Opinions and experts

February 14, 2022
 by Paul McGowan

In Stereophile Magazine’s opening section, As we see it, author Jim Austin has a little title at the top of the page that says, “There are as many opinions as there are experts”.

How true.

Every expert in our field has what I like to call a point of view—their opinion of the best way to get from here to there and why they chose that destination in the first place.

Take my long-held conviction that recording the analog output from a microphone preamp is best when captured with DSD. I can spout off how this conclusion was reached, from as far back as my introduction to the idea by our own Gus Skinas, Ted Smith, and Cookie Marenco, to years of experimenting, listening, testing, and verifying.

All that work has helped me form a point of view that guides my journey through the recording process.

I haven’t enough fingers to count those that argue the exact opposite.

Or, take my heretical point of view that breaks with a long-held tradition that, once captured in DSD, it is best to analog low pass filter the DSD through a DAC for mixing and mastering.

Our friends, the Murrisons (of Bit Perfect fame), have devised a phase-perfect digital low pass filter that performs the same function without ever leaving the digital domain and suffering the slight degradation of the DAC, the miles of circuitry in the analog mixing board, and a return trip through another A/D converter.

This new all-digital technology sounds to my ears remarkably better than the tried-and-true method of converting back and forth between digital and analog.

Few agree with me and most are unwilling to even entertain the idea.

“There are as many opinions as there are experts.”

Thankfully we all have a point of view that differs. From these differing viewpoints come what we call progress. It’s how we go from “that’s batshit crazy” to “that’s just how you do it”.


Subscribe to Paul's Posts

34 comments on “Opinions and experts”

  1. An expert is by definition and etymology someone who has tried and practised something at length. Professional training is the same, just under supervision with the benefit of other’s experience. So anyone can be an expert by definition as long as they’ve been in the game long enough and have a meaningful opinion.

    You completely lost me there with the technical recording stuff. As a member of the electronically unwashed, I consider it dangerous to think I have any expertise about DSD. My limited experience of a DSD DAC is that I couldn’t hear the difference, so the truth may be that DSD has merit if the rest of the system is designed for it and good enough.

    I rely on the experience and expertise of Linn. Their recording business were pioneers of DSD 20 years ago, using it extensively for around 15 years, issuing more SACD releases than anyone else. Their audio business, on the other hand, didn’t bother with DSD for years, as it is an extremely niche market. Having weighed up the sonic benefits, cost, customer demand, and continued ability to produce world-class reference recordings, Linn Records now use 24/192 PCM as their master format.

    DSD may have an edge over PCM in some regards, and Linn may well have believed that. But their experience, and hence expertise, is that commercially it did not make sense – to them. One reason is likely to be that they record a large roster of artists that have a wide audience, not just audiophiles.

    So PSA may be correct if it has a small customer base with very high-end systems willing to pay $30 for a download, and Linn or Hyperion may be correct for their larger customer base willing to pay no more than $15 or $20 for a download. If that is their experience, they may both be correct. I don’t accept there has to be one universal truth.

    1. Jesus Steven! Talk about taking the hammer and hitting the head of the Nail!
      Very well said and thank you for this post this morning.
      For the record. Linn DACs really do have a great reputation. Some of my favourite mixing engineers use them exclusively. 🙂

      1. I don’t think Linn have never made a standalone DAC, only streamer/DACs with digital volume control, with a single ethernet input. I bought a complete Linn streaming system in 2009. Purist pure streaming, in my “opinion” a logical approach. Unless you have a record player!

    2. We just have to separate the question of „what sounds better or is the better technology“ from „what doesn’t make sense commercially in the eyes of the one or other larger audio company“. Both topics have their truth but make no sense to combine fore other things than to muddy the waters.

  2. Paul. I hope you keep exploring and challenge some narratives about DSD to definitively find what really works the best in your experiences. Your love for audio will always prevail to solve so many problems or confusions.
    You have a great many here who support you while you have access to some very sophisticated, advanced and privileged equipment. 🙂

  3. However, finally the perceived sound quality mainly depends on tow primary dominant factors: the optimizations of the mics’ and instruments’ and singers’ arrangement in an acoustically optimized studio/concert hall and the quality and room acoustics in the listening room. Most unfortunately I never could detect the superiority of SACD over other high-res PCM formats – not in the best dealers’ show rooms nor at home. And many recording and mastering engineers have confirmed my own findings. Maybe Paul should list the hardware components needed to significantly reveal the differences in sound quality in everybody’s listening room?

    1. There are also opinions on where music should be recorded. I suspect the majority of the music I listen to was not recorded in a recording studio. Churches seem to prevail, and three large recording studios near me (Air Studios, Livingston Studios and Church Studios) were all originally churches. Church Studios is the one where Bob Dylan got lost looking for it and ended up having a cup tea in some old lady’s living room.

      1. And sometimes better yet if it is recorded while playing for a live audience where the performers feed off the audience’s reaction. Oh how I wish I had recordings of live performances I’ve attended at a jazz festival when the music soared in response to audience reactions. A recording of a great live performance will trump a technically great studio recording every time in my “expert opinion” garnered from listening to music all my life. LOL

      2. Chad Kassem’s Blue Heaven Studio in Salina, Kansas was originally the First Christian Church. Short version, the old stone building had a leaking roof and and the heating system was always breaking down. The church clergy and congregation decided that would be more economically sound for them to just build a new church and put the old one up for sale as is. Kassem was originally looking for cheap warehouse space for his growing disc sales business. When he went to inspect the building, he was greatly impressed by the acoustics of the sanctuary, obvious when just listening to speaking voices as walked around. He bought the building, made the needed repairs, and converted it to a recording studio. He left most of the pews in place and holds an annual (well, pre-CoViD-19) music festival there: Blues Masters at the Crossroads (I-70 and US-81) in the sanctuary. I’ve been there for two of them; great blues, roots, and gospel music.

        1. Brick is best. One of my favourite pre-Covid venues was the appropriately named Church of Sound, where gigs are held on Thursday and Friday evenings, the rest of the time devoted to prayer and communal service. Bands heard there include The Cookers, The Ezra Collective, Shabaka Hutchings and Kahil El Zabar and the Ethnic Heritage Ensemble.

    2. It’s deja vu all over again. Recall:
      Let’s compare the LP with the CD and decide which is best.

      Easy you would think.
      So many variables.

      ——and contrary-wise?—-

      Having said that, I have tracks recorded and processed audiophile-wise in native dsd also as Flac and the dsd was richer.

      1. Great, Peter. Thus you should be able to tell me which hardware I need in order to be able to detect the sonic superiority of a DSD recording. Your advice is highly appreciated. I can only report my personal experiences and that of some recording engineers who claim that even the superiority of 192/24 over 96/24 isn’t audible. And even Paul stated that a conversion from DSD to 192/24 has a nearly identical sound quality when listened to using his top end highly revealing stereo system (a virtual magnifying glass). I do not deny the technical differences or the audible influence of digital filter topologies! Did you also hear differences in sound quality after having downloaded the demo tracks of the Norwegian label 2L?

        1. Paul I wanted to find even -just one- example where dsd sounded better. Proved the concept that was under investigation

          (my gear spectral amplification Qutest dac Yamaha NS1000 speakers. Don’t laugh.)

          I like to rank quantitative improvements. I got more with Fidelizer pro which turns off my laptop background tasks. Wouldn’t be without it.

          I also set up a NAS into UltraRendu plus fancy power supply. Somewhat more flesh on the music

          More: Isolated my twin subbies from the tympanic membrane flooring via low bounce isolators. Rewarding. Got new better bass pitch.

          In general I’ll listen to a girl singer in super hirez just once. Unless she’s Jacqui Naylor

          Or a Jessica Williams jazz piano. Anything of her Redbook cd anytime, against most beautifully recorded pianists. She makes me happy. Trust me and report back and maybe we can persuade her to work for Cookie. Winner.

  4. I like the thought of straying off the beaten path. If the thought and experience to date is that recording in DSD is a better way to go, and the Murrison’s have devised a way to keep things in the DSD Digital domain for mixing, then I say go for it. Especially if it separates Octave Records from the others with a ‘closer to real’ sound than other methods. Their will always be plenty of naysayers and doubters, remember there was also a time where the sun revolved around the earth.

    Now as many have said before, the format is only the end result. Getting everything else correct before the formatting choice is probably a much bigger issue and has probably has a much bigger effect on the end product.

    I’d be the first to admit that I couldn’t walk into any place playing digital audio and instantly tell anyone exactly what format is playing. What I can say is that in the long term listening of my system I prefer DSD in playback. The rest of the opinions here are the same (black / white with no shades of grey)

    Since many DSD recordings are ‘audiophile grade’ recordings, and we don’t necessarily have the ability to listen to exactly the same material recorded and mixed in the two formats (PCM & DSD) from start to finish, how is the comparison supposed to be made by the end listener?

    Does a direct DSD recording that’s then converted to some mathematical whole number divisor for PCM distribution sound better than one that is converted to a PCM using Fractional numbers?

    Is one better off to use a Ladder DAC for PCM and a Delta Sigma DAC for DSD?

    In the end, for this individual it, comes down to this, even if it’s considered the greatest recording made on the face of this earth, if I don’t like the recorded content I’m not going to buy it. Even if someone gave a copy to me, I probably wouldn’t listen to it more than once.

    1. “Ladder DAC for PCM and a Delta Sigma DAC for DSD?”

      I don’t know what these are. This is where the complexities of technology disassociate with the reality of normal people, even with a decent budget for hifi, who just want to listen to nice music.

      Barsley would say I should get a book and become an expert, but I googled Delta Sigma DAC and after 30 seconds reading wiki my head started to hurt. If I were an expert, then I’d have another opinion. I’ve got enough already, thank you.

      1. Funny how only the things one wants to remember or know are the ones that are important 😉
        If you were in the market for a new standalone DAC you for sure would know all about them 😀

        An expert or just interested… doesn’t matter in audio. Pick what ever sounds best or measures best and every other opinion doesn’t matter.

        Audio – more than just a listen… a subject with never ending opinions, all which have various stages of validity… or not 🙂 ✌️

        1. It’s funny Chris, but the complementary position exists among the academic community, where there are people who really do understand how DSD works, but who are unable to come to terms with the concerns and interests of audiophiles. The experts bemoan the audiophiles who are forthright with their opinions while admitting no understanding of the technical issues, and the audiophiles bemoan the experts for spouting theoretical positions without actually caring a whit about perceived sound quality.

          I have a foot (more of a toehold, really) in both camps, and understand where both points of view come from. Nobody seems to have it entirely right. But at the same time, nobody gets it entirely wrong. At the end of the day, when it comes to perceived sound quality, how you set about implementing digital audio is massively more important than which technical approach you choose to take. My view is that those who believe in DSD tend to be pre-disposed toward making better implementation choices cross the board, and this is, at its root, the primary reason why their efforts tend to bear fruit. In other words, getting it right is what counts, even if your explanation for how you think you got it right does not stand up to close scrutiny.

          1. Richard, no one has explained anything about your digital DSD low-pass filter! Conventional wisdom is that DSD is “uneditable” in the digital domain. Octave converts their DSD recordings to PCM for editing, as does everyone else, I presume. Just love to hear — at a high level anyway — how you can digitally low-pass filter DSD. Maybe a return to Copper for an article about it? Thanks, David Fair

            1. David, the key to it is simply recognizing that DSD’s 1’s and 0’s encode the values +1.00000000 and -1.00000000 respectively. You encode those as 64-bit floats and put them through a low-pass filter. Our low pass filter is one that not only has the best possible amplitude flatness in the pass-band, it also has the best possible phase-flatness. The output of that filter is also in 64-bit Float form at the (very high) sample rate of the original DSD bitstream. Because the signal has been low-pass filtered we can simply decimate it down to a much lower sample rate with zero loss of data. This is the format in which we do all the fade/pan/mix work. It is not the DSD domain. As you state, you cannot edit in the DSD domain. We then re-normalize the end result and put it through an SDM to turn it back to DSD. This SDM is, sonically speaking, the only weak point in all this, and the resultant sound quality will – in principle – be degraded only to the extent that the SDM degrades it, because no SDM is perfect. That’s the core workflow, and you can invoke variations upon that, according to what you’re trying to accomplish.

              I wrote an article in Copper 138 (I think) that you should read. It may not answer the specific question you asked, but if you can absorb it and feel comfortable with the key observations it contains, it should help you 🙂

          2. That’s also how I see it. The format DSD makes its small part of the efforts, those who propagate it take, to produce great sound. The importance given to the relevance of DSD as part of the whole has a good dose of marketing in it.

  5. I think the main issue is a lack of interest in any extra effort to make things sound extra special. Most people listen to earbuds through their phone, or some type of HomePod-like device. When 99% of listening is in that vein, why would the recording industry care about whether or not to go straight to DSD or flip between analog & digital. It doesn’t seem like something people want to spend any “grey matter “ on, with the exception of the now niche audiophile group.

  6. Reading some of these comments, I feel better about myself when my eyes glaze over reading the digital technobabble. I guess that’s why I can’t get real interested in it, If it sounds good, great. And many, if not most, of my plain ol red book CDs sound great.

    1. The majority of my audiophile buddies prefer the RB-CD version of a recording which appears to sounds much “punchier” than the high-res versions. And my nieces and nephews actually prefer analog (vinyl records) over digital! Most strange. 🙂

      1. Good for them!
        Some of my best sounding CDs are HDCD….even when played on a player that doesn’t decode. And, of course, when decoded, they’re really great. Generally better even than SACDs.

  7. I vote for keeping as much of the processing as possible in the digital domain. My Hauptwerk digital pipe organ program mixes thousands of individual organ pipe stereo PCM samples real time in the digital domain. Imagine how noisy the result would be if thousands of analog samples were mixed using analog equipment. It would be unlistenable.

    Furthermore, when it comes to conversion from digital to analog, my best sounding DACs are ladder DACs.

    1. Not unlike the old, but amazing kludge that was the Mellowtron that used loops of recorded analog tape and variable speed control to change the pitch. It was apparently the sound technician’s nightmare to keep it running, but it could do things that nothing else could. Back in the ancient and arcane days, pre-digital, it was a brilliant though problematic way of making music. Now we are blessed with sophisticated computer systems and sound libraries that would have left a young Rick Wakeman (e. g., now the self proclaimed “Grumpy Old Rock Star”) drooling in a fit of techno-lust.

  8. There are situations like this that I feel like a complete failure. Math is something that is simple, logical and elegant. If I can “get it” so easily why can’t I help everyone of you “get it”? Not being able to help all of you understand the simple concepts of Pulse Density Modulation and Delta Sigma Modulators is very frustrating.

    Paul, If this pandemic ever ends ( or becomes less serious ) I would love to meet Murrisons and learn all about their phase-perfect digital low pass filter.

    1. Gifts differing. While somethings (like math and physics) come easily to you, that may not be (is) the case for others. And those other people can probably leave choking in their dust in other mental activities. And probably their interests go off in other directions as well. Leonardo da Vincis are extremely rare on the ground [Thank you, Captain Obvious]. Although we share a high degree of commonality in the basic neurological template, there are significant differences among all individuals. I suspect that this may be behind why some people love pure analog recordings and playback and find even the best digital obnoxious, while others (hi, Paul) key in on digital’s obvious strengths. We simply don’t all hear the same way.

Leave a Reply

Stop by for a tour:
Mon-Fri, 8:30am-5pm MST

4865 Sterling Dr.
Boulder, CO 80301

Join the hi-fi family

Stop by for a tour:
4865 Sterling Dr.
Boulder, CO 80301

Join the hi-fi family

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