Should DSD audio have been supported? POLL

[Mr. Afternoon]

For music not DSD recorded, then you are back to pcm at the beginning.
If starting with pcm, it has to be upsampled, then converted to dff. Then for mch music it has to be losslessy compressed to DST.
The beginning, with pcm, is not 24/96, or even 24/48. it's 16/44 IIRC. Then it's upsampled and converted to DSD.

The goal (I suppose) of Sony/Phillips was to put this format on a standard size CD; i.e. the possibility of 80 minutes of music on one disc.

Most music is not recorded in DSD, and quite a few titles that's been mixed in DSD have gone through a PCM passthrough stage. However the source PCM before DSD conversion can be any resolution, and the DSD will preserve as much as it can. That's why some high-end labels works in 352.8kHz.
A single layer SACD is 4.7 GB/4.3GiB. Same as a DVD.

 

[mandrix]

Yes, I was only referring to goal of playback time, as I understand it: not that it would actually be pressed to a CD.

 

[mandrix]

I should have been more clear and stated at the base level, 16/44 can be converted to DSD.
Higher resolutions can indeed be incorporated if they are already constructed with the proper sample rate.
But in the past that has definitely not always been the case.

 

[mandrix]

Do you think, for example the Japanese SACD pressing of Machinehead used native sample rates of 352kHz? Well no, because they did not exist. So to convert to DSD the music had to be resampled..

 

[KG10]

You know what I can actually encode DSD32 files let me do some tests and let you know.

May I ask what software/setup your using to let you encode those unusual audio types?

 

[omnipotent1]

DSD 512 seems to be king amongst the niche high end classical audiophile labels and that's good enough for me
Not that I can actually hear the difference but I trust those that can and they say that there are differences.
I for one don't listen on a professionally calibrated mega buck reference system in a state of the art treated listening studio so it matters not to me
But I do care that it was recorded/mastered at the highest possible quality before it gets to me on a wine and beer enthused weekend blowout

 

[KG10]

DSD 512 seems to be king amongst the niche high end classical audiophile labels and that's good enough for me
Not that I can actually hear the difference but I trust those that can and they say that there are differences.
I for one don't listen on a professionally calibrated mega buck reference system in a state of the art treated listening studio so it matters not to me
But I do care that it was recorded/mastered at the highest possible quality before it gets to me on a wine and beer enthused weekend blowout

Really wish Mofi would start releasing digitally in DSD256 instead of just SACD and Vinyl.

 

[esimms86]

Most music is not recorded in DSD, and quite a few titles that's been mixed in DSD have gone through a PCM passthrough stage. However the source PCM before DSD conversion can be any resolution, and the DSD will preserve as much as it can. That's why some high-end labels works in 352.8kHz.
A single layer SACD is 4.7 GB/4.3GiB. Same as a DVD.

The Sound Liaison label works in 352.8 DXD, as does 2L. Blue Coast Records records in DSD. Both are niche labels. Then there is the HQ Player software which does DSD oversampling and the PS Audio Directstream DAC that converts everything to DSD and then does DSD over sampling prior to playback. PS Audio also produces and sells SACDs. Blue Coast and PS Audio both work with DSD sound engineering legend Gus Skinas. Finally, NativeDSD sells downloads that range from DSD64 to DSD256, some of which are multichannel. If you purchase a recording in DSD256 they will also let you download the same recording at DSD128 and DSD64 at no extra charge.

 

[Sal1950]

The Sound Liaison label works in 352.8 DXD, as does 2L. Blue Coast Records records in DSD. Both are niche labels.

They have to use some "niche" formats as an excuse to charge niche prices.
Nothing they do that couldn't be exactly sonically equaled by 24/96. PCM
If the sold discs sound verifiably different, the masterings did too.

 

[esimms86]

They have to use some "niche" formats as an excuse to charge niche prices.
Nothing they do that couldn't be exactly sonically equaled by 24/96. PCM
If the sold discs sound verifiably different, the masterings did too.

I'm not going to speculate on their motivations but I do know that (1) Blue Coast's Cookie Marenco is highly respected in the industry and is a firm believer in DSD, and (2) both Sound Liaison and Blue Coast Records have all of their music also available in PCM format, including at 24/96. Their artists are not exactly household names so I'm fairly certain think that their sales don't garner them a large amount of their production investment back (at least, compared to major pop or classical labels).

 

[Sal1950]

I'm not going to speculate on their motivations but I do know that (1) Blue Coast's Cookie Marenco is highly respected in the industry

As is our own member Mark Waldrep of AIX records. He's done thousands wonderful recordings in both 2 and multich formats, along with doing a lot of honest and verifiable testing of the various options.
Take his Audio Challenge and read the things he's written.
https://www.realhd-audio.com/?p=6713

 

[esimms86]

As is our own member Mark Waldrep of AIX records. He's done thousands wonderful recordings in both 2 and multich formats, along with doing a lot of honest and verifiable testing of the various options.
Take his Audio Challenge and read the things he's written.
https://www.realhd-audio.com/?p=6713

Yes, I have read his blog posts, own many of his AIX recordings and have actually spoken with him once. He's a big believer in the unassailability of PCM in 24/96. I respect his work, as I do Ms. Marenco's and I'm not the least bit surprised to see experts disagree with one another. Barry Diament of Soundkeeper Recordings has said that PCM 24/192 is indistinguishable from the microphone feed, so there's another expert opinion. I only take issue with vendors that upsample, are not transparent about their use of upsampling, and then sell the upsample at a higher price. Beyond that, I do appreciate when a vendor like HDTracks offers a download in both 24/96 and 24/192 resolution. This allows the buyer to opt for the 24/96 version if they don't see any reason in going higher, and, conversely, it also allows the buyer who wants to feel secure in the knowledge that s/he has everything at the highest resolution to opt for 24/192.

Edit:

Mr. Diament is a mastering engineer who is reknown for not using compression. His remaster of the first Led Zeppelin album in CD format is legendary among audiophiles who seek out the best version of that album(I own a copy). Here is a link to his discography as posted by a fan on the discogs site. I'm certain that you'll recognize at least a few of these :

https://www.discogs.com/lists/Mastered-by-Barry-Diament/306170?page=1

 

[par4ken]

I only take issue with vendors that upsample, are not transparent about their use of upsampling, and then sell the upsample at a higher price.

You can see that if you open the file in an audio editing program. The spectral display will show black (blank) above the original cut-off.

 

[jimfisheye]

They have to use some "niche" formats as an excuse to charge niche prices.
Nothing they do that couldn't be exactly sonically equaled by 24/96. PCM
If the sold discs sound verifiably different, the masterings did too.

Indeed.

And extraordinary claims beg extraordinary evidence!
We have all the examples of commercially released masters with identical masters put to both formats. How do we know they were identical? They null with each other down to the noise floor. That doesn't happen by accident or coincidence! Only audio that was identical but sample rate converted nulls down to the noise floor. ie. null down to around -100db. The audio band is never touched. You can't produce something that sounds different to the ear but magically has an identical data set.

Someone would need to create a recording that shows the PCM converters suddenly mysteriously outputting corrupt signal in an ABX test.

The goal is to deliver audio without degradation or generation loss. We want the listener to hear what we intended. Fussing over A/B tests of stuff within your perception bias (if even different at all)... So what changed in the sound? Pull the lead vocal bus fader down 0.5db on your mix. Now THAT you could hear right away! (While a few not as attentive listeners in the room still didn't catch even that.)

This is how my thought process goes anyway. Do I hear an actual change that isn't inside my perception bias? Do I hear any obvious generational artifacts? When that's "no" and "no" and a null test is showing me null down to -100db... Well, I'm done here! Y'all can buy DSD DACs and burn SACDs if it seems like a beter idea.

Had I invested in DSD DA converters instead of PCM I suppose I might be interested in transcoding the other way and staying native DSD. It would be a similar sideways move to replace everything with PCM and not worth it. But PCM was and is the main digital format. As a recording engineer needing DAW software and so forth, that's mostly all only PCM. The odd DSD transcode that comes up is still transparent and all is well.

 

[jimfisheye]

FYI, anything with reasonable converters at 24 bit is transparent to the mic feed. Not as reasonable converters run at 96k bypass that dodgy low pass filter and are transparent to the mic feed.

This is the single most groundbreaking change with digital recording!
Tape hiss meant that you couldn't record a raw mic feed. You had to produce it on the fly as you recorded. The recording was as good as you were able to pull that off! And 2" 24 track tape was around $80 a minute.

Now you can record all the raw mic feeds and do absolutely anything you please with them after. And recording is basically free now with hard drive space being effectively unlimited. Roll 100s of hours of recording if you want. And then the copies are clones and not generational. THAT's the stuff to celebrate! Not Sony showing their balls with a new digital language to use for copy protection purposes.

 

[kfbkfb]

https://web.archive.org/web/2007101...tre.org/surround2002/technology/page_07.shtml
^^^
...PCM produces a much more faithful copy of the original...


Currently, I listen to SACD MCH DSD dowmixed to Dolby Surround -> coax digital in -> (original) Dolby Pro-logic, probably too much signal processing to be able to make any sort of determination on DSD sound accuracy.


Kirk Bayne

 

[ssully]

DSD was designed to be converted to PCM for consumer releases, at any multiple of 44.1 kHz.

It should have stayed that way. But the industry freaked out about copy protection, so SACD was born.

SACD was eventually hacked anyway.

DSD should go extinct as a consumer format. Certainly not taking up space on a BluRay set.

 

[KG10]

DSD was designed to be converted to PCM for consumer releases, at any multiple of 44.1 kHz.

It should have stayed that way. But the industry freaked out about copy protection, so SACD was born.

SACD was eventually hacked anyway.

DSD should go extinct as a consumer format. Certainly not taking up space on a BluRay set.

Ironic how most Hi res releases are multiples of 48khz

 

[Mr. Afternoon]

DSD was designed to be converted to PCM for consumer releases, at any multiple of 44.1 kHz.

It should have stayed that way. But the industry freaked out about copy protection, so SACD was born.

SACD was eventually hacked anyway.

DSD should go extinct as a consumer format. Certainly not taking up space on a BluRay set.

Actually the SACD was born as a retaliatory move from Sony over disc royalty agreements of the DVD and DVD-Audio.

 

[esimms86]

Sometimes the consumer doesn't have a choice. I'm referring to SACD's from the likes of Dutton Vocalion, Audio Fidelity, Mobile Fidelity, etc. Some SACDs, of course, have a hybrid layer that allows a standard CD player to output the recording in PCM. And then, of course, you convert the digital file from DSD to PCM using various software options, such as foobar.

As an aside, record companies would prefer not to sell you a high resolution download as they feel that they are effectively selling you a master copy. You can then duplicate(and pirate) this without degradation from one copy to another.