• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

DSD is it of any value

luft262

Senior Member
Joined
Mar 25, 2021
Messages
465
Likes
236
Location
Phoenix
I think it's as simple as this...

For the end consumer DSD had an advantage over CD at the time because the bitrate was higher than Redbook CD. Now that most people are switching over to streaming music, Hi-Res PCM usually has an equivalent or higher bitrate than SACD DSD did in the past. Either format would work fine, but since all consumers are using DACs/Devices that support PCM and only some support DSD the streaming services are unlikely to make a move in that direction. For the consumer it doesn't matter what format is used, either can produce excellent sonic fidelity. To me DSD is only relevant if a consumer is playing local files ripped from a CD/SACD or playing those discs directly. In that case the SACD is likely to have a higher sonic fidelity due to having a higher bitrate. Due to convenience, and a lack of local music files, it's irrelevant to me because I'm going to stream all of my music anyway and Hi-Res FLAC/PCM is going to be the best playback available to me and in generally it will exceed SACD DSD not because PCM is inherently better, but because a higher bitrate was used in the production of that file and it's also what is available to me from streaming services.
 

mocenigo

Major Contributor
Joined
Dec 8, 2018
Messages
1,274
Likes
1,034
It is not just about the bitrate: DSD is very inefficient at information storage, despite the compression. So at the same bitrate, PCM is going to hold more information.

Where DSD wins is if you build a true single bit “chipless” DAC. Just a flip flop, an output filter, a buffer. Then you potentially get all the advantages in terms of sound quality — if subjective reports are to be believed.
 

ZolaIII

Major Contributor
Joined
Jul 28, 2019
Messages
4,071
Likes
2,410
I really don't follow people talking nonsense around hire. Bit is bit no matter if it's written in word or as a stream. DSD 64 bit stream is equal regarding data amount as PCM 176400 Hz 16 bit. DSD as a format is only relevant for music labels because it provides them a good protection. DSD can be compressed as any other data. However PDM stream is great for transport protocol to DAC or from ADC in DSD or any other form because simplicity and efficiency. DSD is not editable without converting and it's stiff (regarding it's layout) and proprietary format. I think another transport protocol based on PDM is needed which will be more flexible and open source but I don't mean repackaged DSD like DoP but fundamentally different regarding unfolding/processing.
 
Last edited:

mocenigo

Major Contributor
Joined
Dec 8, 2018
Messages
1,274
Likes
1,034
I really don't follow people talking nonsense around hire. Bit is bit no matter if it's written in word or as a stream. DSD 64 bit stream is equal regarding data amount as PCM 176400 Hz 16 bit.

Yes, same bit rate and therefore in theory they could contain the same amount of entropy.
BUT.
It wastes a lot of this with high frequency quantisation noise, so a good amount of that is useless - and in fact the output filter discards it.
In fact, DSD64 is considered by some to be equivalent to PCM 96/24, and it is clearly an exaggeration. I would rather say PCM 48/24.
 
Last edited:

mansr

Major Contributor
Joined
Oct 5, 2018
Messages
4,685
Likes
10,700
Location
Hampshire
Yes, it contains the same amount of entropy.
Same number of bits, different amount of information/entropy.

It wastes a lot of this with high frequency quantisation noise, so a good amount of that is useless
Right. DSD has higher entropy, making it hard to compress. High-rate PCM has a lot of empty "space" which compresses well.
 

mocenigo

Major Contributor
Joined
Dec 8, 2018
Messages
1,274
Likes
1,034
Same number of bits, different amount of information/entropy.

You are right / what I meant to say was about the entropy space, i.e. as I corrected above, "Yes, same bit rate and therefore in theory they could contain the same amount of entropy." But as you point out:

DSD has higher entropy, making it hard to compress. High-rate PCM has a lot of empty "space" which compresses well.
 

ZolaIII

Major Contributor
Joined
Jul 28, 2019
Messages
4,071
Likes
2,410
Same number of bits, different amount of information/entropy.


Right. DSD has higher entropy, making it hard to compress. High-rate PCM has a lot of empty "space" which compresses well.
Did a test with 7zip. DSD took longer to compress and archive whose 6~12% bigger than wave one. Used 96/24 masters, converted to 176400/16 wave and from there to DSD. Still compression ratio didn't feel under 50% even for DSD. Still when it comes to transport and simplicity DSD is a right choice, for storage certainly not.

Edit: Interesting part is after converting I let JRiver analyse tracks and results ware as expected DSD had the -6 dB to original DR (used good masters with 12~13 DR without normalisation) whose the same however after normalisation DSD one's had a slightly higher DR (0.1).
If anyone could explain this last part with DR slight increase after normalisation for DSD track's I would love to hear/learn about it.
 
Last edited:

julian_hughes

Addicted to Fun and Learning
Joined
Aug 23, 2020
Messages
657
Likes
901
I have quite a lot of hybrid SACDs (the format is very common with classical releases). I thought this one in particular sounded much better in DSD than when played as CD or as 16-bit 44.1 file:
cover (copy 1).jpg


It really does seem more alive, more real, just better! Anyway, out of curiosity I extracted the two channel DSD as a single Edit Master file from the disc (yes, it can be done) and converted it to 16-bit 44.1 kHz flac using ffmpeg with options
Code:
-af volume=4dB,aresample=resampler=soxr:precision=28:dither_method=lipshitz:dither_scale=0.5:cutoff=1 -ar 44100 -sample_fmt s16
*

The resulting CD quality audio file sounds *identical* to the DSD! It also measures the same when checked vs a 24-bit 88.2 kHz version with dr14 tool, and the waveform looks very nearly identical too. Subjectively, the CD layer of the disc is mastered to play a little louder than the DSD layer but I'm not sure if there are any other differences. What I do know is that side by side I find the DSD and my conversion impossible to distinguish.

*I arrived at this set of parameters by reading the fine manual (more than once :oops:), looking up how other people had approached the same task, and then by some trial and error. This conversion doesn't necessarily give a result from DSD that matches the CD layer of the same disc, but it does always result in a Red Book compliant file that cannot be audibly distinguished from that DSD layer....which is supposedly better!
 

Mnyb

Major Contributor
Forum Donor
Joined
Aug 14, 2019
Messages
2,645
Likes
3,627
Location
Sweden, Västerås
I have quite a lot of hybrid SACDs (the format is very common with classical releases). I thought this one in particular sounded much better in DSD than when played as CD or as 16-bit 44.1 file:
View attachment 125784

It really does seem more alive, more real, just better! Anyway, out of curiosity I extracted the two channel DSD as a single Edit Master file from the disc (yes, it can be done) and converted it to 16-bit 44.1 kHz flac using ffmpeg with options
Code:
-af volume=4dB,aresample=resampler=soxr:precision=28:dither_method=lipshitz:dither_scale=0.5:cutoff=1 -ar 44100 -sample_fmt s16
*

The resulting CD quality audio file sounds *identical* to the DSD! It also measures the same when checked vs a 24-bit 88.2 kHz version with dr14 tool, and the waveform looks very nearly identical too. Subjectively, the CD layer of the disc is mastered to play a little louder than the DSD layer but I'm not sure if there are any other differences. What I do know is that side by side I find the DSD and my conversion impossible to distinguish.

*I arrived at this set of parameters by reading the fine manual (more than once :oops:), looking up how other people had approached the same task, and then by some trial and error. This conversion doesn't necessarily give a result from DSD that matches the CD layer of the same disc, but it does always result in a Red Book compliant file that cannot be audibly distinguished from that DSD layer....which is supposedly better!

Yes the CD layer is usually not the same master , because how to sell DSD then if they sound the same as you discovered ;)
 

julian_hughes

Addicted to Fun and Learning
Joined
Aug 23, 2020
Messages
657
Likes
901
It could perhaps be the same master but with the CD layer and DSD layers playing at slightly different output level by default. As is typical with SACD, the disc includes both multi channel and two channel DSD as well as the two channel Red Book, so who knows? Clearly the multi channel audio is different from either of the two channel offerings and does not sound the same. There are things you can get with SACD that you can't get with Red Book. There is no separate DSD edition, it's a hybrid disc that plays on any standards compliant CD player and is priced just like any other CD so nobody is trying to pull any dastardly sales tricks.
 

Mnyb

Major Contributor
Forum Donor
Joined
Aug 14, 2019
Messages
2,645
Likes
3,627
Location
Sweden, Västerås
Regarding compression of 24bit files may disappoint due that many of the bites are random due to noise as most bit below 13-14 bits are random noise in many recordings.
 

Mnyb

Major Contributor
Forum Donor
Joined
Aug 14, 2019
Messages
2,645
Likes
3,627
Location
Sweden, Västerås
All the good ones are multi-level sigma-delta.
Is it not sometimes 5 level sigma delta some internal format that does not at all look like any consumer or pro format just something suitable for the inner workings of the chip ?
 

mansr

Major Contributor
Joined
Oct 5, 2018
Messages
4,685
Likes
10,700
Location
Hampshire
Is it not sometimes 5 level sigma delta some internal format that does not at all look like any consumer or pro format just something suitable for the inner workings of the chip ?
Yes, you would not normally store data in that encoding, if you even have access to it. I'm aware of only a few chips that expose these signals externally.
 

mansr

Major Contributor
Joined
Oct 5, 2018
Messages
4,685
Likes
10,700
Location
Hampshire
I do not claim to know all DACs. Apparently you can ;-)
The top performers from AKM, Cirrus, ESS, and TI are all multi-level designs. Achieving comparable performance with a single-bit design would be very difficult if not impossible.
 
Top Bottom