I greatly appreciate this channel and the info Paul shares but I think every person who considers themselves an audiophile should take a course to record and mix a couple of songs themselves so they can get a better sense for what actually goes on behind their favorite music. I see too many people commenting here that don’t seem to understand just how much gear and processing your favorite music has gone through before it reaches your desired format...and how much budget gear at that (beat up SM57s, gunked up old tube amps, countless pieces of rack gear both digital and analog and of all levels of quality through multiple busses and summing, etc etc etc). The music making process is not what I would call “pristine” even when great care is taken. The “magic” we hear and love is amazing performances put through gear that the engineers felt helped those performances. After all, what is art without its color choices? Mixing engineers use that gear like an artist uses their colors. It’s not about a “perfect recording” in the end. It’s about how it makes us feel when we listen to it and if it took a thousand pieces of gear to get it there, then that’s what it takes. I think we should all remember this when we listen to our favorite music. Modern DAWs can do 32 bit 384khz so I don’t think there is any limitation to PCM as a format anymore...I would be curious to hear a comparison of DSD and PCM in its highest resolution.
DSD is great but the quality of the artist, composition, instruments used and sound of the recording room is far more important. DSD recordings sound very real but if the reality is a crappy song recorded in a small rehearsal room with rented gear and mediocre musicians I prefer a great recording from great musicians in PCM.
I prefer recording great performances in DSD, as it so so far better than our 2” 24 tracks can record, let alone PCM. If you haven’t experienced DSD 256+, just go to a great live performance and you get the idea. It is NOT a minor difference.
sorry, bro, but all you'll find on this channel is opinions. actual issues, hypotheses, design implementations, and results-gathering will not be covered on this channel.
Best format is PCM because it works through the entire plumbing of digital audio. The audio fidelity of PCM is as good as you make it to (higher sample rate and higher sample resolution). The problem with DSD is that it doesn't actually work with any codecs or DSP audio processing in modern audio architectures. DSD over PCM is a hack. It's fair to say that 44.1kHz 16 bits is not perfect, but PCM is not limited to that. Nowadays you can stream (e.g. from Amazon HD music) bit perfect 192kHz 24 bits. Besides modern DACs actually are not optimized for DSD but rather use a combination of sigma delta and other decoding principles in combination.
I champion DSD as the best way of capturing an audio signal in the most accurate and robust way but the bottom line is to get the benefit from that you need to keep it 'pure' - so basically I agree with you. The thing I've learned recently is many labels record in DSD and keep these captures in the hope they can release this material in the future when tools exist to process them (they consider it does sound better) in the meantime going down the ubiquitous DXD route. Why not just record into DXD? for the reason stated above - you would lose the DSD capture and also I'm reliably informed, converting the DSD captures to DXD post recording delivers possibly a better DXD conversion as the processors are not doing the whole job on the fly. I don't agree with your DAC comment, the argument is sigma delta ended DAC's benefit from having DSD fed into them which is why many folks are converting their PCM files before sending using HQ player or whatever...
@@angelwars3176 This article sums it up well: benchmarkmedia.com/blogs/application_notes/41262017-audio-myth-dsd-provides-a-direct-stream-from-a-d-to-d-a
@@ThinkingBetter yet the article is very measuring and technical, but it is about how it sounds. If a cap is a cap and a cable is a cable, a cow is a cow and it is all the same, yet 'the same' far from true.
Unfortunately the format won’t matter much if the music is not interesting. Conversely, if the music is good, the playback format matters less. Find good music you enjoy and all the techie stuff just fades away.
So you take the DSD files, then convert them to PCM for editing, then convert them back again to DSD? How is that better than just recording directly to PCM in the first place? Inquiring minds want to know.
@@johnholmes912 Please, site your scientific proof of statement. (I know there are lossy methods of conversion, often used by people trying to listen to DSD recordings on hardware that isn't built for it. I'm far from an expert on the subject myself, but do trust someone I know IS an expert - Paul McGowan- to know what he's talking about.)
Labels that record in DSD then convert to PCM(DXD) often want to keep the DSD captures for future release( if they can process them in the future). Another reason is converting to DXD after first capturing to DSD can lead to a better more robust conversion rather than recording directly on the fly to PCM - I've been reliably told.
See Douglas Blake's thread starting with "Interesting ... that console Paul is sitting in front of ..." , where he explains how mixing and mastering is usually done. In another video Paul has explained how they record in DSD, mix in analog on that console and record the output in DSD, thus avoiding a conversion to PCM. In the same thread Angelwars states "... DSD is arguably the best way to CAPTURE an analogue audio signal in an accurate and robust format that can be stored in 'digital' equipment and the folks that capture their recordings in DSD do so for this reason - they believe it sounds better!" That is the reason the RECORDINGS are made in DSD and not PCM. HOWEVER, once the conversion is made from DSD to PCM, I cannot see the point of converting it back to DSD.
What we do is we mixdown to tape one quarter-inch running at 15 IPS then we go to the DSD from the tape and master from DSD. That's gives us a dsd and tape premaster for archive. We usually master to pcm 24 96 but for our upcoming vinyl we will master back to tape. Wow this stuff makes your head hurt😉
I wish we audiophiles can pick one based on ultimate sound quality, but in the end, we are at the mercy of the decisions made by major labels. Thank God for PS Audio's Octave Records.
@@vylbird8014 this is the truth for all you "audiophiles." it's in your head, bro. or your walls. or your fingertips. but not the technology, that would be an insane idea.
The DSD to PCM and back to DSD conversion reminds me of Sprockets on SNL when something was translated From English to German then back into English. Or was it the other way around?
It seems a silly procedure, DSD-PCM-DSD. Better then to record DSD-multitrack, mix analog en master to stereo DSD. No you can't use your terrible audio work station pcm effect plugins then ;-) And if it is true that the difference between DSD and PCM is so small then why don't get rid of DSD altogether? (DSD is better of course)
@Ed Fort The big problem is that almost all music is recorded in PCM 48khz or 96khz and a lot of music is mixed and eq adjusted poorly! Proper adjustment of the EQ. and other tIng when making a song is more important than if it is PCM. or DSD. !
DSD is digital and can hopefully correct for media/stream errors, so possible both signal values (bits) and the timing (sampling rate, pulse width or ...) _must_ be quantized in some way: that's just, albeit complex, physics. That also means that the digital information rate is limited. I can imagine that at the same price-point and information rate, it is easier to construct a better ADC or DAC for DSD when compared to some other formats. Studios often use internal formats internally that can contain much more information than DSD. However, recording the source in DSD might profit from a better ADC. And also converting it to DSD might improve playback at the consumer, because of the relatively better DAC. In general, however, I guess a studio "downgrade" to 192/96 kHz at 24-bit and a very good DAC also works for the consumer. Pay attention, however. A lot of equipment uses professional studio-grade ADC/DAC but a crappy analogue path and a shitty power supply. That gives good specs under ideal conditions. However, it might voice radio signals for you or require laboratory-grade cleaned power sockets. My advice: buy something else, even if you ditched your fridge already.
"Digital audio has used somewhat the same format for nearly 40 years, is there something better on the horizon?" How is DSD "on the Horizon?" SACD was launched over 20 years ago, and is now nearly extinct.
I remember SACD being pronounced dead 20 years ago, and yet it's still here. I still buy new titles regularly. Many would consider high-end audio as a whole nearly extinct, yet we're still here.
@@brianmoore581 They're not still "here" in any practical way. The percentage of titles released on SACD is pretty dismal. I think I've bought one in the last year. For general / public purposes, it's still pretty much dead.
@@ReflectedMiles I buy one or two a SACDs a month. It's not dead. They're not hard to find. Sure, it's not a mass market product, but neither is any brand of audio equipment I own. You're right that you aren't going to be picking up Lorde's latest on SACD. Not a problem for me. Lots of Classical, and lots of Classic Rock, is available on SACD. You can't get everything you might want, but there are thousands of titles you can get, and most of them sound spectacular. Just because the format didn't take over the world doesn't mean you should ignore it. People still listen to vinyl records and reel to reel tapes, too. You can't get everything on either of those formats, either.
So if you have converted the DSD to PCM for editing/mixing etc., why not just leave it as PCM, don't convert back to DSD as one less conversion stage is surely better. If you want DSD as a consumer/end user format, then do your mixing etc live (as the musicians performs) then convert the live analogue to DSD and leave it alone, basically the same as straight to disk. If the resultant DSD file has issues, delete it and get the artist back into the studio and do it all again.
DSD all the way. I don't know all the techincal stuff, but since I heard DSD music files, even when I upsample my Flac and Wav music files to DSD, it just sounds better, clearer and easier to hear for my ears.
@@jamotter8967 You don't upsample. You can convert PCM (FLAC is a PCM container)to DSD using music playing software like HQ player, 'on the fly' before sending it to your DAC
Octave Records DSD recordings are the best of the best. Silky where there's silk, gritty where there's grit. It's as real as it gets, for me at least. Whatever your taste in music, it's surely high fidelity.
Musicians being able to play together? Listener fatigue of the same old pop effects and treatments being overused? Direct to stereo or ambisonic recordings in acoustic environments that suit the performance? Engineers placing microphones correctly and not “fixing it in the mix”? All of the above?
They need to do what 3d modellers did and create a unified format like fbx...it alliws for interoperability between apps and its always the same no matter what platform you are viewing on...
Listen to DSD and you´ll never return to other formats. And right, the biggest problem is the volume without converting back. But engineers would find a way in the future, so we could listen to PDM GaN FET amplifiers.
@@johnholmes912 Are you talking about the same thing that happens with digital volume controls? I believe the proper term is bit stripping. The more you lower the volume, the more resolution you loose.
@Taco I don't get what you're getting at. He said dsd is the superior recording code but people record in pcm because it's easier to manipulate. Then he said that octave records in dsd and then converts to pcm for mixing.
@Taco that comment makes even less sense. He also specifically says that if you record in dsd and then convert to pcm afterwards you lose basically nothing or very little.
If I can't hear a difference on some top notch headphones, then I see no advantage. Has anyone detected a difference between DSD and 16/44.1 PCM or Hi-Res PCM?
I can hear night-and-day differences between DSD and CD PCM on my $80 Sennheiser headphones. Speakers are even better, as even a decent $300 set of them tends to have higher sound resolution capability than headphones.
DSD is an obsolete and technically inferior format for music releases and is no longer supported by Sony. A good DSD recording and a good PCM recording are audibly indistinguishable. However, the DSD file will be enormous and so unsuitable for streaming. Converting DSD to PCM for editing then back to DSD is highly questionable. Why not leave out the last step and go with the PCM version? The answer is marketing. Audiophiles mistakenly think they are buying a superior DSD recording.
No it's not marketing. Labels that record in DSD do so because they consider this capture to be superior and often archive these captures for a future in which hopefully they can process them for release without conversion. Also going DSD then conversion to PCM post recording rather than straight to PCM appears to create a better sounding conversion as the processors are working easier. Bottom line is DSD is a superior way of capturing an audio signal but cannot be mixed without some kind of conversion. Pure DSD released recordings are as rare as rocking horse shit.
One question: have you actually compared DSD to PCM on a high res system? I always doubt myself and am not capable of stating any truth with certainty, yet every time I listen to my DSD albums I am shocked by the overwhelming natural and easy sound. Therefore I have difficulty understanding your statement 'are audibly indistinguisshable' Perhaps my system is inferior on PCM playback? To me, PCM is good, DSD is so much better...
A question for the experts. Be it DSD or PCM are the files restricted to the cloud or can they be stored on physical media such as CDs or DVD? I frankly abhor the cloud. Call me a fetishist perhaps but I need to collect an object.
Is there any group/authority out there that sets universal standards for recording? From my small understanding, there was sort of a "gentlemens agreement" of sorts for some time until the early 00's when Rick Rubin decided to start the loudness wars. As other worldly the man is, its been sort of downhill since.
Yes. The RIAA set some technical standards for the music industry, though it isn't their principle purpose. ITU do standards in telecom, so there's some overlap.
Paul is absolutely 99,9% right. DSD is magnificent. Truly amazing. Please start using DSD. There are devices for playback in many pricesranges. Thank you Paul and others advocating for DSD. Hopefully many start using DSD so more artists and labels produce DSD to enjoy.
DSD is only going to stay a niche recording market, you cannot mix/master and carry out DSP without converting it which defeats the object. Pure DSD recordings are extremely rare!
@@CJNooberson That is a matter of choice. Old recordings will never be full DSD. I've got a one-mic session by Carmen Gomes, full DSD, which is a treat. A master-tape to DSD of for example Eric Bibb sounds like a high speed tape, wonderful all the same. If you hear what Gus Skineas (Octave) can achieve with DSD, the richness, texture and sound.... After all, records, tapes, (SA)CD's do cost a lot more than streaming. Quality will have a price. I would love to pay for DSD quality albums which I love, even though they weren't recorded in DSD.
It's been shown pretty conclusively that 1-bit high-sampling DSD is not in any technical or measurable way better than 24-bit/96khz PCM - it's different, but it's not intrinsically better. Sure, you can (in theory) design simpler DACs when using DSD, but in practice that doesn't matter, because everyone has gotten really REALLY good at designing cheap, excellent PCM DACs. I collect SACDs, I don't pretend DSD is better than garden variety 96/24 or higher PCM cuz it simply isn't. PCM works great! Even 44.1Khz covers more dynamic range than any of us can actually hear, and the extra headroom you get from 96/24 and 192/24 is excellent at capturing ultrasonics in the analog source, if you believe those matter.
The whole recording methods want changing to a much simpler approach to capture a live sound you would hear in front of you well rehearsed in one take only simple analogue mixing no editing but no big record companies do this pro tools etc rule
Hi Paul! I bought your Don Grusin album a while back and I have to say I don't like it. I used my CD player to play the album and it has some sort of nasty flavor to the sound. I don't know if it is the DSD format or your Sonoma DSD recording system that is adding that. I remember when I would play SACD's on my old Sony SACD player that I heard more resolution but really didn't enjoy the sound. Actually I have a recording of Carmina Burana recorded in DSD and it is the worst sounding classical recording that I have. It just sounds off. Like the notes aren't in the right place or something.
I had a costumer coming in for an audio system and did the obvious selling trick of playing a 96kbps MP3 vs a FLAC on a decent consumer-grade system. Yet, I could tell from the guy's face, he just couldn't hear the difference. He bought the set-up, so I did my job, but it's all so subjective in the end...
For digital interconnects USB is everything anyone would ever need ever again. It's got bandwidth for all the channels you are ever going to hook up at any sample rate you ever need and it's got error correction in case a bit ever got lost. I'm not going to buy anything without USB ever again!
@@vylbird8014 It's already 25 years since USB 2.0 was introduced, time to move on. Even USB 3.0 is now 10 years old. Audio manufacturers are true fossils, it's no wonder the industry is dying.
@@jukkamaljanen6644 Even USB3.0 can't do guaranteed low-jitter. There's simply no call for it: USB is aimed at the consumer space, and no consumer cares about having sub-millisecond-precise timing in their audio. Why complicate the design for a feature that hardly anyone wants?
@@vylbird8014 I'm no expert on FireWire, but USB does have adjustable buffer sizes etc which helps with latency. It's more up to the capabilities of the connected devices than the bus itself. I'd guess anything that anything you do to a sound signal in between interconnects takes more time than the transfer from box to box. Anyway, live applications do require minimal latency so any technology that emphasises on it can surely be useful.
@@jukkamaljanen6644 It's not just about latency, it's jitter - variation in latency. You might have an audio device on your USB bus, but what if you have a storage device too? Then when the storage device has some data to move, it'll occupy communications for a few miliseconds, and cut off the audio device. The only way to handle this is by putting a buffer in the audio, which means the effective latency becomes equal to the worst-case latency. For most applications this doesn't matter. No person watching TV is going to mind if their audio is five milliseconds off. For professional audio it does matter, when you are dealing with multiple audio devices and have to worry about keeping signals captured with such timing precision as to preserve the phase. For this reason, high-end professional mixing desks used to have firewire interfaces. Today they may have thunderbolt. This will cost you, of course. It's far in excess to what any home studio will need, but if you're managing a professional recording studio with enough microphones to voice every instrument in an orchestra individually, you're going to need that kind of gear.
i got the sacd cranberries zombie and didn't like how they mixed it. I got the diana krall non-dsd quiet album and it blew me away. Sometimes having the sacd tech is good, other times, requiers good mix men to make good product. same think with pink floyd. the dvd was awesome. the sacd not so much.
That's a special case. For the SACD launch, they had Tim P from EAR remaster the entire album with a bunch of tube gear. That remaster doesn't sound anything like the others, or the original. I have nothing against tubes. I use them in my own system, but the remaster for DSOTM just didn't come out right. It was a well done album to begin with, and just didn't need to be fixed. In my opinion, where they went wrong with SACD is how they structured the format. There were 2 choices. Single layer and dual layer. They should have all been dual layer. You would have had people building their SACD collections even if they didn't know what it was. Eventually, some may have taken interest in the format and realized that they already have a bunch of them. Just a missed opportunity, I think.
well mastered 44.1 kHz 16 bit digital CD quality audio is already perfect audio. but it must be well mastered and not "brickwalled to shit" and "de-noised to death"^^
It depends on what is meant by "better." Better sounding in terms of detail / density / preserved original signal? That has long depended on how much one believes in the placebo effect ("Of course I can hear 192kHz resolution. That's why I paid $120,000 for this German-built, Amish-finished passive amplifier art.") If by "better" one means the perfect reproduction of the original performance including its environment, as if sitting there, that will require more than just a different format or additional channels and speakers.
you try to make speakers that have as flat a frequency curve as possible in relation to the ear. and then music should have as flat a frequency as possible without sharp frequencies. Good dynamic is also important and is the variation between high and low sound that can be compared to the contrast on a tv.
There's a giant flaw in your assumptions, which also happens to be (or at least is slowly becoming) the elephant in the room for the audio industry. That flaw is the false idea that the point of higher sample rates in audio is to capture higher frequency sounds than the top audible sounds for the human ear. It's not. The point of higher audio sample rates (audio industry's hi-res audio BS marketing notwithstanding) is to still capture mostly only audible sounds, within the 20Hz-20kHz spectrum, but to capture more of them at the same time, to create better detail, dynamic range, quality and fidelity to the real music/audio being recorded.
@@edfort5704 That is the oldest argument there is about higher sampling rates, along with the ability to reproduce square waves accurately, and the net effects have been shown to be indistinguishable even to those who say they can distinguish them. I have conducted quite a number of such experiments on such claimants myself. There's no way to get increased audible detail from a 192 kHz sample, even with the signal up at 5 or 10 kHz (the practical top end of musical hearing, especially over the age of 40 or with a live-music background), than there is at half that sample rate. That's because half that rate already delivers fantastic levels of detail very accurately. One thing audiophile companies and proponents always need to be vehemently opposed to and dismissing at every opportunity is blind A/B listening. If it were to ever become routinely available with regard to any technology or equipment, it might well ruin the industry. It is that big of a reality check.
@@edfort5704 The big problem is that almost all music is recorded in PCM 48khz or 96khz and a lot of music is mixed and eq adjusted poorly! Proper adjustment of the EQ. and other tIng when making a song is more important than if it is PCM. or DSD. !
@@ReflectedMiles 192kHz? Psst. Well there's your problem. First, drop the PCM paradigm and let"s just talk about 1-bit Modulation, like the way DSD works, since the PCM talk only serves to confuse people (which, I know, is useful to many). Try recording sound at a couple of hundred megahertz (yes mega, not kilo) sample rate, run it through an entire equipment workflow capable of handling such sample rates (from mic diaphragms capable of vibrating a few hundred million times a second, to connectors able to handle the amount of bits and ADC/DACs designed for it, plus amplifiers and speakers able to resolve such sound res that would total a few hundred million vibrations/second of soundwaves mostly still in the 20Hz-20kHz range). Your ears and mind would be dumbfounded by the difference in sound with the right setup able to handle all this sound resolution. And it would still consist of soundwaves in the audible range, no 120kHz audible sound bs. Fact is what I've said is incredibly hard to implement though, so people prefer the alternative of bs-ing themselves and others with the same old trope that 44.1kHz is all you need, when in fact it's like playing in a puddle when you could be swimming and exploring the ocean instead.
Yes but what does dsd sound like on vinyl can you not print some vinyl better for artists as well get more money! just put a download link in the record. What about a dsd player is there such a thing? It could be recorded on DAT I suppose what does that sound like?
Here we go again. "DSD is a Form of PCM"....What? Paul PLEASE do a better job of explaining what PDM is. You can edit DSD without conversion though you cannot mix it like PCM because it isn't actually digital. Most people who record in DSD already convert it to a high rate PCM for mixing and mastering and then back to DSD - it's called DXD (PCM 352.8/24bit) and has been around for years! DSD is certainly the best most accurate and robust way of capturing an audio signal but you just cannot mix and manipulate it like digital PCM which is why PCM has become the ubiquitous format for music recording and distribution. The vast majority of labels that record in DSD convert to DXD and create their finished project (edit master) in that format. Many keep the original DSD captures as they consider them superior and hope to be able to release them in the future when proper tools are available to mix and work on this material. Also I am reliably informed capturing first to DSD then converting to DXD after recording can give a better result than straight to DXD (less 'on the fly' processing) The bottom line is there are very, very few pure DSD recordings available to buy - last count on nativedsd.com was 40 out of a massive catalogue.
No, there is no better digital audio format. There are all computerised manipulated sound. Listen to music as it should be heard in full analogue from source through processing and through analogue circuitry, turntable , tape player onto loudspeakers. No tampering.
"computerised manipulated sound" ...Analogue gets manipulated by the analogue medium (hiss, pops etc.), the player (speed/pitch fluctuations), and the preamp (such as tube "warmth" or RIAA equalization etc.), it's hardly 1:1 true to the source. Enjoyment of these imperfect characteristics is absolutely fine, but making objectively false statements on the internet; "That's a paddlin'". Also, to avoid digital you would be denying yourself the majority of music produced in the last 30 years.
@@johnb7009 I am not denying digital. I listen to digital a lot. That is why I know that analogue sound is better. BTW, the so called digital you hear is also analogue (DAC). It is the processing of the signals that is done using digital technology so called fuzzy logic of 1 and 0s. Used in fridges, washing machines, TVs etc.
I greatly appreciate this channel and the info Paul shares but I think every person who considers themselves an audiophile should take a course to record and mix a couple of songs themselves so they can get a better sense for what actually goes on behind their favorite music. I see too many people commenting here that don’t seem to understand just how much gear and processing your favorite music has gone through before it reaches your desired format...and how much budget gear at that (beat up SM57s, gunked up old tube amps, countless pieces of rack gear both digital and analog and of all levels of quality through multiple busses and summing, etc etc etc). The music making process is not what I would call “pristine” even when great care is taken. The “magic” we hear and love is amazing performances put through gear that the engineers felt helped those performances. After all, what is art without its color choices? Mixing engineers use that gear like an artist uses their colors. It’s not about a “perfect recording” in the end. It’s about how it makes us feel when we listen to it and if it took a thousand pieces of gear to get it there, then that’s what it takes. I think we should all remember this when we listen to our favorite music. Modern DAWs can do 32 bit 384khz so I don’t think there is any limitation to PCM as a format anymore...I would be curious to hear a comparison of DSD and PCM in its highest resolution.
DSD is great but the quality of the artist, composition, instruments used and sound of the recording room is far more important. DSD recordings sound very real but if the reality is a crappy song recorded in a small rehearsal room with rented gear and mediocre musicians I prefer a great recording from great musicians in PCM.
I agree wholeheartedly. There are so many variables that could swallow any inherent "benefits " that DSD may have.
I prefer recording great performances in DSD, as it so so far better than our 2” 24 tracks can record, let alone PCM. If you haven’t experienced DSD 256+, just go to a great live performance and you get the idea. It is NOT a minor difference.
DSD is better, ok, but why exactly? more accurate reconstruction filters, no time smearing? Please give infos, not only opinion.
sorry, bro, but all you'll find on this channel is opinions. actual issues, hypotheses, design implementations, and results-gathering will not be covered on this channel.
Best format is PCM because it works through the entire plumbing of digital audio. The audio fidelity of PCM is as good as you make it to (higher sample rate and higher sample resolution). The problem with DSD is that it doesn't actually work with any codecs or DSP audio processing in modern audio architectures. DSD over PCM is a hack. It's fair to say that 44.1kHz 16 bits is not perfect, but PCM is not limited to that. Nowadays you can stream (e.g. from Amazon HD music) bit perfect 192kHz 24 bits. Besides modern DACs actually are not optimized for DSD but rather use a combination of sigma delta and other decoding principles in combination.
im deep into dsp now, do not want to lose that and go to another format like dsd.
I champion DSD as the best way of capturing an audio signal in the most accurate and robust way but the bottom line is to get the benefit from that you need to keep it 'pure' - so basically I agree with you. The thing I've learned recently is many labels record in DSD and keep these captures in the hope they can release this material in the future when tools exist to process them (they consider it does sound better) in the meantime going down the ubiquitous DXD route. Why not just record into DXD? for the reason stated above - you would lose the DSD capture and also I'm reliably informed, converting the DSD captures to DXD post recording delivers possibly a better DXD conversion as the processors are not doing the whole job on the fly.
I don't agree with your DAC comment, the argument is sigma delta ended DAC's benefit from having DSD fed into them which is why many folks are converting their PCM files before sending using HQ player or whatever...
@@angelwars3176 This article sums it up well: benchmarkmedia.com/blogs/application_notes/41262017-audio-myth-dsd-provides-a-direct-stream-from-a-d-to-d-a
@@sudd3660 Yes, but rest assured the world will not shift back to DSD. It died ages ago because it doesn’t actually allow any modern processing.
@@ThinkingBetter yet the article is very measuring and technical, but it is about how it sounds. If a cap is a cap and a cable is a cable, a cow is a cow and it is all the same, yet 'the same' far from true.
Unfortunately the format won’t matter much if the music is not interesting. Conversely, if the music is good, the playback format matters less. Find good music you enjoy and all the techie stuff just fades away.
100% agree with you.
Yes.. does not matter if your favourite track is not available in dsd format.
Are there any scientific studies to support dsd being better than pcm?
So you take the DSD files, then convert them to PCM for editing, then convert them back again to DSD? How is that better than just recording directly to PCM in the first place? Inquiring minds want to know.
it is not better, the conversion process is very lossy
Exactly
@@johnholmes912 Please, site your scientific proof of statement. (I know there are lossy methods of conversion, often used by people trying to listen to DSD recordings on hardware that isn't built for it. I'm far from an expert on the subject myself, but do trust someone I know IS an expert - Paul McGowan- to know what he's talking about.)
Labels that record in DSD then convert to PCM(DXD) often want to keep the DSD captures for future release( if they can process them in the future). Another reason is converting to DXD after first capturing to DSD can lead to a better more robust conversion rather than recording directly on the fly to PCM - I've been reliably told.
See Douglas Blake's thread starting with "Interesting ... that console Paul is sitting in front of ..." , where he explains how mixing and mastering is usually done.
In another video Paul has explained how they record in DSD, mix in analog on that console and record the output in DSD, thus avoiding a conversion to PCM.
In the same thread Angelwars states "... DSD is arguably the best way to CAPTURE an analogue audio signal in an accurate and robust format that can be stored in 'digital' equipment and the folks that capture their recordings in DSD do so for this reason - they believe it sounds better!"
That is the reason the RECORDINGS are made in DSD and not PCM.
HOWEVER, once the conversion is made from DSD to PCM, I cannot see the point of converting it back to DSD.
What we do is we mixdown to tape one quarter-inch running at 15 IPS then we go to the DSD from the tape and master from DSD. That's gives us a dsd and tape premaster for archive. We usually master to pcm 24 96 but for our upcoming vinyl we will master back to tape. Wow this stuff makes your head hurt😉
I wish we audiophiles can pick one based on ultimate sound quality, but in the end, we are at the mercy of the decisions made by major labels. Thank God for PS Audio's Octave Records.
Blue Coast Music has been recording in DSD for years as have many other labels.
You can't, because the limiting factor in sound quality is no longer in the technology. It's in your ears.
@@vylbird8014 this is the truth for all you "audiophiles." it's in your head, bro. or your walls. or your fingertips. but not the technology, that would be an insane idea.
The DSD to PCM and back to DSD conversion reminds me of Sprockets on SNL when something was translated From English to German then back into English. Or was it the other way around?
It seems a silly procedure, DSD-PCM-DSD. Better then to record DSD-multitrack, mix analog en master to stereo DSD. No you can't use your terrible audio work station pcm effect plugins then ;-)
And if it is true that the difference between DSD and PCM is so small then why don't get rid of DSD altogether? (DSD is better of course)
🤗 HONESTY IS THE BEST POLICY …. But seldom popular 😉😍😍😍
@Ed Fort The big problem is that almost all music is recorded in PCM 48khz or 96khz and a lot of music is mixed and eq adjusted poorly!
Proper adjustment of the EQ. and other tIng when making a song is more important than if it is PCM. or DSD. !
DSD is digital and can hopefully correct for media/stream errors, so possible both signal values (bits) and the timing (sampling rate, pulse width or ...) _must_ be quantized in some way: that's just, albeit complex, physics. That also means that the digital information rate is limited.
I can imagine that at the same price-point and information rate, it is easier to construct a better ADC or DAC for DSD when compared to some other formats. Studios often use internal formats internally that can contain much more information than DSD. However, recording the source in DSD might profit from a better ADC. And also converting it to DSD might improve playback at the consumer, because of the relatively better DAC. In general, however, I guess a studio "downgrade" to 192/96 kHz at 24-bit and a very good DAC also works for the consumer.
Pay attention, however. A lot of equipment uses professional studio-grade ADC/DAC but a crappy analogue path and a shitty power supply. That gives good specs under ideal conditions. However, it might voice radio signals for you or require laboratory-grade cleaned power sockets. My advice: buy something else, even if you ditched your fridge already.
Pure DSD256 is pure gold.
"Digital audio has used somewhat the same format for nearly 40 years, is there something better on the horizon?"
How is DSD "on the Horizon?" SACD was launched over 20 years ago, and is now nearly extinct.
I remember SACD being pronounced dead 20 years ago, and yet it's still here. I still buy new titles regularly.
Many would consider high-end audio as a whole nearly extinct, yet we're still here.
@@brianmoore581 They're not still "here" in any practical way. The percentage of titles released on SACD is pretty dismal. I think I've bought one in the last year. For general / public purposes, it's still pretty much dead.
@@ReflectedMiles I buy one or two a SACDs a month. It's not dead. They're not hard to find. Sure, it's not a mass market product, but neither is any brand of audio equipment I own. You're right that you aren't going to be picking up Lorde's latest on SACD. Not a problem for me. Lots of Classical, and lots of Classic Rock, is available on SACD. You can't get everything you might want, but there are thousands of titles you can get, and most of them sound spectacular. Just because the format didn't take over the world doesn't mean you should ignore it. People still listen to vinyl records and reel to reel tapes, too. You can't get everything on either of those formats, either.
All my DSD collection is direct from the original reel to reel tapes.
Here I am looking at reel to reel and cassette tape to buy.
So if you have converted the DSD to PCM for editing/mixing etc., why not just leave it as PCM, don't convert back to DSD as one less conversion stage is surely better.
If you want DSD as a consumer/end user format, then do your mixing etc live (as the musicians performs) then convert the live analogue to DSD and leave it alone, basically the same as straight to disk. If the resultant DSD file has issues, delete it and get the artist back into the studio and do it all again.
What are the speakers in the background ???
DSD all the way. I don't know all the techincal stuff, but since I heard DSD music files, even when I upsample my Flac and Wav music files to DSD, it just sounds better, clearer and easier to hear for my ears.
How does one "upsample" FLAC to DSD? Can I then stream my newly created DSD files? Please advise.
@@jamotter8967 You don't upsample. You can convert PCM (FLAC is a PCM container)to DSD using music playing software like HQ player, 'on the fly' before sending it to your DAC
When is octave records going to release albums on vinyl?
Yes, 'analog kid, digital man, DSD grandpa ! Thanks Paul. :-))
Octave Records DSD recordings are the best of the best. Silky where there's silk, gritty where there's grit. It's as real as it gets, for me at least. Whatever your taste in music, it's surely high fidelity.
How's the music? At the end of the day, that's the only thing which really counts.
dsd is awful, the information loss is quite apalling
@@johnholmes912 ?
What will it take for the industry to record in DSD again? What's out there in DSD is so extremely limited, it's massively depressing.
Musicians being able to play together? Listener fatigue of the same old pop effects and treatments being overused? Direct to stereo or ambisonic recordings in acoustic environments that suit the performance? Engineers placing microphones correctly and not “fixing it in the mix”? All of the above?
They need to do what 3d modellers did and create a unified format like fbx...it alliws for interoperability between apps and its always the same no matter what platform you are viewing on...
Did he just say "Fucks wit it?"
Listen to DSD and you´ll never return to other formats. And right, the biggest problem is the volume without converting back. But engineers would find a way in the future, so we could listen to PDM GaN FET amplifiers.
the signal degrades whenever you try to do anything with it.......there's a reason why studios avoid it
@@johnholmes912 Are you talking about the same thing that happens with digital volume controls? I believe the proper term is bit stripping. The more you lower the volume, the more resolution you loose.
So no link to octave records where we can purchase some dsd?
@Taco uhm he also said it needs to be initially recorded in dsd first which is what octave does. It's like you didn't watch the video
@Taco No, you can buy the recording in DSD64 on an SACD or as a download.
@Taco I don't get what you're getting at. He said dsd is the superior recording code but people record in pcm because it's easier to manipulate. Then he said that octave records in dsd and then converts to pcm for mixing.
@Taco that comment makes even less sense. He also specifically says that if you record in dsd and then convert to pcm afterwards you lose basically nothing or very little.
If I can't hear a difference on some top notch headphones, then I see no advantage.
Has anyone detected a difference between DSD and 16/44.1 PCM or Hi-Res PCM?
I can hear night-and-day differences between DSD and CD PCM on my $80 Sennheiser headphones.
Speakers are even better, as even a decent $300 set of them tends to have higher sound resolution capability than headphones.
DSD is an obsolete and technically inferior format for music releases and is no longer supported by Sony. A good DSD recording and a good PCM recording are audibly indistinguishable. However, the DSD file will be enormous and so unsuitable for streaming. Converting DSD to PCM for editing then back to DSD is highly questionable. Why not leave out the last step and go with the PCM version? The answer is marketing. Audiophiles mistakenly think they are buying a superior DSD recording.
No it's not marketing. Labels that record in DSD do so because they consider this capture to be superior and often archive these captures for a future in which hopefully they can process them for release without conversion. Also going DSD then conversion to PCM post recording rather than straight to PCM appears to create a better sounding conversion as the processors are working easier. Bottom line is DSD is a superior way of capturing an audio signal but cannot be mixed without some kind of conversion. Pure DSD released recordings are as rare as rocking horse shit.
One question: have you actually compared DSD to PCM on a high res system? I always doubt myself and am not capable of stating any truth with certainty, yet every time I listen to my DSD albums I am shocked by the overwhelming natural and easy sound. Therefore I have difficulty understanding your statement 'are audibly indistinguisshable' Perhaps my system is inferior on PCM playback? To me, PCM is good, DSD is so much better...
A question for the experts. Be it DSD or PCM are the files restricted to the cloud or can they be stored on physical media such as CDs or DVD? I frankly abhor the cloud. Call me a fetishist perhaps but I need to collect an object.
Of course you can buy DSD on SACD.
Compact disc-all the way, baby! 40K and counting!
I love the red book 💿, too….still no 1 ….and so inexpensive right now as every fool and his brother jumps on vinyl 🥱
@Taco yeah, the so called loudness war, that startet sometime in the mid 90's , gave CD's a bad name that this format does not deserve!!!
DSD is the future!
Mmm what about these multi track tape recorders in the studios. If that’s the zenith well let’s have some of that please
Is there any group/authority out there that sets universal standards for recording? From my small understanding, there was sort of a "gentlemens agreement" of sorts for some time until the early 00's when Rick Rubin decided to start the loudness wars. As other worldly the man is, its been sort of downhill since.
Yes. The RIAA set some technical standards for the music industry, though it isn't their principle purpose. ITU do standards in telecom, so there's some overlap.
Paul is absolutely 99,9% right. DSD is magnificent. Truly amazing. Please start using DSD. There are devices for playback in many pricesranges. Thank you Paul and others advocating for DSD. Hopefully many start using DSD so more artists and labels produce DSD to enjoy.
So purchase all my music again in DSD?
DSD is only going to stay a niche recording market, you cannot mix/master and carry out DSP without converting it which defeats the object. Pure DSD recordings are extremely rare!
@@CJNooberson What you can do is convert PCM to DSD on playback - this can be beneficial depending on your DAC.
@@CJNooberson That is a matter of choice. Old recordings will never be full DSD. I've got a one-mic session by Carmen Gomes, full DSD, which is a treat. A master-tape to DSD of for example Eric Bibb sounds like a high speed tape, wonderful all the same. If you hear what Gus Skineas (Octave) can achieve with DSD, the richness, texture and sound.... After all, records, tapes, (SA)CD's do cost a lot more than streaming. Quality will have a price. I would love to pay for DSD quality albums which I love, even though they weren't recorded in DSD.
It's been shown pretty conclusively that 1-bit high-sampling DSD is not in any technical or measurable way better than 24-bit/96khz PCM - it's different, but it's not intrinsically better.
Sure, you can (in theory) design simpler DACs when using DSD, but in practice that doesn't matter, because everyone has gotten really REALLY good at designing cheap, excellent PCM DACs.
I collect SACDs, I don't pretend DSD is better than garden variety 96/24 or higher PCM cuz it simply isn't.
PCM works great! Even 44.1Khz covers more dynamic range than any of us can actually hear, and the extra headroom you get from 96/24 and 192/24 is excellent at capturing ultrasonics in the analog source, if you believe those matter.
The problem is not the audio format but how the music is made and adjusted. A format does not get better than what you put into it if you understand?
The whole recording methods want changing to a much simpler approach to capture a live sound you would hear in front of you well rehearsed in one take only simple analogue mixing no editing but no big record companies do this pro tools etc rule
Hi Paul! I bought your Don Grusin album a while back and I have to say I don't like it. I used my CD player to play the album and it has some sort of nasty flavor to the sound. I don't know if it is the DSD format or your Sonoma DSD recording system that is adding that. I remember when I would play SACD's on my old Sony SACD player that I heard more resolution but really didn't enjoy the sound. Actually I have a recording of Carmina Burana recorded in DSD and it is the worst sounding classical recording that I have. It just sounds off. Like the notes aren't in the right place or something.
it is the least stable format of all, whenever you do anything to it the information degrades
See if you can get a copy of the studio's half-track master (assuming it was just an analog feed). 😊
44.1/24 PCM in a Flac container CAN be as good as it gets... Unless someone is trying to sell you something
@Taco .wav = PCM...
@Taco FLAC is lossless !!!
I had a costumer coming in for an audio system and did the obvious selling trick of playing a 96kbps MP3 vs a FLAC on a decent consumer-grade system. Yet, I could tell from the guy's face, he just couldn't hear the difference. He bought the set-up, so I did my job, but it's all so subjective in the end...
For digital interconnects USB is everything anyone would ever need ever again. It's got bandwidth for all the channels you are ever going to hook up at any sample rate you ever need and it's got error correction in case a bit ever got lost. I'm not going to buy anything without USB ever again!
Only one limitation: Jitter. USB has too much for some really time-critical purposes. Firewire was better in that regard.
@@vylbird8014 It's already 25 years since USB 2.0 was introduced, time to move on. Even USB 3.0 is now 10 years old. Audio manufacturers are true fossils, it's no wonder the industry is dying.
@@jukkamaljanen6644 Even USB3.0 can't do guaranteed low-jitter. There's simply no call for it: USB is aimed at the consumer space, and no consumer cares about having sub-millisecond-precise timing in their audio. Why complicate the design for a feature that hardly anyone wants?
@@vylbird8014 I'm no expert on FireWire, but USB does have adjustable buffer sizes etc which helps with latency. It's more up to the capabilities of the connected devices than the bus itself. I'd guess anything that anything you do to a sound signal in between interconnects takes more time than the transfer from box to box. Anyway, live applications do require minimal latency so any technology that emphasises on it can surely be useful.
@@jukkamaljanen6644 It's not just about latency, it's jitter - variation in latency. You might have an audio device on your USB bus, but what if you have a storage device too? Then when the storage device has some data to move, it'll occupy communications for a few miliseconds, and cut off the audio device. The only way to handle this is by putting a buffer in the audio, which means the effective latency becomes equal to the worst-case latency.
For most applications this doesn't matter. No person watching TV is going to mind if their audio is five milliseconds off. For professional audio it does matter, when you are dealing with multiple audio devices and have to worry about keeping signals captured with such timing precision as to preserve the phase.
For this reason, high-end professional mixing desks used to have firewire interfaces. Today they may have thunderbolt. This will cost you, of course. It's far in excess to what any home studio will need, but if you're managing a professional recording studio with enough microphones to voice every instrument in an orchestra individually, you're going to need that kind of gear.
👍
i got the sacd cranberries zombie and didn't like how they mixed it. I got the diana krall non-dsd quiet album and it blew me away. Sometimes having the sacd tech is good, other times, requiers good mix men to make good product. same think with pink floyd. the dvd was awesome. the sacd not so much.
That's a special case. For the SACD launch, they had Tim P from EAR remaster the entire album with a bunch of tube gear. That remaster doesn't sound anything like the others, or the original. I have nothing against tubes. I use them in my own system, but the remaster for DSOTM just didn't come out right. It was a well done album to begin with, and just didn't need to be fixed.
In my opinion, where they went wrong with SACD is how they structured the format. There were 2 choices. Single layer and dual layer. They should have all been dual layer. You would have had people building their SACD collections even if they didn't know what it was. Eventually, some may have taken interest in the format and realized that they already have a bunch of them. Just a missed opportunity, I think.
@@solarfall2728 i was playing it on tubes - 300b SET!!!!! oh well.
This is not about music anymore. More of a huge waste of time.
well mastered 44.1 kHz 16 bit digital CD quality audio is already perfect audio. but it must be well mastered and not "brickwalled to shit" and "de-noised to death"^^
Why is so much crap being thrown at DSD? Speakers and microphones create significantly greater distortion.
Ill pass on Digital Audio ,,, Ive still got analogue ears.
It depends on what is meant by "better." Better sounding in terms of detail / density / preserved original signal? That has long depended on how much one believes in the placebo effect ("Of course I can hear 192kHz resolution. That's why I paid $120,000 for this German-built, Amish-finished passive amplifier art.") If by "better" one means the perfect reproduction of the original performance including its environment, as if sitting there, that will require more than just a different format or additional channels and speakers.
you try to make speakers that have as flat a frequency curve as possible in relation to the ear.
and then music should have as flat a frequency as possible without sharp frequencies.
Good dynamic is also important and is the variation between high and low sound that can be compared to the contrast on a tv.
There's a giant flaw in your assumptions, which also happens to be (or at least is slowly becoming) the elephant in the room for the audio industry.
That flaw is the false idea that the point of higher sample rates in audio is to capture higher frequency sounds than the top audible sounds for the human ear. It's not.
The point of higher audio sample rates (audio industry's hi-res audio BS marketing notwithstanding) is to still capture mostly only audible sounds, within the 20Hz-20kHz spectrum, but to capture more of them at the same time, to create better detail, dynamic range, quality and fidelity to the real music/audio being recorded.
@@edfort5704 That is the oldest argument there is about higher sampling rates, along with the ability to reproduce square waves accurately, and the net effects have been shown to be indistinguishable even to those who say they can distinguish them. I have conducted quite a number of such experiments on such claimants myself. There's no way to get increased audible detail from a 192 kHz sample, even with the signal up at 5 or 10 kHz (the practical top end of musical hearing, especially over the age of 40 or with a live-music background), than there is at half that sample rate. That's because half that rate already delivers fantastic levels of detail very accurately. One thing audiophile companies and proponents always need to be vehemently opposed to and dismissing at every opportunity is blind A/B listening. If it were to ever become routinely available with regard to any technology or equipment, it might well ruin the industry. It is that big of a reality check.
@@edfort5704 The big problem is that almost all music is recorded in PCM 48khz or 96khz and a lot of music is mixed and eq adjusted poorly!
Proper adjustment of the EQ. and other tIng when making a song is more important than if it is PCM. or DSD. !
@@ReflectedMiles 192kHz? Psst. Well there's your problem. First, drop the PCM paradigm and let"s just talk about 1-bit Modulation, like the way DSD works, since the PCM talk only serves to confuse people (which, I know, is useful to many).
Try recording sound at a couple of hundred megahertz (yes mega, not kilo) sample rate, run it through an entire equipment workflow capable of handling such sample rates (from mic diaphragms capable of vibrating a few hundred million times a second, to connectors able to handle the amount of bits and ADC/DACs designed for it, plus amplifiers and speakers able to resolve such sound res that would total a few hundred million vibrations/second of soundwaves mostly still in the 20Hz-20kHz range).
Your ears and mind would be dumbfounded by the difference in sound with the right setup able to handle all this sound resolution. And it would still consist of soundwaves in the audible range, no 120kHz audible sound bs.
Fact is what I've said is incredibly hard to implement though, so people prefer the alternative of bs-ing themselves and others with the same old trope that 44.1kHz is all you need, when in fact it's like playing in a puddle when you could be swimming and exploring the ocean instead.
snake oil....mike drops.
Yes but what does dsd sound like on vinyl can you not print some vinyl better for artists as well get more money! just put a download link in the record.
What about a dsd player is there such a thing? It could be recorded on DAT I suppose what does that sound like?
A DSD player is called an SACD player.
@Douglas Blake Nothing to stop you from converting a DSD file to analog then cutting it to vinyl.
@@gotham61 are you sure about that
@Douglas Blake I meant after it's converted to analogue
DSD EXTREME!?
Here we go again. "DSD is a Form of PCM"....What?
Paul PLEASE do a better job of explaining what PDM is. You can edit DSD without conversion though you cannot mix it like PCM because it isn't actually digital. Most people who record in DSD already convert it to a high rate PCM for mixing and mastering and then back to DSD - it's called DXD (PCM 352.8/24bit) and has been around for years!
DSD is certainly the best most accurate and robust way of capturing an audio signal but you just cannot mix and manipulate it like digital PCM which is why PCM has become the ubiquitous format for music recording and distribution. The vast majority of labels that record in DSD convert to DXD and create their finished project (edit master) in that format. Many keep the original DSD captures as they consider them superior and hope to be able to release them in the future when proper tools are available to mix and work on this material. Also I am reliably informed capturing first to DSD then converting to DXD after recording can give a better result than straight to DXD (less 'on the fly' processing)
The bottom line is there are very, very few pure DSD recordings available to buy - last count on nativedsd.com was 40 out of a massive catalogue.
nobody cares
@@zebunker I do.
DSD travels disguised on the common path of PCM, in the DSD device it throws of its mask and rulez in all glory alike xD
No, there is no better digital audio format. There are all computerised manipulated sound. Listen to music as it should be heard in full analogue from source through processing and through analogue circuitry, turntable , tape player onto loudspeakers. No tampering.
"computerised manipulated sound" ...Analogue gets manipulated by the analogue medium (hiss, pops etc.), the player (speed/pitch fluctuations), and the preamp (such as tube "warmth" or RIAA equalization etc.), it's hardly 1:1 true to the source. Enjoyment of these imperfect characteristics is absolutely fine, but making objectively false statements on the internet; "That's a paddlin'".
Also, to avoid digital you would be denying yourself the majority of music produced in the last 30 years.
i'm sorry but dsd is pretty crap, the signal degradation makes it worse than analogue
@@johnb7009 The hiss an pop is in your head. You know that. Get off your butt and clean the records before you play it, you lazy sod.
@Douglas Blake The analogue sound is recorded from the source by a recording engineer using a sound mixer.
@@johnb7009 I am not denying digital. I listen to digital a lot. That is why I know that analogue sound is better. BTW, the so called digital you hear is also analogue (DAC). It is the processing of the signals that is done using digital technology so called fuzzy logic of 1 and 0s. Used in fridges, washing machines, TVs etc.
And then there’s MQA … with MQA CD’s sounding better than SACD that in turn sound better than Redbook CD’s. Shhhh..,