I went to a short seminar by Dr Waldrop at an audio show a few years ago. He made it pretty clear that does not care that much about soundstage and imaging. And to me, that is where DSD is very noticeably superior.
I agree with Paul. DSD sounds so much more natural. I think it is better to compare PCM against DSD. Although PCM in higher resolutions sounds smooth, the 'lightness' of DSD, the fragile details are so missed in PCM and all the effort that goes in getting to sound PCM nice (like on real high-end DAC's) are like lipstick on a pig. Thank you guys for advocating for DSD!
Which is pretty much all of us, unless the either of them has faults and adds stuff that wasn't there in the recording to begin with. I can pretty much guarantee that Octave Records productions will be mixed and mastered using PCM. And it will be fine. DSD creates more problems than it solves in terms of music production. Well designed PCM implementations are already totally transparent to our ears, even at a low price point.
@@ericbrammer3845 I suspect DSD limits you a lot in what kind of DAWs you can use, because all the usual ones (ProTools, Cubase, Logic etc) are PCM. The same goes for the used audio plugins, which are all PCM. If there is a DAW that supports DSD and that allows using 3rd party plugins, then I suppose the recorded DSD will be internally converted back-and-forth between PCM and DSD, at which point you could ask yourself why use DSD in the first place? In my humble opinion there's no real benefit in recording in DSD format instead of PCM. DSD however can be useful in for example streaming music over the internet (it's good at dynamically optimizing quality for the available data bandwidth, without switching between static sampling-rates and such).
Without going through the technicalities, which truthfully I am not knowledgeable about.. I do hear the difference between DSD recordings and regular recordings, and I can hear more details on DSD recordings. However, a bad recording will still be a bad recording even if it's in DSD.
digital conversion chips available now , are mostly 24 bits , delta to sigma conversions technique , the latest conversion technique , the best of all it is in a stereo format.
One thing people miss is the EXPERIENCE in the mind that DSD can give you, versus other format is exceptional - There is an assumption that better must always sound 'better' but then dont clarify what 'better' actually entails ! DSD for me does sound better but its the absorption INTO the music that my mind does is different from flac for example - This is subtle - not everyone listens like this
High res PCM and DSD capture the same information, mathematically speaking. One does not capture more than the other, one does not encode more information than the other. The only difference between them is how they capture that information. Both DSD and PCM's means of capturing that information produce artifacts (also provable with math), which each format deals with differently. Which means someone that says "DSD is better" needs to show mathematically that the way DSD vs PCM captures information and deals with artifacting produces effects on the audible spectrum - which to date no one has been able to do, because neither format does. Otherwise we're all just talking about subjective experiences, which aren't reliable data and can't prove anything. If your brain thinks DSD is supposed to sound better, your brain will make DSD sound better - whether or not DSD actually sounds better in a measurable sense. My personal theory is that DSD has gotten a reputation for being "better" precisely because it *is* hard to work with in the studio - editing DSD is so hard people usually don't bother to manipulate it much post-recording, so DSD recordings tend to be pretty "flat" and less-manipulated (unless people recorded in PCM and then converted to DSD after mixing and recording, which often happens, in which case it doesn't matter) It's not a better format in any technical way at all, it's just so hard to work with it discourages recording and mastering engineers from monkeying with it :D
In Paul's case, I believe him that his DSD recordings really do sound better than his PCM recordings. If you listened, I suspect that you would hear the difference, too. I suspect that the difference is not subtle. If we go with the assumption that what Paul said really is true, then... ...the question is why? Why are Paul's DSD recordings better sounding than his PCM recordings? Until Paul provides a list of all of his associated gear, the answer to that question remains a mystery. There could be any number of reasons why Paul's DSD sounds better than his PCM. It could be that he uses different transports, or different interconnects, or different analog to digital converters. Paul said @3:43 that he records both DSD and PCM at the exact same time. Does Paul have two of every box needed for that purpose? -- Two of the exact same microphones? -- Two of the exact same digital to analog converters? -- Two of the exact same interconnects? -- Two of the exact same power conditioners? Etc. And has Paul switched them around, to see if, perhaps, two supposedly identical pieces of gear really are identical in their sound quality? Did Paul purchase two of everything at the same time? Or did Paul purchase one set, and later a second set, leaving the possibility that changes were made by the manufacturers, resulting in different sound quality? When Paul plays back the DSD and the PCM, is he using the exact same equipment to compare the two? As it pertains to the above: Who knows? Paul knows that all of this is key, and yet he completely ignores the topic. Until Paul lays out the answers to the above, I am not buying his assertion that DSD is better than PCM. Maybe it is? But until I have affirmative confirmation, by way of having all of the above answers, I will not accept Paul's generalized statement. For all we know, Paul might not even realize that he has a mismatch in his gear when he does the comparisons.
@@NoEgg4u Spot on. The difference in conversion alone can be enough to prefer one over the other. Using the same exact signal chain and identical converters I haven’t been able to tell the difference, and I’m a full time mastering engineer monitoring in a fully built mastering room. The math suggests the same as both sampling methods should capture everything in the audible spectrum (assuming that the sampling method is correctly set up for the material). Coincidentally, all of the mastering engineers I know (including the absolute best of the best in the industry like Bob Ludwig and Bob Katz) are more than happy with PCM and don’t give a damn about DSD.
@@NoEgg4u Oh I think his recordings *do* sound good - I own the Zuill Bailey release from Octave Records, it's lovely. But that's because they're well-recorded, and/or the chain for DSD recording or PCM recording is different - not because DSD is inherently better than high-res PCM.
@@davegbro Thanks, I was going to post a comment asking if Paul thinks Claude Shannon's mathematical proofs about digital audio were wrong but apparently they are still valid.
For some reason, my ears are not as discriminating.. I've had many opportunities to listen to both DSD and PCM (including Octave Record releases) and can hardly tell the difference.. It is possible that my stereo is not resolving enough but regardless both sounds AMAZING to me. It is always good to have CHOICES !
Solid video, Paul! Dr. Waldrep is a fountain of knowledge, and he has convincingly established that most people cannot reliably distinguish between hires formats in blind tests (although I don't remember if DSD was part of that study). But my biggest concern with him is that it sometimes feels like he's got a horse in the race since he also has a label. On my Grado SR325is headphones, with my CHORD Mojo using the UAPP app, DSD sounds somewhat cleaner, more transparent. But it's by no means worth the headache to find free or affordably priced files to download. I've also been trying to build an SACD collection but short of spending thousands of dollars, it'll remain limited in size and it'll never be comprised of albums I actually chose, just the best titles currently available out there. (If I was into classical music, it'd be easy. But for jazz and classic rock, foggetaboutit.) So yeah, DSD is a pain in the a$$ at the user end as well. Cheers from Montreal.
There's a very simple rule behind blind tests: the bigger the difference in the two formats/standards being tested, the easier it will be to detect that difference. The more you understand how the digital world works, not just in audio, but in video and even photo as well, the more the above rule becomes a universal law. High-res audio needs to be orders of magnitude higher in sample rate and bit rate compared to what we've had as standard in the past few decades (aka CD PCM - 16bit/44.1 kHz audio) for the difference to just overwhelm anyone undergoing a blind test. What we have with blind tests in the present is quasi-high-res or lite-high-res vs standard res audio. If we were to compare something like DSD 1024 or its PCM equivalent versus CD audio, the difference would be astonishing to anyone undergoing a blind test - on a sound system capable of fully resolving the high-res version, of course. Gonna be a while though, before that can happen.
Blind testing can only determine what people are conscious of, which turns out to be 1% of what they experience. There are techniques to make them more aware of experience, and thus more conscious, but they require a psychological saavy that the jihad-objectivist types who champion AB testing, happen to be the worst personalities in the world for having any competence in understanding.
my guess that in theory DSD's advantage is it's PDM signal can be directly fit into the last stage of a class-D power amplifier, cutting out the decoding and encoding DAC processes thus "direct" in DSD ..But in reality there're very few of such special amplifier devices existing as product. most people use some kind of DAC that takes DSD as input and outputs analog signal then regenerates PDM signal by a common class-D amp at downstream, much like a traditional PCM audio system.. resulting almost no advantage on the consumer market..
Thanks Paul,, this session of Ask Paul got me to start follow up reading. about the best explanation I came across was in an article from back in Aug 2015 by Benjamin Zwickel. I can easily see your point, it's listening to the playback in the original recorded format and using a format that works best.
It depends on what DAC you use. For example the Chord Dave sounds better with PCM whilst various others sound better with DSD. If your DAC suits DSD then it will sound better than DSD & vice versa
What I miss is some blind tests and some scientific proof. Other than that this is to me just an opinion and not a fact at all. DSD died long time ago and for good reasons.
For me when I listen to DSD I get a better sense of space in the recording. I hear more of the trail off after a note is released giving the music a more realistic feel. It’s like the space between the notes is more active but in a natural way even on my modest setup.
I actually agree with your assessment. The only PCM albums that I have that sound just as good or even a tad bit better than DSD are the DXD albums I own. Namely, the Master of Puppets album by Metallica
Seems to me the only value to DSD is that you could directly drive a class D amplifier with it. Seems silly to use the format for mastering as you'll always be converting it to another format for editing and mixing. What is the point? Sounds like misinformation being propagated. If you went from DSD to PCM and then back to DSD and claim it sounds great, then the PCM format must be okay.
These listening tests indicate that as a rule, no significant differences could be heard between DSD and high-resolution PCM (24-bit / 176.4 kHz) even with the best equipment, under optimal listening conditions, and with test subjects who had varied listening experience and various ways of focusing on what they hear. Consequently, it could be proposed that neither of these systems has a scientific basis for claiming audible superiority over the other. This reality should put a halt to the disputation being carried on by the various PR departments concerned. DVD-Audio versus SACD Perceptual Discrimination of Digital Audio Coding Formats Listening Comparison Test between DSD and High Resolution PCM (24-bit / 176.4 kHz) by Dominik Blech and Min-Chi Yang Erich-Thienhaus-Institute (Tonmeisterinstitut), University of Music Detmold, Germany Presented at the 116th Convention of the Audio Engineering Society
Paul, I have heard big differences with .flac files, when, for example, a laptop was used as the transport to a high-end DAC vs. the Innuos Statement used as a transport, to the same DAC, and both used the exact same .flac files. And the laptop sounded really good, until I heard the Innuos Statement. Why am I pointing this out? The Innuos Statement transport virtually eliminates jitter and noise, and the result is amazing. Would this be the case with DSD? Meaning, is the DSD technology such, that it does not suffer from the jitter that is associated with PCM? Also, which transport were you using, when you asked your artists to choose the better sounding recording? The transport is critical in achieving a bit-perfect data stream. I hope your DSD vs. PCM challenge was not skewed by different quality transports. And then there is the analog to digital converter. Which one did you use to create the DSD files and the PCM files? Was it at the caliber of the Lynx Hilo? Were the same interconnects used? The same power conditioning used? The same DAC used for playback, etc? Please provide details so that your viewers can know that your test was fair, as well as it just being helpful to us to know which equipment you used. All Absolute Sound, Stereophile, etc reviews include all of the gear in their reviews. Please do the same. Please consider making a video, walking us through your process for both DSD and PCM. Thank you.
It's simple: Record in DSD (best for preserving the music with what the sigma delta ADC outputs). Play in PCM (best for distribution and device support, still with scalable audio fidelity).
To me PCM sounds great but I can tell it still sounds like a "record" a bit edgy and ringing a bit on highs especially when things get busy while DSD are much more dynamic and natural sonically still the closest thing to Analog recording. DSD really showcase its advantages in genres that have a lot of instruments, big band, jazz fusion and classical when things gets to crescendo DSD remains clear and detailed while PCM would struggle at highs and smear and ringing harshness starts to occur, brighter masters on PCM really exacerbate its weakness. Too bad SACD specifically are so hard to rip due to its proprietary nature and high level copy protection makes them hard to archive for people who just wants to listen to their SACD collection elsewhere.
I agree with Paul on this one - DSD does sound better IMHO. Hey Paul- "Leslie" from the UK is probably a male, not a female. In the UK, the feminine version is most often spelled "Lesley."
@@YnotNomis do you know how to actually convert and what frequency to use?I use 2x rate and it’s beyond amazing 🤩 I use 32x176.4kghz and then convert to 128mhz and then back to 256mhz and the difference is night and day apart from 128 which is Greta but I love the wide sound stage of 256mhz and I’m not using more space either
The whole system needs to be good enough to reveal the difference. DSD isn't something I miss if it's not presented AB alongside PCM. I don't have an incredible system by any means (mostly NAD class AB ampifiers, late 80s era JBL monitors), and even I can hear the difference between DSD and PCM on that stuff; to me, it's especially noticeable on very fast higher frequencies, like busy hi-hat work. The difference is likely considerably more profound on better playback gear.
Mark Waldrep is correct. In an effort to maintain revenue Sony developed dSD many years ago when their licensing stream for the CD was expiring. PCM has improved greatly over recent years and the PCM problems that DSD was supposed to address no longer exist. DSD has serious technical problems. Such as very high noise levels just above audible range requiring sophisticated filters and the need to convert to analog or PCM to edit and mix. When a PCM file is played on a native DSD single-bit converter, the single-bit DAC chip has to convert the PCM to DSD in real-time. This is one of the major reasons people claim DSD sounds better than PCM, when in fact, it is just that the chip in most modern single-bit DACs do a poor job of decoding PCM. A good DSD and a good PCM recording can be audibly indistinguishable.
Really nice channel and content, thanks for doing this! True-fully you can tell the difference! If you have the ears ;-) I am hearing mostly classic and sacd stereo even makes a huge difference! I just started with DSD and SACD and its a complete different story! After this I cant hear any more cd's or even mp3s. They sound like garbage to me. And DSD is really hard to be handled by the music editors, thats why editors came up with DXD etc. Thanks for bringing people to DSD and SACD. 👏
I have never heard DSD content before. Is it really that marvellous? Don't you need super high end equipment to even begin to hear the differences between that and PCM? Does it *really* sound "live" ?
DSD sounds fantastic compared to mainstream PCM (a la CD quality). You don't need much in terms of a DAC and AMP to hear its qualities - relatively entry level ones will suffice in fact ($100 dac/$100 amp), but your speakers should preferably have at least 50 kHz of sound resolving capabilities - I mean in terms of total amount of vibrations they're capable of in the 20Hz-20kHz range, not in the sense of having tweeters capable of producing 50 kHz soundwaves.
Just focus on the music. Philips' and Sony's engineers knew what they were doing. CD format is all you'll ever need. The rest is just extending the product/technology lifecycle to make a couple of extra bucks.
Paul, maybe with all the engineering prowess at your disposal you might focus on coming up with a way to do the things like editing DSD directly that you currently cannot do. Greater achievements have been made. I'm not convinced that it's physically impossible. It might take massive amounts of computing power at least at first but hey, in the words of your boy, "C'mon man!."
Single-stage, 1-bit sigma-delta converters are in principle imperfectible. We prove this fact. The reason, simply stated, is that, when properly dithered, they are in constant overload. Prevention of overload allows only partial dithering to be performed. The consequence is that distortion, limit cycles, instability, and noise modulation can never be totally avoided. We demonstrate these effects, and using coherent averaging techniques, are able to display the consequent profusion of nonlinear artifacts which are usually hidden in the noise floor. Recording, editing, storage, or conversion systems using single-stage, 1-bit sigma-delta modulators, are thus inimical to audio of the highest quality. In contrast, multi-bit sigma-delta converters, which output linear PCM code, are in principle infinitely perfectible. (Here, multi-bit refers to at least two bits in the converter.) They can be properly dithered so as to guarantee the absence of all distortion, limit cycles, and noise modulation. The audio industry is misguided if it adopts 1-bit sigma-delta conversion as the basis for any high-quality processing, archiving, or distribution format to replace multi-bit, linear PCM. timbreluces.com/assets/sacd.pdf
I'm sure DSD is better than PCM because of the much higher sampling rate. The trouble is DSD recordings have a very limited catalog. I can get a CD for just about any recording.
Paul, I can hear the differences And I just got into this hobby doing the pandemic I can hear the difference between PCM, MQA and DSD and I agree with the a 1000% DSD is the better format hands down With pcm and MQA. I hear it but I do have a musical Fidelity Integrated and vanderstine's model CE signature so I have a very resolving system.
I listen mostly to rock and it's difficult to find a cd where the cymbals sound reasonably natural. Mostly it sounds like a compressed saucepan lid. But I think a lot has to do with the recording and mixing of a cd, cymbals need a lot of headroom. Listen to the album Heritage by Opeth. Great cymbal sound, you can even hear it at low resolution on Spotify. And the music is good, I don't just listen to the cymbals ha,ha😆
@Lloyd Stout Yeah, some remasters sounds better in the bass. But then they turn up the recording levels so much there is no dynamic headroom left for the sounds that need it.... It becomes a big "moosh" a.k.a. Deep Purple/ Made in Japan. I don't really get it.... why? Music without dynamics is boring. It's not remasters we need, it's remixes😉
What about more-modern PCM DACs and ADCs that internally convert to/from pulse-density modulation? This would make the differences largely irrelevant, correct?
How thick are the walls in the new studio going to be? 12 inch with treatment? We already know your going to home run every AC outlet to the panel, and Lighting? What are you doing in that direction? And the CDs I have from Octave Records sound pretty darn good, so am looking forward to the new Studio. Wish the CDs where 10 bucks less, will you do that for Black Friday? I'd like another one before Christmas.
ABX will tell if they are different, not really which is better. Seeing as the would come from different masterings or via DIY file conversion...I am not sure you would learn anything new since thwy will probably be different anyway.
Mastering is a lost art and the digitized mixing process has turned sound into an artificial soup that is difficult to qualify. From autotune to quantized drums, sampled acoustic sounds, synth based orchestral augmentation etc... I mean what the hell is that supposed to sound like? Everything is processed so much as to sound perfect. And nothing in the real world is perfect. Our brain knows the difference and some smart folks do too. Unfortunately, the biz is fixated on AI in just about everything and it's sucking the soul of our world. So thanks Paul, thanks for standing up for quality, the real and tangible contribution of your vision.
Much more to read, but Paul's post was enough to prompt me to track down and purchase Mark Waldrep's book. I jumped right to the chapter on DSD... and so far (many pages downstream) MW has done nothing but explain how DSD is *inferior* to high bit rate PCM. ???????
What about conversions? Does PCM converted to DSD sound better than PCM? Do people tell the difference between DSD recorded and DSD converted in blind tests? Does DSD converted to PCM sound provably worse than DSD?
I don't think anyone can really hear the difference between hires PCM (e.g. 24/96) and DSD. Its also difficult to find a sample which is recorded, mixed and mastered all in the same native formats using the same source (performance). Most DSD recordings (SACD) are transfers from PCM or had a PCM interim step for e.g. mixing.
I don't have a choice how most of the music I buy was originally recorded. Most hardware supports both so I just take what I can get. I don't spend time worrying about something I can't change.
Is it possible that it's not necessarily PCM that's worse but rather all the implementations? There's still a lot of work being done on digital filters even though the Red Book standard has been around since the 80s!
94 khz/24 bit instead of 44.1 Khz/16 bit sounds better to me, deeper and more airy. Maybe because of issues with the low-pass filters, I don't know. Even when I generate high frequency sine waves (10Khz and up, and I'm 43 but can still hear fine up to 18Khz, tested) I hear some lower frequency artifacts from my DAC. Also higher bit depth increases dynamic and lowers quantization noise. Anyway, CD sound was really good compared to older analog formats, but there's always a sensation that there is something wrong with the sound, like something missing, unsatisfying, stressful, hard to explain. Maybe human hearing can only oscillate up to 20Khz, but perhaps the brain perceives changes in a time scale smaller than a 44,1 khz grid, who knows for sure? I prefer DSD myself. It's the closest to analog we can get, but with the benefits of digital. Pulse density, in my opinion mimics in a way, the grain in a magnetic tape, but with a precision and depth that only digital can achieve. It sounds so musical and satisfying to my ears. Also the method for filtering DSD is a bit different from PCM.
Adjusting the equalizer and mixer correctly in the audio studio is much more important than the audio format itself. It is not the actual audio format that determines whether there is good audio quality or not. if you are going to use DSD. then you also need to record in DSD. You can not convert from PCM to DSD to believe you are getting DSD audio quality. Sound quality between DSD and PCM is not as big as people think. Is not philips bitstream a bit similar to the way DSD works?
I'm recently downloading sacd iso albums, and just discovered how good they sound, more detailed warmer and with more texture. Plus, there's a vast catalog of popular artists in sacd. And I'm just using the chord mojo dac/amp and HD 660s headphones. 👍
to be able to read the dsd and decode it you need all the necessary equipment, that is to say speakers and amps which go up to more than 40khz in the characteristics in order to be able to hear the harmonics, without forgetting the dac and the source👍👍
You won't be getting into a pissing-match not at your ages' not with your,,prostates lol. yes but what you're saying with DSD you could do with eqing with reverb in PCM and make it sound live so what's the difference? What is the difference between PCM and dsd noise floor and DB ratings? Both do the same dynamic range I don't know about noise floor
I did a few years ago, but I had to do it "in the box" which meant re-converting the DSD to hi-res PCM. The test was really to compare SACD to CD. I recorded the analog output of the SACD player into hi-res PCM, downsampled it to CD quality (which filtered out all that ultrasonic noise) then burned it to CD. Null testing the CD version against the original capture left some residual signal above 20 kHz (down -80 dB or more) and noise in the audible range (down -100 dB). There was no audible difference (i.e. the nulled result was perfectly silent).
But we're do bits come in? I seen somewhere that most of DSD is still only 24bits. I know that bitrates were covered in a video not too long ago. It's just I'm a big believer in 32 bit original recordings. Then converting it down when nessisery.
Isn’t Abbey Road Studios going through an entire rebuild? Who has an idea of the new recording equipment that they’re going to be using? They may have already finished.
Did you watch the video? He said 176kHz PCM, which really, is high compared to what most other studios in the world probably still use, a la 24bit/96kHz.
@@edfort5704 I'm aware most other studio's run at 24bit/96kHz. I run one myself. However, most times I've seen DSD compared to PCM, it's usually against the 384 bit rate. BTW, It's okay to provide an answer without being a smartass.
Roons DSD upsampling is poor tbh , try HQ player that’s some beefy DSD upsampling , makes some stuff a lil soft i use it. mostly for poor recordings as it can take glare out , although on well mastered stuff it can be very nice but depends on filters and PCM on well recorded stuff can be amazing too
@@RossKyle95 Thanks. I am a newbie to HQ Player and always thought that it's very complicated to settle on one profile. Is there a generally limed setting in HQP that most people prefer that I can set up and try HQP? Also, is the difference between HQP and Roon DSD upsampling as noticeable in speaker setups as in headphones? Thanks
DSD is a far superior format to PCM and anyone that wants to test that assertion out for themselves doesn't have to spend $50,000 on ultra-high-end audio equipment (from any company) to do it - or anything even close to that. Simply go out and buy a new Sony X800M2 4K Blu-ray player for $325 and a new Sony STR-AN1000 receiver for $900. Plug the 4K player into the receiver with an HDMI cable that can pass the DSD audio bits directly over to the DAC in the receiver. Get a copy of an album on CD that also has a version available in SACD. First, pop in the CD and listen. Then pop in the SACD and listen. Problem solved. Once you do that, you'll understand what DSD's value proposition is and what it brings to the table. Once you've done that for yourself, you'll finally get why so many people absolutely LOVE what DSD (and the SACD format that contains it on disc) offers the world of high-end audio. It's a huge step up in both sound quality and the overall listening experience. Try it yourself and let your ears be the judge. THEN you can decide if you want to go out and spend $50,0000 on an ultra-high-end audio setup for your home.
The "advantages" of Hi-Res, DSD and all that are only in your mind. It's called confirmation bias. You hear a difference, because you paid for it. I listen to new music on Qobuz in Hi Res. If I like it, I buy it on CD (oh no, only 16 bit/44.1k), or vinyl. Trust me, the CD sounds the same as Hi-Res. Maybe even a little better, because it's got a nice booklet 😉.
When using my DAC player I hear no difference with playing in 24bit FLAC recording or having everything played via DSD. DSD the DAC gets quite warm that I prefer to leave the DSD off.
@@maxcardenas8257 For example, the playback is in flac file, and then playback with the built-in dsd mode. (You can switch the dsd on or off on your DAC player)
I think there are mainly two issues we are dealing with here, 1) people are not more evolved that others, they have opinions on things too just like everyone else and, 2) the equipment we have access to makes a big difference on what you get out of content you are watching or listening to, in that case most low end equipment does not cannot resolve ultra small differences. About two decades ago I bought a racing game to play on my not top end PC, it had beautiful font floating count down numbers. One day I put it on a more powerful PC and it turned out that it actually had a hot girl standing in front of the cars doing the count down (not floating numbers).
Unsuitable for manipulating audio in the studio, but the most suitable for capturing sound in all its exquisite nature, including for the most complex sounds, of orchestras and instrument concerts.
@@edfort5704 Recording engineers all over the world disagree. DSD is far less stable than PCM and much more susceptible to artifacts caused by clock jitter. While the artifacts can be avoided to an extent, the clocking accuracy required needlessly complicates the recording system with no real benefit. A PCM system with the same clock as a DSD system will always be more accurate and have higher fidelity. I think that the idea of DSD being more “natural” comes from a fundamental misunderstanding of how digital audio sampling works. I don’t mean to be confrontational, I’m just tired of hearing audiophiles complain about PCM as if the people making records for a living wouldn’t move to DSD if it had a real benefit
@@davegbro Recording engineers, for the most part, have a dog in the PCM vs DSD race. At gigantic sample rates (into the Megahertz) DSD and PCM may prove to have identical sonic qualities but, the way things are now with PCM standards set in place decades ago and having gospel status, DSD is a pioneer standard that help push down the old, false paradigms that are still peddled as truth: like the one claiming 16bit/44.1kHz covering all you can ever hear or need to hear. DSD is easier to work with in the giant paradigm shift that the audio world needs. Working with PCM means just hitting the wall of stuborness and refusal to evolve that the audio industry displays, such as the engineers you referred to.
@@edfort5704 I’m one of the recording engineers you speak of (with a background in electroacoustic engineering as well) and I’d switch in a heartbeat if there if there was any demonstrable advantage to DSD. I have yet to see it in practice and in theory. If you think that Nyquist, Shannon, and Whittaker were wrong about sampling (and the 16bit 44.1kHz standard that came as a result) please write a legitimate challenge to their work and publish it in the AES. You’d be a legend in the audio engineering community from that alone, assuming you could propose a real challenge. Contrary to what many audiophiles seem to believe, the people designing audio hardware know what they’re doing. Most of us actually went to school for this stuff and had to learn the theory and science behind it, and we use the science and theory on a daily basis to make a living. 16bit/44.1kHz will absolutely capture everything in the audible spectrum and do so while leaving some headroom at the top (22kHz ceiling in 44.1kHz sampling vs the 16kHz ceiling of the average adults ears). Higher sampling rates and bit depth do not increase the amount of resolution in the audible spectrum. That’s an old myth based on a serious misunderstanding of how quantization works in sampling. Unless you’re listening to music with information above 20k (that you can actually hear) 44.1k is more than enough. I do prefer to record at 24bit 96k, but that’s because of the options it opens for heavy audio manipulation, not because of “higher quality”. The only point I can concede in regards to 44.1kHz is that some old, very low quality converters really didn’t do well with 44.1kHz because of how low their anti-aliasing filters had to be, causing a noticeable reduction in high end clarity and detail due to the premature cutoff. Those converters were long gone by the 2000s though, so it’s a moot point in 2021. Here’s a question for you: what is your listening room equipped with to correct for room issues like resonance, reflections, etc?
@@davegbro If you, especially as an audio engineer, claim there is no demonstrable advantage to DSD over current PCM standards like CD 16bit/44.1 kHz, then you either have: - never heard a DSD recording or a good Vinyl transfer to it; - or you've not compared DSD and PCM on a system that does not manipulate or tamper with the original sound in any way by upsampling or other tricks; - or you're being disingenuous. To answer here early and expand a little on your question at the end, I have the following audio system: $80 DAC, $120 amp, $275 speakers that can resolve 50kHz of sound resolution, as claimed by the manufacturer. My listening room is my bedroom and does not have any special sound modifying features from a normal living bedroom. On my entry-level audiophile system, DSDs sound phenomenal compared to anything PCM (CDs, Flac, nevermind MP3s) - not perfect, but a huge increase in sound quality from old PCM standards. If you say you have never seen or heard DSD being better than PCM in practice and theory, then I pity you, because what I've heard with many DSD recording is just eyewateringly great by comparison to PCM. Here's why the audio industry's favourite sampling theorem does not apply to real life audio as it is being claimed to: In simple terms, the audio establishment has claimed for decades, based on the Nyquist sampling theorem, that in order to capture all 20 thousand or so frequencies that humans can hear (roughly and relatively speaking), that is 20Hz-20kHz, you only need to sample at twice the rate of the highest frequency in the range, so 2 X 20Khz = 40kHz, plus some headroom until 44.1 kHz for sonic-ultrasonic interaction and some filtering. The fundamental misconception about this is based on the false assumption and subconscious claim that we could only hear one sound frequency at a time, which is fundamentally false. Our ears and our bodies can hear and perceive multiple different sound frequencies happening at the same time. Any properly-designed speaker can also produce multiple sound frequencies with the same diaphragm.A mid-range woofer can produce multiple mid-range tones concurrently. A tweeter can produce multiple high-frequencies concurrently. The same with a sub and its low frequencies. And that's what music and real-world audio is made up from: multiple sound frequencies happening at the same time. For us to capture with digital audio what goes on in real world audio, we actually need to apply the Nyquist sampling theorem (2 x sampling) to EACH individual soundwave that happens at a given instant in time, so that the actual sampling necessary to capture sound in its entirety is the sum of Niquist samples applied to each concurrent soundwave that we can hear - and we can hear and perceive a lot of soundwaves going on at once! As an example, picture an instant of time in a real-life song that has these frequencies happening at once: a 100Hz, an 800Hz, a 6000Hz, a 12000Hz and a 16000Hz. Our ears could hear something of this sort and we would need digital audio to sample each sound wave and add them together in a sum of their samples in order to capture the full sound happening in that instant of time. So you would need a sample rate of 2x100Hz + 2x800Hz + 2x6,000Hz + 2x12,000Hz + 2x16000Hz = 69,800Hz sample rate, which exceeds the PCM CD standard by quite a lot. And it may be entirely possible that we could perceive a huge amount of soundwaves at the same time - maybe all 20 thousand or so soundwaves in the audible 20Hz-20kHz spectrum, if not all instantly, then at least by tuning our ears to different smaller sets of frequencies happening in the same instant - studies need to be done in this area. If it will someday be proven that we can hear and perceive all the 20thousand(ish) soundwaves of the audible spectrum hapenning concurrently, either together at once (though unlikely) or, more likely, by tuning/focusing our ears&hearing to different smaller sets of frequencies happening at once, then we may need samples rates that equal the Nyquist rate of the sum of all the soundwaves in the audible spectrum, which would be 2(Nyquist) x 200Mhz (the sum of all soundwaves in the 20Hz-20kHz audible spectrum) = 400Mhz - an enormous sample rate compared to the measly 44.1kHz in CD PCM. Standard rate DSD has 2.8 Mhz, still not enough if it turns out we can perceive the full 20Hz-20kHz audible spectrum at once, and I can definitely hear limitations in it as well, but at least it's a huge increase and progress from CD PCM. Now do you get it? As a side note, it's funny you suggesting that I write to some guild of audio engineers about these issues on 44.1kHz Nyquist sampling being inadequate for capturing sound in its perceivable entirety. These issues are so simple to get once you understand how digital audio works, that I believe most of them probably know or suspect my claims are true, but would prefer not to talk about them much given their huge technical requirements. Professional guilds of any sort have a history of being less than open to innovation, competition and free discussion of new ideas that challenge established dogmas. Their aim is more to create standards within their industry and enforce rules and acceptance of them by the people in their field. Read about the history of mercantile professional guilds in Europe, as it's an interesting one, quite ugly but funny - I recommend reading Frederick Bastiat's works on this. TLDR: DSD rules over established PCM standards like CD 16bit/44.1kHz because it captures more soundwaves in the audible 20Hz-20kHz range that we can actually hear and perceive, thus capturing more resolution and realism of real-world sound.
It seems what you actually "proved", if anything, is that "DSD is the best _playback_ medium in the world." I recently shelled out the cash for an ADC that can output PCM @ (maximum) 768 khz . Would that not be even better than 352 khz, for recording, etc.? My goal is simply to record as close to analog as possible, using 21st century tech, while bypassing the "pain in the (DSD)", since it is now feasible using ultra high resolution PCM. Once you have the original "in the (digital) can", just DAC to whatever you prefer. Even down-sample for folks who listen on earbuds ...
DSD is analog audio, the frequency and amplitude are modulated like analog FM on a square wave's "width". Very easy to capture that in the digital domain! Just sample for "false" or "true" (low or high signal) on a high frequency (2,8MHz) and there you are! That's also the reason why it is a 1 bit system. PCM is more complex. It's sampling at a lower frequency, like 192kHz, to determine the frequency of the analog signal. But that's only the frequency per sample. Not the amplitude per sample. Every sample also has a value in bits for the amplitude. In CD standard, 44.1kHz has 16 values (bits) available for the amplitude information per sample. 192kHz has 24/32 values (bits) available. Now you can feel what's going on. 16 or 24 amplitude values per sample can record and reproduce audio on a decent way, but it is far less detailed than a true analog signal. DSD, which is actually the "digital" sister of FM, will record and reproduce the sound like an FM radio. FM radio is sampled, DSD is sampled.
DSD all day and everyday for me. It changed the way I listen to digital recorded music. It the closest I ever heard to analog music on a digital recording without the "noise" that is captured on vinyl. PCM formats sound to rough and "unnatural" for me.
Here’s a dumb idea: record in PCM at whatever insane frequency (2+ MHz) and bit depth you want, that matches the desired DSD frequency, then edit it as PCM before downsampling to DSD. Would that work (barring the insane amount of data that would need to be processed), or am I crazy?
Ears that can, do! Ears that can not, disagree! I avoided Dsd for the reasons you highlighted about editing, i'll just stick to 192 and leave the headache pills in the cupboard! 😉😅👍
To me, DSD seems to be a bit of an elitist thing, which requires high-end(very expensive) non portable equipment. I only listen to music when I'm out and about, away from home.