equivalent PCM rates for converting DSD files?

[bidn]

Dear all,

I have some big, space-wasting DSD files which I want to convert to PCM (and then erase), they have various rates: mostly DSD64, but also some with DSD128 or DSD256.
Even if anything beyond Red Book cannot be heard, I would prefer to minimize resolution loss.
I can't find answers on the internet on the most appropriate recommended conversion PCM rates.
e.g. should DSD64 be converted into 24 bits-96 kHz and DSD128 into 24 bits-192 kHz ?
I have no idea...

I would be very thankful if those knowledgeable about this could help.
Already thank you, a merry Christmas and happy new year in advance,
bidn

 

[sarumbear]

Dear all,

I have some big, space-wasting DSD files which I want to convert to PCM (and then erase), they have various rates: mostly DSD64, but also some with DSD128 or DSD256.
Even if anything beyond Red Book cannot be heard, I would prefer to minimize resolution loss.
I can't find answers on the internet on the most appropriate recommended conversion PCM rates.
e.g. should DSD64 be converted into 24 bits-96 kHz and DSD128 into 24 bits-192 kHz ?
I have no idea...

I would be very thankful if those knowledgeable about this could help.
Already thank you, a merry Christmas and happy new year in advance,
bidn

Search the forum for @amirm’s high resolution track analysis. It makes clear that anything more than 20-bit at 48kHz resolution signal is available on any of those recordings.

 

[Kal Rubinson]

It makes clear that anything more than 20-bit at 48kHz resolution signal is available on any of those recordings.

Anything or nothing? Either is moot.

 

[sarumbear]

Anything or nothing? Either is moot.

I don’t get your meaning.

 

[DVDdoug]

I don't understand the math but I believe mathematically-perfect conversion is not possible.

I think the equivalent bit-depth depends on frequency so you'd have very high "bit accuracy" at low frequencies and less at high frequencies.

I found this:

Though you can’t make a direct comparison between the resolution of DSD and PCM, various experts have tried. One estimate is that a 1-bit 2.8224MHz DSD64 SACD has similar resolution to a 20-bit 96KHz PCM. Another estimate is that a 1-bit 2.8224MHz DSD64 SACD is equal to 20-bit 141.12KHz PCM or 24-bit 117.6KHz PCM.

In other words a DSD64 SACD has much higher resolution than a 16-bit 44.1KHz Red Book CD, roughly the same resolution as 24-bit 88.2KHz PCM recording, and not as much resolution as a 24-bit 176.4KHz PCM recording.

(I didn't read the whole thing.)

 

[Kal Rubinson]

It makes clear that anything more than 20-bit at 48kHz resolution signal is available on any of those recordings.

I don’t get your meaning.

I was not sure of your meaning. Does your sentence mean that there is "more than 20-bit at 48kHz resolution signal is available on any of those recordings" and that you are advising the OP to convert to PCM at DSD64 into "24 bits-96 kHz and DSD128 into 24 bits-192 kHz?" If so, i would have expected the word "something" instead of "anything." If you meant the opposite, I would have expected "nothing."

 

[mansr]

I don't understand the math but I believe mathematically-perfect conversion is not possible.

I think the equivalent bit-depth depends on frequency so you'd have very high "bit accuracy" at low frequencies and less at high frequencies.

24-bit PCM will perfectly capture the content of a DSD signal up to the Nyquist frequency. Now music doesn't contain anything above 30 kHz (and certainly not above the DSD noise), and even if it did, few speakers would reproduce it, and even then, nobody would be able to hear it. This means 96 kHz, 24-bit is more than you'll ever need regardless of the DSD rate.

 

[sarumbear]

I was not sure of your meaning. Does your sentence mean that there is "more than 20-bit at 48kHz resolution signal is available on any of those recordings" and that you are advising the OP to convert to PCM at DSD64 into "24 bits-96 kHz and DSD128 into 24 bits-192 kHz?" If so, i would have expected the word "something" instead of "anything." If you meant the opposite, I would have expected "nothing."

I am sorry for my unclear sentence. I should have said there’s hardly anything in the recordings that is worth to convert the DSD stream higher than 20 bits at 48kHz PCM stream.

 

[escksu]

24-bit PCM will perfectly capture the content of a DSD signal up to the Nyquist frequency. Now music doesn't contain anything above 30 kHz (and certainly not above the DSD noise), and even if it did, few speakers would reproduce it, and even then, nobody would be able to hear it. This means 96 kHz, 24-bit is more than you'll ever need regardless of the DSD rate.

Higher frequencies do have an effect on wave form of lower frequencies. What you are hearing are not individual frequencies but resultant of all the frequencies put together (think of it like fourior transformation of all the frequencies).

 

[escksu]

Dear all,

I have some big, space-wasting DSD files which I want to convert to PCM (and then erase), they have various rates: mostly DSD64, but also some with DSD128 or DSD256.
Even if anything beyond Red Book cannot be heard, I would prefer to minimize resolution loss.
I can't find answers on the internet on the most appropriate recommended conversion PCM rates.
e.g. should DSD64 be converted into 24 bits-96 kHz and DSD128 into 24 bits-192 kHz ?
I have no idea...

I would be very thankful if those knowledgeable about this could help.
Already thank you, a merry Christmas and happy new year in advance,
bidn

Why not simply convert them into various resolutoin/samping rate and listen for yourself? IF you cant hear any difference, then it makes no differnce to you.

Btw, I am more of surprise that you say the files are big.... With such ample storage space today, audio files hardly take up space.

 

[Blumlein 88]

Well if you look at performance for SNR, and bandwidth then DSD 64 is more or less like PCM 88.2 20bits. So 88.2 khz 24 bit is what you'd want.

DSD128 still has ultrasonic noise levels rise due to noise shaping. There are various implementations. Even those don't have low SNR at much more than 40 khz. So again 88.2 or 96 24 would do.

So just to preserve the signal of the recordings 88.2 or 96 khz rates are fully sufficient in the audio band and beyond for DSD64, and nearly fully sufficient for even DSD128. So simple answer is just do 96/24 and don't worry about it anymore.

 

[radix]

Personally, I think 24-bit 48 kHz is plenty. If you want to splurge, 96 kHz. 192 kHz I think is a waste. My understanding is some studios use 96 kHz so they have headroom to filter down to 48 kHz.

What do you listen to? If it was anything back in the AAD or ADD days, it was recorded in a low-rez format. Upscaling it is mostly a gimmick, I think. Even stuff recorded in all digital likely was not recorded above 96 kHz, but that's a guess on my part.

 

[Blumlein 88]

Personally, I think 24-bit 48 kHz is plenty. If you want to splurge, 96 kHz. 192 kHz I think is a waste. My understanding is some studios use 96 kHz so they have headroom to filter down to 48 kHz.

What do you listen to? If it was anything back in the AAD or ADD days, it was recorded in a low-rez format. Upscaling it is mostly a gimmick, I think. Even stuff recorded in all digital likely was not recorded above 96 kHz, but that's a guess on my part.

I actually agree with that. However, if your goal was to have a technical performance envelope that more or less equals or slightly exceeds DSD you are going to need at least 88.2 khz rates. That way the signal output of your converted PCM files can pretty much equal the signal output of the DSD format.

 

[Zoomer]

If space is your main concern, why not compress with wavpack instead of converting?

Compressing DSD files with Wavpack 5

Last updated 05/03/2024 If you are reading this you already know that DSD files (DSF and/or DFF) have rather big sizes and large collections of these require plenty disk space. Although the cost of…

diyaudioheaven.wordpress.com

 

[kchap]

If space is your main concern, why not compress with wavpack instead of converting?

Compressing DSD files with Wavpack 5

Last updated 05/03/2024 If you are reading this you already know that DSD files (DSF and/or DFF) have rather big sizes and large collections of these require plenty disk space. Although the cost of…

diyaudioheaven.wordpress.com

It's not just that DSD is a waste of bandwidth, it's the total inconvenience of DSD. Active speakers, digital volume controls, room equalizers etc become a pain with DSD. It should be banned.

Sorry, do I go too far?

 

[mansr]

Higher frequencies do have an effect on wave form of lower frequencies. What you are hearing are not individual frequencies but resultant of all the frequencies put together (think of it like fourior transformation of all the frequencies).

Uh-huh.

 

[ZolaIII]

It's CD rate/16 bit. Example 64/16= 4 that's 176400 (or 88200 per channel) Hz at 16 bit PCM. Take a look what WavPack can do as a container (even more regarding PCM 24 bit over 48 KHz) in lossy hybrid mode with log) but keep in mind you will need software support for it (FFmpeg, Foobar2000...).

 

[escksu]

Uh-huh.

Uh-huh. Uh-huh.

 

[escksu]

It's not just that DSD is a waste of bandwidth, it's the total inconvenience of DSD. Active speakers, digital volume controls, room equalizers etc become a pain with DSD. It should be banned.

Sorry, do I go too far?

No... You have every right to think and believe in what you want.

 

[kchap]

No... You have every right to think and believe in what you want.

Thanks. There was a little bit a smile on my face when I wrote the last line but, as you can tell I'm not a fan of DSD