DSD to PCM conversion

[@johnnytoobad7785]

The vast majority of these "Re-mastered" and "Reissued" LP's on thick vinyl are from DSD-to-PCM sources. Then D->A, EQ'd and compressed for Vinyl.

[@FrancisEVidal]

"pro audio" is at 48khz to sync perfectly with movie that were filmed at 24 fps.

[@angelwars3176]

Absolutely. Paul should know this.

[@JonAnderhub]

Except that movies are filmed at 22 fps and NTSC video was originally done at 29.97 fps (now 30 fps). The modern frame rates also include 24 fps and 60 fps.

[@FrancisEVidal]

@@JonAnderhub Movies filmed on actual film (not digital) were all at 24 fps.

[@JonathanGerardDias]

Been working in 88.2/176.4khz 64bitF/32bitF, honestly it sounds like dsd that I've heard at a friend's house.

[@stephenbrown2367]

Exactly I have added differing and I use 32bit float WAV and boy is it amazing once conveyed to DSD and again overwriting the DSD file and getting the true in studio feeling and I can tell the difference now☝️

[@stephenbrown2367]

@@Wizardofgosz are you saying 16bit 44.1 is better ? I am confused!

[@stephenbrown2367]

@@Wizardofgosz ok 👍 cool but I can tell the difference.And I used to produce my music so I knowing what good sound quality sounds like because I worked with wave and recording my own stuff at 192kghz

[@ThinkingBetter]

The vast majority of DAC chips are optimized for PCM conversion and DSD is internally converted to PCM first to do multi bit sigma delta conversion and then converted to analogue.

[@housepianist]

What I'm interpreting here is that, technically, you can hear differences between DSD and PCM *if* you use PS Audio designed and manufactured equipment to capture it and (presumably) play it back using that same equipment. Although I will never hear it in this sense, I don't doubt Paul's assessment. However, the likelihood of anyone having the exact same set up as in the Octave Records studio is extremely unlikely, therefore, the question becomes more about the resolving power of your average audiophile system versus the same in Paul's studio. With all the different systems out there, all with various degrees of resolving power, it could make detecting any DSD/PCM differences obvious at one end, and completely undetectable at the other. It's sort of like putting your ear close to a tweeter with your amp turned on. Most tweeters will produce some kind of faint (or not so faint) hissing noise. That's easy to hear because of where your ear is located. That's highly resolvable. But when you step back at a more normal listening distance, you don't hear it at all. Your ear simply cannot resolve that hiss from farther away. I think that's a metaphor for this issue. All the equipment in the studio is so high quality, it's like putting your ear up close to the music, which could certainly lend itself to hearing these DSD/PCM differences. If you do not posses such equipment, you may not hear any differences at all. Maybe it's more about our own systems (and ears!) than the digital files themselves.

[@NoEgg4u]

PS Audio is one of many high-end component / speaker manufacturers. Any other reputable, high-end equipment, that is professionally set up, will reveal what Paul spoke about in this video. And depending on what stereo equipment is used, it can / will be more apparent than Paul's flagship setup. Equipment from, for example, Aesthetix, Audio Research, Aurender, Ayre Acoustics, Belles, Innuos, PSB, Vandersteen, Magico, Wilson, Kef, etc, are all up to the job. "All the equipment in the studio is so high quality..." It depends on the studio. Some studios use mass produced, mid-fidelity (being generous) equipment. "Maybe it's more about our own systems (and ears!) than the digital files themselves." It is not a choice of one vs. the other. If it sounds better on a high-end system, then it will sound better on any other system. Lower-end systems will do a lousy job at revealing the differences (that goes for any listening test). But better sounding content sounds better than poorer sounding content, on any stereo. If the stereo is of too poor quality, none of this matters (like putting high-octane gasoline into an engine that runs badly -- the better gasoline might help, but so insignificantly that it serves no purpose). Better sounding content will never hurt, and can only help. How much help depends on the quality of the stereo. The better the stereo, the more it will reveal the subtleties between varying quality content.

[@housepianist]

@@NoEgg4u well most of what you've said is certainly your opinion and not a de facto standard. You can only theorize that your listed equipment, especially when set up properly, "will reveal what Paul spoke about". And I'm aware that there are other high-end manufacturers. But that does not leave the conclusion that equip like this will reveal what "Paul spoke about". Keep in mind that Paul said the recording and playback equipment, designed and set up by him, will reveal how much better DSD is over PCM. So unless you have some empirical evidence that the equipment produced by the aforementioned manufacturers can reveal "what Paul spoke about", it's all speculative. And besides, while Paul has spoken highly of many manufacturers products in the past, he will also tell you if some products come up short of what HE thinks should be right. In the end, all the fancy, high-end equipment in the world may or may not display any aural differences between DSD and PCM. The end user may or may not hear those differences. The only thing that we can be reasonably sure of is that we can most likely hear the most obvious characteristics in the music when using high-end equipment: a more distinct bass, clearer midrange, darker/brighter tone, realistic timbre, etc, etc. But with something as discerning as high-quality digital recordings, it's not so clear. And some people will simply not hear any immediate differences, regardless of the set-up. And that's a fact! In the event that you may possess some "other-worldly" knowledge and insight onwhat people can hear, I can only dismiss your statements as speculative and theoretical - unless of course you have all the equipment mentioned and have set it all up to actually do some comparative listening to DSD and PCM files. There's a much bigger point you're missing here but I'll reserve that for later, assuming you'll reply to this comment. 🤷‍♂️

[@NoEgg4u]

@@housepianist "So unless you have some empirical evidence that the equipment produced by the aforementioned manufacturers can reveal "what Paul spoke about", it's all speculative." I have an amazing sounding pressing of Led Zep II (the hot cut). It sounds amazing on my stereo. I took it to the store where I purchased my stereo, "speculating" that it would also sound amazing on their flagship stereo (equipment that is better than mine, and a custom room that is better than mine). Yes, I "speculated" that it my Led Zep II pressing would sound amazing on their stereo -- and it did. It surpassed my expectations. But how could I have known that it would sound amazing on the store's stereo (different speakers, different turntable, different everything)? When you gain enough experience on a topic, you get to know what is what -- how things work. I do not need to drive a Corvette to know it would be more fun than driving my Sentra. Ergo, I know about all of the equipment that I listed above (most of it I have heard in high-end stores). That is why I listed those brands. They will all do what Paul says his equipment does. I can slam on my car's brakes on the highway, for no reason, with someone behind me, and I could "speculate" that they are going to run into me. You can make the argument that I am speculating, or guessing. I will argue that I am certain, that at 65 MPH, with someone following me, that they will hit me if I mash down on my brakes for no reason. I am right. I know I am right. And the same goes for knowing I am right about the equipment that I listed. "...empirical evidence..." That is your way of dismissing what you have no meaningful reasons for being combative. Should I get a signed, notarized letter, accompanied with a video of me listening? Even with the above, you would likely state that I imagined what I heard. Why not try to learn, rather than impulsively taking contrary positions? Everything else you wrote evidences that you never heard a stereo on the level of Paul's FR30 room. Whereas, I have. I have heard approximately a dozen or more stereos, ranging from a few $thousand, to hundreds of $thousands, all of which would reveal what Paul spoke about in this video. I have also heard far more stereos, at wide-ranging prices, that would not reveal what Paul spoke about in this video (or my ear might pick it up, whereas the general public's ear will not, because I know how to listen, critically). From what you have written, my conclusion is you have only heard the latter -- non realistic sound-fields.

[@housepianist]

@@NoEgg4u oh good grief, another audiophile "pissing contest". 🤦‍♂️ I guess you're going to pull this sentence out and tell me how using term "pissing" is a euphemism for urinating but the result is the same - emptying one's bladder! Lol You are utterly and completely missing the point! It's a mystery why you fail to address the DSD/PCM questions and instead choose to fill your replies with pointless comparisons about slamming car brakes, or saying a Vet is more fun than your Sentara, or driving 65mph, blah, blah, blah. You are using physical, tangible examples to make a point about a completely subjective matter. Don't use physical science to make a subjective point about hearing difference between DSD and PCM. I use an example of listening to a tweeter up close and you're telling me about physical applications in driving a car lol. Ooooook..... 🤷‍♂️ Since you have not addressed why most people, even with high end systems, won't be able to detect differences in DSD and PCM, I have to assume that you know I'm correct but won't admit it. But hey, at least I will know that a Corvette will be more fun to drive than a Sentra lol. Or will it? Hmm....maybe I'll take that to a car forum and see. But someone will probably compare that to analog vs digital recordings. 😂

[@rbartsch]

Multiples or divisions of 48 kHz are standard in the digital world because they are divisions of the system frequencies of devices. Most devices use multiples of 48 kHz internally and interpolate audio with multiples of 44.1 kHz. Sony's odd 16 bit/44.1 kHz (Philipps suggested 14 bit/44.056 Hz) was used because bitrates over 45 kHz did not fit into the vertical blanking intervall of video tape recorders used in studios before solid state-memory.

[@glenncurry3041]

Yes few realize that 44.1K (Hz is "cycles per second" for signwaves, not clock cycles) was chosen to fit 3 samples into a horizontal line on a 3/4" Umatic VTR. I sold the PCM-F1 when it came out.

[@Bassotronics]

So… DSD only goes up to 30k? 🤷🏻‍♂️

[@JonAnderhub]

You don't have to know math to know that what Paul is saying doesn't add up because, 1. It has been factually proven in double-blind listening tests that DSD does not sound better than PCM. 2. Paul says 8:34 "...so DSD is the way to capture it, but then if you (alter the original signal) decimate it, if you put it through a low pass filter (to eliminate the large quantities of noise that must be added to DSD to quantize it), convert it to PCM there's very little loss..." so wait there's very little loss when converting DSD to PCM? Wouldn't it make sense that there would be no loss starting out recording in PCM? 3. Paul says at 8:48 ...go the other way record in PCM you've already lost the information from the live microphones..." WAIT, WHAT??? Paul just told us that converting DSD to PCM there is very little loss, now suddenly there is some kind of loss even though PCM uses the same type of converter and has better resolution than DSD and doesn't require low pass filters? Something isn't adding up here Paul. 4. BUT WAIT THERE'S MORE!!! Converting PCM to DSD requires more noise to be added and filtered yet again on playback (PCM requires no filtering on playback) and there's a significant loss. 5. Paul neglects to mention that his current system requires DSD to be converted to PCM in order to do any editing, and then converted back to DSD (wait didn't Paul tell us that converting from PCM to DSD incurs significant loss? Sorry, Paul but what you are saying doesn't add up at all.

[@davidcross890]

PCM DONE WITH ABSOLUTE TOTAL PROCESS OVERALL LOWEST DISTORTION QUALITIES AS THE VERY 1ST HELD PRIORITY IS EXTREMELY DIFFICULT TO BEST. I KNOW THEIR ARE ALL KINDS OF FLAVORS POSSIBLE WITH THE ALTERNATIVE DIGITAL CHOICES BUT AS A PLEASE GIVE ONLY THE REAL FACTS PLAYBACK I CHOOSE PCM IN ITS BEST POSSIBLE FORM UNTIL IT CAN ACTUALLY BE BESTED AND PROVEN TO BE BESTED IN A REAL COMPARATIVE AND ALTERNATE METRICS MEASURING TEST. I LIKE REAL COMPARATIVE TESTS AS THEY EXPOSE REALITY. WE ALL GET TO BENEFIT WHEN WE ACTUALLY IMPROVE OUR SOUND EXPERIENCE AND ENGAGEMENT. OF COURSE ANYONE IS ENTITLED TO LISTEN TO GREATER LEVELS OF DISTORTIONS OF THEIR CHOICE IF THEY WISH.

[@scottwolf8633]

Paul, I'm more interested in, HOW you design Topology without a strong Math background? I was Math/CS, for my first undergrad and hung out with the EE undergrads, as we shared the love for the bong and the misery of Calculus, I-IV, Differential Eqs. and then a semester of the Laplace Transform and Convolution Integral along with the computational language CSMP, AKA, Operational Mathematics for Systems Analysis, course. Post the OP Math, the EE's were done with their Math requirements, but unfortunately, I was not. BTW, I use an AudioNote Kit DAC employing R2R Topology that reads what was recorded, not a Delta-Sigma based Algorithm where a machine guesses at what was," Laid down". Then a Resistor/Transformer/Resistor, Topology for I/V conversion.

[@octaverecordsanddsdstudios1285]

Thanks, Scott. R2R is accurate to about 16 bits, 18 on a good day. One of the reasons as an industry we moved away from R2R is getting 24 bit dynamics and there you need a SDM. Resistors cannot be made with tolerances good enough to hit those numbers. But, I agree, like vinyl, there's a sonic goodness to RTR. As for HOW I design without heavy math there's probably a book to be written about learning by doing and through mentoring with skilled designers. Not room enough in this small comment section.

[@richardboyce4921]

please ramble some more. really love your talks Paul. Thank you

[@stimpy1226]

How did you do analog design without math? Before the HP programmable calculators, you would’ve needed a pencil and lab notebook to keep notes on your analog circuit designs which includes quite a bit of mathematics.

[@octaverecordsanddsdstudios1285]

Actually, not. Aside from calculating currents with Ohm's law-which is simple arithmetic-a -3dB points and filter values with 1/2πRC there's not much math involved.

[@johnnytoobad7785]

@@octaverecordsanddsdstudios1285 Less than 99% True. Only if you are using circuit simulation software...which does the "math" internally.

[@magoostus]

if the dsd low-pass filter is analog. that means its minimum-phase. and you can just as easily convert dsd to pcm with either a minum-phase OR linear-phase low pass filter. thus making pcm superior

[@Error2username]

Its like trying to get a sacd pressed into a 12"vinyl disc. The tecnology is getting better and better toghether with quality.

[@carlstineman274]

Knowing nothing about the actual DSD to PCM conversion process, I would guess that it goes something like this. Divide the DSD stream into time increments of 1/44100 seconds. Add up the number of 1s in that increment, multiply by 256 (number of SPL increments available in standard PCM) and divide by 64 (the number of DSD samples in a PCM time increment) and you would get the PCM amplitude value for that time increment, convert that value to binary and there you have it. Well within Paul's mathematical abilities (grin).

[@angelwars3176]

So to be clear Octave records record in DSD but mix/master in DXD (the dreaded PCM folks!!) the vast majority of DSD recording go this way and have done for many years now. So, question is simply why not just record in DXD to start with? (Many folk do just that these days but it's their own project choice and nothing wrong in that). Reasons are: 1) Because DSD itself is a better capture of the original audio signal regardless of what you then do with it. 2) It leaves you with the Pure DSD streams to work with in the future when possibly you could actually mix/master it without conversion maintaining its purity. If you've gone direct to DXD/PCM you've lost these forever. 3) Converting your DSD captures to DXD post recording is easier on the processors so renders a better DXD conversion to work with. Don't shoot the messenger folks this has all come from very reliable sources and I aint getting paid!

[@JonAnderhub]

I question your "reliable sources" because there is no scientific or factual proof to show that DSD is a better capture of the original audio. 1. In fact, there is significant evidence to show that PCM is actually better at capturing original source audio because of better quantization, resolution, and the lack of a need to add large quantities of noise to the original signal to maintain quantization. 2. DSD "streams" are not pure. The Delta-Sigma analog to digital conversion introduces distortion along with the need to add extreme quantities of noise which must be added to each DSD "stream" or channel (multiple channels must have noise added to each channel and then when these channels are combined the noise doubles with each channel added and then the need to use filters to remove the noise in the digital to analog conversion does not create a "pure" signal. PCM (DXD) on the other hand provides significantly better resolution at lower sample rates, thereby not overworking the Delta-Sigma converter (or the Sigma-Delta digital to analog converter) and PCM only requires the use of one Nyquist filter per "stream" or channel at twice the recorded frequency (even at 48 kHz that filter only has to reach down to 24 kHz and at higher sampling rates such as 352.8 kHz used for DXD conversion the Nyquist filter only has to filter out frequencies above 176 kHz (virtually no need for a filter on most audio circuits.) and requires NO filter for the conversion back to analog. 3. Recording direct to high-resolution PCM is much less stressful on Delta-Sigma converters and there is no need for processors to work at putting large quantities of noise into the above audio spectrum with each stream. Because PCM files are in the native processor language, and because PCM file sizes are smaller, and because there is no need to convert PCM files for processing, processors are much less stressed and PCM capture is significantly "purer" and provide much better replication of the original source that DSD or even analog equipment.

[@angelwars3176]

As said, don't shoot the messenger! If PCM 'direct' capture is sonically better as you are suggesting the simple question is why are many highly experienced, professional recording engineers bothering to do something that is complicated and tedious and loses money?

[@jimmyolsenschannel6263]

I recently bought the Simply Vinyl re-release of Procol Harum's 1969 LP "A Salty Dog". Supposedly it has gone through the same DSD-involved procedure as mentioned by Paul here. Comparing with my old US original copy of the same album, some of the differences I hear are hard to put a finger on technically, but let me try to explain it in layman's terms. The SV release has a kind of combined two- and three-dimensionality to it, akin to what you can experience visually with computer animation (or in a 1950s "Peep Show" or a 3D movie), i.e. the elements can seem to be in front of or behind each other (3D), yet individually they are flat like cut-out cardboard figures (2D). You may like or dislike it, but in my opinion it can hardly be classified as "High Fidelity".

[@geddylee501]

Oh god, more digital shenanigans lol

[@housepianist]

I know, right? 🙄😁

[@tomehCanada]

Paul, In "Analog to DSD, Vinyl and MoFi" you said at 4:18 that comparing a live Mic in the studio against a PCM stored and then DAC to monitors was "splippppppppp." Can you make available a small sample of a DSD capture and the same on PCM? I'm very curious since you displayed such a negative response to PCM. Best regards, Tom

[@digggerrjones7345]

This...again??

[@Mark-lq3sb]

I know, the arguments will never stop. Sometimes I do enjoy the Keyboard Bravado Boys. Paul can bring-up one subject and suddenly there's 10 people with 10 different explanations and each one would swear in court that they are correct, and no one else is, especially Paul, LOL! Poor Paul seems like more than 50% of the commenters believe he's always wrong. Opinions, that's what makes the world go around... Wait, that would be 10 different facts, LOL! 🤣

[@imkow]

I do DSD downsampling to 48khz all the time

[@davidcross890]

An interesting possibly creative take or opinion on reality

[@dangerzone007]

Has bit perfect been proven to make any difference?

[@NeilDSouza7]

Coming to think of it .. LaserDisc or LD actually recorded video signals in Analog .. If Laserdiscs made a comeback instead of Vinyl to record PURE ANALOG signal and reproduce it using modern laser transport mechanisms and digital error tracking correction. Then we would have a Perfect Analog reproduction of the Master Recording or Master Tape and DSD and PCM would not be needed .. What do you think ???

[@glenncurry3041]

Not entirely true. All "disc" storage (optical capacitive,...) used some form of binary "pits". In the case of LD the analog video signal was converted into Pulse WIDTH Modulation.

[@angelwars3176]

Laserdisc was brilliant for its time. Bladerunner put through my then simple stereo with a set of phono plugs bloody amazing!

[@stephencooke4569]

@@angelwars3176 I have a THX demonstration disc that sounds astounding, with a montage of clips from various Lucasfilm productions. It beat the pants off of anything similar on DVD.

[@louissilvani1389]

It’s amazing to me that would hold his conversion it doesn’t leave a signature 🤔

[@yvesmartin8259]

I hope GNR would have Octave Records remastered their UYI1 and 2.

[@davidfromamerica1871]

I am waiting for the dancing tower speakers to come out that dance to the rhythm of the music. Even Twerking.

[@dangerzone007]

But then you have to convert DSD to PCM to do EQ compression etc.

[@angelwars3176]

Well you don't have to but most folks do.

[@loonation2185]

this kind of questions never die lol

[@NeilDSouza7]

DSD to PCM conversion You get DPCMSD or whatever that is ... LOLOL

[@grumpy9478]

since the rise of the modern age of science & technology (heck, spiritual practice for that matter), there have been those who assert, "it's all figured out!", others who believe "there's more to it!" &/or "here's an alternative path". mixed in among all are technicians, discoverers, deluded, demented, visionaries, scammers, fabricators, inspired, observers, purists, combos of all.... & more. it has struck me (over seven decades) there are even those who bitch somewhat more than teach or create.

[@geddylee501]

A comment more than worthy of your name -;)

[@grumpy9478]

@@geddylee501 thx. happy was taken, sleepy was too accurate & dopey... well...

[@logtothebase2]

There is very good explanation the DSD to PCM conversion on this video kzbin.info/www/bejne/nompeqRnZq90l9U I think this is important as one thing that I think gets lost in the DSD is best narrative, is DSD is converted, and is wholly dependent on PCM for music production. Although at very high resolution (DXD) compared to domestic audio, there is no DSD without PCM conversions.

[@geddylee501]

It's true, hearing it all in a dutch accent gives it more gravitas ...

[@carmonben]

Y'all would be horrified by my main setup for music I run optical out of a Chromecast into a resampler to force the output to be 44.1 so my mixer can mix it with my other audio sources including Bluetooth from my phone and from my two computers...

[@hoobsgroove]

the why you took Sony be knocking on your door soon offering you a job 😂

[@angelwars3176]

Nah Sony aint interested in DSD any more sadly.

[@davidcross890]

My Comment is if PCM IS DONE WITH ABSOLUTE TOTAL PROCESS OVERALL LOWEST DISTORTION QUALITIES AS THE VERY 1ST HELD PRIORITY IS EXTREMELY DIFFICULT TO BEST. I KNOW THEIR ARE ALL KINDS OF FLAVORS POSSIBLE WITH THE ALTERNATIVE DIGITAL CHOICES BUT AS A PLEASE GIVE ONLY THE REAL FACTS PLAYBACK I CHOOSE PCM IN ITS BEST POSSIBLE FORM UNTIL IT CAN ACTUALLY BE BESTED AND PROVEN TO BE BESTED IN A REAL COMPARATIVE AND ALTERNATE METRICS MEASURING TEST. I LIKE REAL COMPARATIVE TESTS AS THEY EXPOSE REALITY. WE ALL GET TO BENEFIT WHEN WE ACTUALLY IMPROVE OUR SOUND EXPERIENCE AND ENGAGEMENT. OF COURSE ANYONE IS ENTITLED TO LISTEN TO GREATER LEVELS OF COLORATIONS, ENHANCEMENTS, OR DISTORTIONS OF THEIR OWN CHOICE IF THEY WISH.

[@gg.6967]

48 & 96 khz are a derivative of the digital Video / Audio Recording Industry.

[@gg.6967]

I’ll assume the audio recording industry adopted the Video/ Audio standards possibly due to the preferred clock. But anyway please correct this for my own educational purposes. I’m enjoying discovering & learning, listening etc.😃

[@gg.6967]

😉192 khz is derived from 96khz but is not a derivative of the Video industry like that of 48 & 96 kHz , it is an audio industry standard.

[@MufasaRizz]

And CD quality is all you need no matter what they tell you. 44.1 to 16 bit the human hearing. The rest is mambo jumbo.

[@shipsahoy1793]

To your point, all this extra nonsense really is ridiculous.. it’s kind of like saying I have 100 dB signal to noise ratio, but 110 dB is better, or 30KHz audio bandwidth is better than 20 ! 🤣 It really gets ridiculous after a while with this multitude of “standards,” and lack of attention to realistic pragmatism. I look at it as a mental gymnastics game that some “digital” people need to keep garnering a paycheck…

[@joaoandre4158]

This channel is pure snake oil

[@joaoandre4158]

What they dont tell you is that this dsd SACDs have DIFERENT MASTERRING from regular cd and thats why it sounds diferent just like vinyl.

[@AT-wl9yq]

Do you have any science to back that up or are we just supposed to take your word for it?

[@MufasaRizz]

@@AT-wl9yq Read all about the human hearing. We hear 20hz to 20khz, thats it and its on a cd at 44.1 to 16 bits. The rest is all snake oil.

[@tombuehner6119]

Who cares!

[@MiklosKoncsek]

I myself, find this kind of audio engineering trivia interesting. One fact I heard, is that you have, say, 16 bit reproduction with the humble CD. Yet, how much better than a studio reel-to-reel machine is it? CD is 16 bit depth, but you take the biggest, latest analogue tape machine, running at the fastest speed, with the widest tape, you're only gonna reach a maximum theoretical bit depth of...13 bits. Another, when Hi-fi stereo vcrcame along in the early 80's, audio enthusiasts flocked to them, just for the audio recording ability. Their audio reproduction trumped most home reel-to reels, what with 70dB noise to signal ratio, stereo seperation of also 70dB, to top that al off it had a dynamic range of 90dB! This is BIG stuff to me...

[@geddylee501]

@@MiklosKoncsek that's true, people noticed how much better the wider tape in a VCR sounded than a regular cassette, man that was years ago lol