Paul you are absolutely killing it with the videos lately man...Thank you for them. I can only imagine how unreal and awesome it must be for you right now...top of the world, via a life time, dedicated to audio nirvana amid these recent achievements. At least that is what I hope it is!
The best I can do right now in my home studio is capture the master in DSD. My DAW Cakewalk has the export option of a DSD format. On the tracking side of things, the best that can be done is DoP. Oh well! However limited my setup is, even the master export being an upsample sounds fantastic! Thanks all you at Octave and Paul, for turning me on to this.
80 minute CD is 700 MB. You must have meant GB when you were talking about CD being just under 1 MB ? When I was with the CBC in the late 1990's and early 2000's we were running a system called DALET for broadcast program packaging but in the studio's we were running Sonic Solutions, Pro-Tools, Soundscape, and Pyramix depending on the operations required by the recordings. Drama and documentary production used Soundscape and later Pro-Tools. Serious music used Sonic Solutions and then moved to Pyramix. I was very happy using Soundscape but gradually transitioned to Pro-Tools once their HD systems were proven. The classical music people used the Sonic Solutions system but once they saw the early versions of Pyramix they lobbied for that to be their preferred system so when I left we were basically using Pro-Tools and Pyramix with DALET being the broadcast system used for current affairs and network delay recording. Since DALET was based on essentially mp3 recording we as recording engineers were required to know how to use it but we found it very limiting in its' ability to edit and crossfade. It was also a two channel system at the time whereas the others, Pro-Tools and Pyramix were multi-track capable as was Soundscape previously. Sonic Solutions had essentially been a four channel system (2 stereo tracks or four mono tracks).
They are Samsung Android tablets. I hate them but they were cheap. I can't figure out half of how to make them just do their job. They continually demand to be upgraded, etc.
@@Paulmcgowanpsaudio Hmmm... Paul, Thank you for your graciousness in taking time to respond. So Paul, aside from the software... is there a list of specific hardware which would be required to capture DSD within the audio-chain as well (or is it specifically the software)?
@@sekritskworl-sekrit_studios But Nuendo, REAPER, SoundForge, Samplitude/Sequoia, and Studio One Pro can record in DXD (Did i mention if Sequoia, Samplitude, and SoundForge "should" able to record in DSD?)
Funny in another video you said that DXD used by pyramix was blah. So what changed Paul? What happened to the zephyr filter? Think you need a video to square this.
In older versions of Pyramix yes, but with 14 installed it's pretty amazing. I am not certain if they changed anything or we got our act together in the long learning curve required. Also, we've changed setups in the meantime. Too many changes all at once to claim anything one way or the other. Comparing Zephiir to DXD straight out of the DAW box I don't hear enough of a difference to warrant the extra use anymore.
If I couldn't sleep and I wanted someone to tell me a fairy tale this video would be it. Sony began developing PDM (DSD) recording systems in 1999. Sony was already using PCM multitrack and mastering systems in the 80s. The 24-track PCM-3324 was released in 1982, and the later 48 track PCM-3348 was released in 1989. Sony WAS NOT looking for a digital system that was more "analog-like" when they developed DSD. In fact, Sony developed the DSD format or 2.8 MHz 1-bit PCM so that they could transfer their catalog of analog master tapes to a digital format without compromising the quality of those source tapes. By recording the 2.8 1-bit bitstream output Sony could produce a digital version of a master tape without committing to a specific sample rate and word length and so they could keep smaller digital files. DSD (PDM) failed miserably in this last regard due to the large quantity of noise needed in a DSD file. Additionally, while Sony and Phillips had their hands deep in the CD technology licensing business Sony had no financial ties to the DVD market so they hoped to use PDM as a multi-channel format for the DVD market. Sony and Philips took DSD to the DVD Forum. The forum had issued a “request for proposal” in the late 90s to all technology companies and research institutions looking for the next audio encoding scheme to move consumer audio to high-resolution and multichannel. The DVD Forum considered higher resolution PCM (96 and 192 kHz and 24-bits) and Sony/Philip’s DSD 64 (2.8224 MHz 1-bit PCM). After extensive evaluation and independent investigations done by experts in the field, the forum determined that PCM met the standards of their RFP and DSD did not. DSD was rejected because of its inherent and substantial high-frequency noise, the lack of available production tools, and the lack of production hardware. Let's also clear up this false notion that DSD (PDM) is more "analog like". Above 25kHz the quantization noise rises sharply, requiring far more sophisticated filters and/or noise-shaping algorithms. When filtered, the output of DSD64 with a simple low-pass filter, the result is distorted phase/time and some rather nasty artifacts in the audible range.
There is no need to add filter at 20kHz, we can't hear the noise. It is -120db real world in audible frequency. Now it is near -150dB in DSD512, of course PCM can do the same at 386kHz. Both DSD and PCM consume about the same binary data. Practically don't need any filter at all. Because of PWM nature of SDM DAC, DSD has it edge over PCM w/o addition conversion. Your phone sound chipset is already SDM AD/DA. Your phone is constant covert any PCM to DSD like data for the converter. And lots of phone chip is DSD like data direct drive to class D like amplification. PWM and PDM propagation in the same matter, DSD and Class D amp work seamlessly.There is one audiophile loudspeaker manufacture employ direct DSD driven to Class D amp active loudspeaker. Not like PCM->PDM/PWM->Filter->Analog stage-> Class A, A/B,D, H,G amplification as standard chain.
Jon, thank you for this information. It took some time. It makes perfect sense the DVD forum rejected DSD for exactly the reasons you state. However, not among them is how much better it sounds, at least to these ears and those of people we trust.
@@kongwee1978 While human hearing (on average by the way, not definitively) may be limited to 20 kHz, audio equipment is not. Ultrasound noise can cause audible non-linear distortions in playback on an audio system therefore a low pass filter is absolutely necessary. Additionally, The noise used to quantize DSD reaches as low as 15 Khz, well within the audible spectrum, and while the signal level may be as low as -150 dB at 15 kHz when added to an audible signal the signal is altered, and distortion is created. The noise level increases significantly as the frequency gets higher and is as high as -36 dB in the upper spectrum. When multiple DSD signals are combined, such as in multitrack recording, the noise level doubles for each signal added. DSD files are considerably larger than PCM files because of the need to record the noise in the upper spectrum to maintain dithering. PCM dithering noise is in the "audible spectrum" but well below the threshold of hearing, especially at higher bit rates, and therefore there is no need to record or place ultrasonic information in the file. Your example of the phone is correct, though it should be noted that the phone does not use DSD for file storage or use. The phone converts PCM to DSD at the output stage so what you are hearing is a DSD representation of a PCM digital signal. PWM and PDM propagation are not the same. PWM has a clearly distinguished period, PDM does not have a clearly distinguishable period. Pulse density modulation can be adjusted at any clock cycle without any extra work; unlike pulse width modulation, which can only be modified effectively once per period.
@@JonAnderhub DSD noise is -120db w/o filer at audibe frequency. Whatever bits above 22kHz, it not a concern until you want to put a filter in to get rid of aliasing. Way back before DSD. 1 bit SDM is already at propagate at 2.8224mhz on oversampling PCM. Indirectly you are using PCM like PDM/PWM. PDM and PWM are about the same. That why they can use directly at Class D. DSD is PDM because it need time interval is needed. Orelse you can't track your songs. Class D amp is PWM as they don't care about the time interval. PDM signal is really PWM for Class D amp. 0-2.8Mhz frequency, Class D amp won't care. They don't need to sync with clock frequency. If you read the Class D amp, they don't have clock frequency at amplifying stage. They only understand on and off state.
ahhh the tired, old and wrong ...anti DSD marketing hype....."DSD was rejected because of its inherent and substantial high-frequency noise" - this is utter and pure BS.(show me your audible "noise" hearing capabilities at north of 25Khz ...also now that DSD is at 4x (256). your claim "By recording the 2.8 1-bit bitstream output Sony could produce a digital version of a master tape without committing to a specific sample rate" is false -- 2.8Mhz IS the sample rate...just like 2X DSD is 5.6Mhz and 4x DSD is 11.2 Mhz .....this guy knows nothing of modern DSD recording and the emerging multi bit DSD future!
Good luck yall... its a smart business move in a way since there are no longer any real players in the DSD game anymore. Hopefully you can get some really desirable releases, bc at the end of the day it doesnt matter how much marketing speak you can lay on for the audiophiles or how good it sounds... its all about the songs and catalogue. Get yourself A&R perhaps ...
Paul you are absolutely killing it with the videos lately man...Thank you for them.
I can only imagine how unreal and awesome it must be for you right now...top of the world, via a life time, dedicated to audio nirvana amid these recent achievements.
At least that is what I hope it is!
How very kind. Thank you! And yes, I am having a blast.
The best I can do right now in my home studio is capture the master in DSD.
My DAW Cakewalk has the export option of a DSD format.
On the tracking side of things, the best that can be done is DoP. Oh well!
However limited my setup is, even the master export being an upsample sounds fantastic!
Thanks all you at Octave and Paul, for turning me on to this.
At 1:30, for those new to the game, that should be 1 Gigabyte, not one Megabyte.
You right
80 minute CD is 700 MB. You must have meant GB when you were talking about CD being just under 1 MB ? When I was with the CBC in the late 1990's and early 2000's we were running a system called DALET for broadcast program packaging but in the studio's we were running Sonic Solutions, Pro-Tools, Soundscape, and Pyramix depending on the operations required by the recordings. Drama and documentary production used Soundscape and later Pro-Tools. Serious music used Sonic Solutions and then moved to Pyramix. I was very happy using Soundscape but gradually transitioned to Pro-Tools once their HD systems were proven. The classical music people used the Sonic Solutions system but once they saw the early versions of Pyramix they lobbied for that to be their preferred system so when I left we were basically using Pro-Tools and Pyramix with DALET being the broadcast system used for current affairs and network delay recording. Since DALET was based on essentially mp3 recording we as recording engineers were required to know how to use it but we found it very limiting in its' ability to edit and crossfade. It was also a two channel system at the time whereas the others, Pro-Tools and Pyramix were multi-track capable as was Soundscape previously. Sonic Solutions had essentially been a four channel system (2 stereo tracks or four mono tracks).
Doh! Thanks for the correction.
Really enjoy your video's!
Do you use Merging Technologies Horus DSD256 AD/DA converter?
The Hapi versions, yes.
Really interesting video - thank you! Quick question: what tablets are you using with your Avid control surfaces? They don’t look like iPads. Thanks!
They are Samsung Android tablets. I hate them but they were cheap. I can't figure out half of how to make them just do their job. They continually demand to be upgraded, etc.
So can Nuendo, Reaper, and other Daws not capture in DSD?
That is correct. They do not.
@@Paulmcgowanpsaudio Hmmm...
Paul, Thank you for your graciousness in taking time to respond.
So Paul, aside from the software... is there a list of specific hardware which would be required to capture DSD within the audio-chain as well (or is it specifically the software)?
@@sekritskworl-sekrit_studios
But Nuendo, REAPER, SoundForge, Samplitude/Sequoia, and Studio One Pro can record in DXD
(Did i mention if Sequoia, Samplitude, and SoundForge "should" able to record in DSD?)
nope ...they can not.
Funny in another video you said that DXD used by pyramix was blah. So what changed Paul? What happened to the zephyr filter? Think you need a video to square this.
In older versions of Pyramix yes, but with 14 installed it's pretty amazing. I am not certain if they changed anything or we got our act together in the long learning curve required. Also, we've changed setups in the meantime. Too many changes all at once to claim anything one way or the other. Comparing Zephiir to DXD straight out of the DAW box I don't hear enough of a difference to warrant the extra use anymore.
Forgot to mention how tiny Pyramix and not GUI resizable and how glitchy it is :-))
If I couldn't sleep and I wanted someone to tell me a fairy tale this video would be it.
Sony began developing PDM (DSD) recording systems in 1999.
Sony was already using PCM multitrack and mastering systems in the 80s.
The 24-track PCM-3324 was released in 1982, and the later 48 track PCM-3348 was released in 1989.
Sony WAS NOT looking for a digital system that was more "analog-like" when they developed DSD.
In fact, Sony developed the DSD format or 2.8 MHz 1-bit PCM so that they could transfer their catalog of analog master tapes to a digital format without compromising the quality of those source tapes. By recording the 2.8 1-bit bitstream output Sony could produce a digital version of a master tape without committing to a specific sample rate and word length and so they could keep smaller digital files.
DSD (PDM) failed miserably in this last regard due to the large quantity of noise needed in a DSD file.
Additionally, while Sony and Phillips had their hands deep in the CD technology licensing business Sony had no financial ties to the DVD market so they hoped to use PDM as a multi-channel format for the DVD market.
Sony and Philips took DSD to the DVD Forum. The forum had issued a “request for proposal” in the late 90s to all technology companies and research institutions looking for the next audio encoding scheme to move consumer audio to high-resolution and multichannel. The DVD Forum considered higher resolution PCM (96 and 192 kHz and 24-bits) and Sony/Philip’s DSD 64 (2.8224 MHz 1-bit PCM). After extensive evaluation and independent investigations done by experts in the field, the forum determined that PCM met the standards of their RFP and DSD did not. DSD was rejected because of its inherent and substantial high-frequency noise, the lack of available production tools, and the lack of production hardware.
Let's also clear up this false notion that DSD (PDM) is more "analog like".
Above 25kHz the quantization noise rises sharply, requiring far more sophisticated filters and/or noise-shaping algorithms. When filtered, the output of DSD64 with a simple low-pass filter, the result is distorted phase/time and some rather nasty artifacts in the audible range.
There is no need to add filter at 20kHz, we can't hear the noise. It is -120db real world in audible frequency. Now it is near -150dB in DSD512, of course PCM can do the same at 386kHz. Both DSD and PCM consume about the same binary data. Practically don't need any filter at all. Because of PWM nature of SDM DAC, DSD has it edge over PCM w/o addition conversion. Your phone sound chipset is already SDM AD/DA. Your phone is constant covert any PCM to DSD like data for the converter. And lots of phone chip is DSD like data direct drive to class D like amplification. PWM and PDM propagation in the same matter, DSD and Class D amp work seamlessly.There is one audiophile loudspeaker manufacture employ direct DSD driven to Class D amp active loudspeaker. Not like PCM->PDM/PWM->Filter->Analog stage-> Class A, A/B,D, H,G amplification as standard chain.
Jon, thank you for this information. It took some time. It makes perfect sense the DVD forum rejected DSD for exactly the reasons you state. However, not among them is how much better it sounds, at least to these ears and those of people we trust.
@@kongwee1978
While human hearing (on average by the way, not definitively) may be limited to 20 kHz, audio equipment is not.
Ultrasound noise can cause audible non-linear distortions in playback on an audio system therefore a low pass filter is absolutely necessary.
Additionally, The noise used to quantize DSD reaches as low as 15 Khz, well within the audible spectrum, and while the signal level may be as low as -150 dB at 15 kHz when added to an audible signal the signal is altered, and distortion is created.
The noise level increases significantly as the frequency gets higher and is as high as -36 dB in the upper spectrum.
When multiple DSD signals are combined, such as in multitrack recording, the noise level doubles for each signal added.
DSD files are considerably larger than PCM files because of the need to record the noise in the upper spectrum to maintain dithering.
PCM dithering noise is in the "audible spectrum" but well below the threshold of hearing, especially at higher bit rates, and therefore there is no need to record or place ultrasonic information in the file.
Your example of the phone is correct, though it should be noted that the phone does not use DSD for file storage or use.
The phone converts PCM to DSD at the output stage so what you are hearing is a DSD representation of a PCM digital signal.
PWM and PDM propagation are not the same.
PWM has a clearly distinguished period, PDM does not have a clearly distinguishable period.
Pulse density modulation can be adjusted at any clock cycle without any extra work; unlike pulse width modulation, which can only be modified effectively once per period.
@@JonAnderhub DSD noise is -120db w/o filer at audibe frequency. Whatever bits above 22kHz, it not a concern until you want to put a filter in to get rid of aliasing. Way back before DSD. 1 bit SDM is already at propagate at 2.8224mhz on oversampling PCM. Indirectly you are using PCM like PDM/PWM. PDM and PWM are about the same. That why they can use directly at Class D. DSD is PDM because it need time interval is needed. Orelse you can't track your songs. Class D amp is PWM as they don't care about the time interval. PDM signal is really PWM for Class D amp. 0-2.8Mhz frequency, Class D amp won't care. They don't need to sync with clock frequency. If you read the Class D amp, they don't have clock frequency at amplifying stage. They only understand on and off state.
ahhh the tired, old and wrong ...anti DSD marketing hype....."DSD was rejected because of its inherent and substantial high-frequency noise" - this is utter and pure BS.(show me your audible "noise" hearing capabilities at north of 25Khz ...also now that DSD is at 4x (256). your claim "By recording the 2.8 1-bit bitstream output Sony could produce a digital version of a master tape without committing to a specific sample rate" is false -- 2.8Mhz IS the sample rate...just like 2X DSD is 5.6Mhz and 4x DSD is 11.2 Mhz .....this guy knows nothing of modern DSD recording and the emerging multi bit DSD future!
Good luck yall... its a smart business move in a way since there are no longer any real players in the DSD game anymore. Hopefully you can get some really desirable releases, bc at the end of the day it doesnt matter how much marketing speak you can lay on for the audiophiles or how good it sounds... its all about the songs and catalogue. Get yourself A&R perhaps ...
You own the most expensive microphones in the world...and your desk looks like it's being held together with bubblegum, lol.
Sorry Paul, I think rrecording studios should go back to the old analog recording consoles , like that Studer board
Another tape and vinyl limit: wow & flutter.