Great insight’s here and, somehow, the attitude in the industry about the complexity of the current DSD platforms, rings of past transitions….

  6. I’ve been looking for an excuse to post this. Well, if DSD isn’t conventional neither is this.
    It’s the theme to the new BBC four part series ‘Marriage’. The program is loved by the critics but most viewers seem less enamoured. After an hour of some mumbling dialogue this bursts from your tv at some volume. Initially I found it annoying but confess it’s almost becoming a grower. Shows what overexposure can do.

  7. The record companies don’t really care about sonic quality and that even includes many of the ones that claim they do and charge more. About 20 years ago my friend George Bischoff developed a simple, cheap(pennies at most) CD treatment that improved CD reproduction. It was demoed to me at a Stereophile show and I heard the improvement in seconds. George took it to big and small CD producers and the reply was they didn’t care. No one gave a damn. George wasn’t looking to get rich, just make some money to retire more comfortably so money wasn’t the problem. Perhaps if Octave records existed then since the process works on any disc.

  8. The DSD talk is always about a format. Now a days it’s about the recording side. 5ish years or more ago it was all about the sonic advantages of conversion to DSD on the playback side.

    Call DSD a superior format for analog capture. How the format is mixed and manipulated from there seems to be where all the hoopla is. Personally I don’t care what format I end up with if the recording is first rate and more importantly I like the music.

    So DSD is a differentiator for Paul and Octave Records. He / They believe it’s a technically superior way of doing things. Part of that belief is convincing others to come along for the ride.

  9. My problem with DSD is not the format itself, but the inconsistency and complexity of playback.

    Through my music server, I have tried several playback softwares with several DACs. For starters, you have to have a DAC that can play native DSD. It’s not always straightforward to understand if the DAC can do it. Second, getting playback software to deliver native DSD is not straightforward, some won’t support it. Finally, I don’t think a lot of the DACs have implemented Native DSD correctly, as most of the time PCM sounds better. I have tried downsampling DSD to various PCM sample rates with mixed results. Sometimes the downsampled DSD sounds better, sometimes not. Cafe Blue sounds way better on DSD, but the soundstage of Time Out is messed up.

    I2S seems to be a better delivery method for DSD (or PCM for that matter), but there are not many source or DAC options. To further complicate it, I2S has several connector options. Some are HDMI like, some Coax, others more Ethernet type, etc.

    Player disc transports that support DSD are getting less and less, and physical media seems to be on the decline.

    It’s a lot to go through for what is a relatively small available catalog of music. From my experience to date, it feels like what makes a recording special is mostly technique vs. format.

  10. After reading every comment this morning, I believe that every comment is close to or spot on “valid” and what I have gleaned from all of these comments is that there are so many reasons why these new audio technologies are not going to move forward quickly if at all.

    Keep in mind that it took many of us over 10 to 15 years to purchase our first CD player if we were smart and didn’t want to waste our money on the ridiculously terrible sound of the beginning of digital audio. That’s how long it took me to “buy in” since the quality at that time was horrifyingly horrid. This new and evolving digital audio technology is a different story, one that requires someone like me who is two years older than Paul to choke up the bucks to purchase the type of CD transport or Player and have yet another technology added (not replaced) in my music system. I continue to think about how many quality CDs I have in my vast collection even though I have many poorly recorded CDs as well. Then, I think about yet another method that I can bring into my system… streaming individual recordings (in addition to a quality delivery system like Qobuz) which I would consider once the high end audio industry come together with solutions on how the next generations of listeners will start to “buy in” and start listening tohigher quality music reproduction when they have their iPhones and earbuds and they can move around wherever they want while listening to music that they consider have acceptable SQ. I’m in the process of working with the down loading process and trying to figure out if I really want to continue to download digital audio and integrate these particular recordings with my current PCM streaming subscription that does not integrate well with metadata provided by ‘in my case’ Roon in addition to my current extensive CD collection On a high-quality CD player only to wind up in a technology maze. For me, this is a very difficult time to choose. Hopefully I’ll live for another 20 years and before that time I may evolve in your direction completely, then again maybe not. I have downloaded two of the Audiophile Society’s limited choices at $9.95 each (thank you David Chesky) and it has opened my eyes to the possibility of seeing some light at the end of the tunnel only to think about how I can easily homogenize ‘the old with the new’ instead of replacing the old with the new. there is no easy solution at the moment for me.

    I’ve got a decent amount of time on my hands to enjoy listening to my music and I’m not as inquisitive as I used to be nor am I willing to put in the time to integrate everything to my satisfaction.

    Work on the younger generations Paul, they may come around as they get married and earn a better living to discover what we have discovered over the course of a lifetime.

    1. It took me 1 year to buy my first CD player (1984) & it was a glorious little Toshiba (I can’t remember the model).
      It cost me AU$400 in 1984 money & it made vinyl sound noisy, flat & lifeless.
      Best AU$400 that I ever spent.
      By 1993 I had a Marantz – ‘CD80’, a 33lb beast that I was using as a transport with an
      outboard (as they were called back then) Audio Alchemy – ‘DDE v1.0’ DAC that sounded
      sweet, detailed & helped to create an amazing holographic 3D soundstage.

      You type, “…over 10 to 15 years to purchase our first CD player if we were smart & didn’t
      want to waste our money on the ridiculously terrible sound of the beginning of digital audio.”
      I have to say that I completely disagree with that assessment Neil.
      Buying that little Toshiba, back in ’84, was one of the smartest home-audio purchases that
      I ever made.

      1. I agree, it took me about 20 minutes to decide to buy a CD player. I must have bought my Sony within days. It cost £150 or £200, I forget, but CDs in those days cost £15, which is a huge amount of money, much more than $50 for an Octave SACD that never arrives.

        DSD is double or treble the price of PCM downloads, presumably because sales volumes are so low, the label has to bump up the prices to recover their costs.

        1. Initially they were AU$30 here, later going down to AU$20.
          AU$30 is/was about 15GBP.
          Now AU$8 on Discogs ($4 – $12)…average.
          Using the on-line ‘Inflation Calculator’ AU$30 in 1985 is equal to AU$95 now (2022)
          Conversely AU$8 now was AU$2.50 in 1985.

          1. In a word, bigbucks.

            My memory was going into HMV and thinking LP’s are a lot cheaper, so why did I buy this machine? Plus I was still a relatively poor student and for £15 I could buy 7 concert tickets (last minute and returns at the Festival Hall for £2 a pop).

  11. When I discovered SACD’s in 2001, it change my view of digital audio.
    I was one of those people who could not listen to CD’s for more than 45 minutes without getting a headache. Hi-rez PCM is not as annoying, however, I find that DAC’s that up sample to DSD before conversion and playback do the best job on CD’s.

  12. My silence connotes my ignorance on the subject matter. Speaking of curtains closing, in researching near-death experiences, most people have described bright lights. Some have reported beautiful music. I wonder what format it is. 🙂

    1. Like your thinking there Joseph in that last sentence, obviously a true audiophile. 🙂
      I’m guessing your next question would be, was it recorded by Octave? 😉

  13. SACD came far closer to happening than most people realize. Two of the three major labels were completely on board. Unfortunately, the replication plants capable of manufacturing CD-compatible SACDs were still under construction when Napster appeared. Label executives watched their CD sales tank in each town just as soon as broadband internet became available locally. That was the end of any new release format.

    Piracy has consequences.

    1. Interesting!

      But I think the expectable end of any discs has many reasons, all connected with the availability of online or local streaming. For high end freaks, the CD and even SACD format must have been dead as soon as it was clear that both can’t carry the data of meanwhile available sample rates and their benefit anymore.

      1. Surround was a major component of SACD and seen as its biggest selling point to consumers as opposed to better sound quality. The industry was already selling mostly catalog reissue CDs and it was hoped that SACD would lead people to buying their favorite albums for a third time. ATMOS streaming is another attempt at that.

        The immediate problem was that the managers of new artists didn’t want to wait for the SACD plants to be brought online and the stores were not interested in stocking multiple formats of the same album.

      2. jazznut, I think I qualify as a high end freak. My two principle sources of music are vinyl ( 1st ), SACD ( 2nd ) and only if I cannot find it an SACD of the music do it listen to CD’s. If I want background music a tune into a local independent radio station on my FM tuner. Streaming does not exist in my world.

        1. Ok, I understand there are some of us freaks who intentionally pass on network audio and I understand the reasons not to start with that can of worms. But otherwise it’s not typical for us to pass on a difference hires vs CD or higher res vs. SACD makes.

          And I think it must be said that the later generations of freaks rather think about discs as the older generations think about streaming. I guess the reason to pass on streaming for complexity reasons or because of a reluctance towards change dies out sooner than later.

          It’s not meant to offend, this decision is ok for now, but no perspective for major labels. We noticed when PSA gave up disc drive development, let’s see how long discs are still offered for a reasonable price 😉

          1. Being an Audio-enthusiast, as opposed to being an Audiophile, my 1,000+ Redbook/HDCD CD library, & no intention of streaming, will do me nicely 😀

Leave a Reply

© 2022 PS Audio, Inc.

linkedin facebook pinterest youtube rss twitter instagram facebook-blank rss-blank linkedin-blank pinterest youtube twitter instagram