A toast to PGGB, a heady brew of math and magic

[Miska]

1 hour ago, littlej0e said:

I would only caution against using HQPlayer upscaling to evaluate DSD and use native/converted files instead.

 

Why? HQPlayer is not holding back any algorithms or making any shortcuts because it is performed in realtime. I have never restricted any algorithms because of something happening in realtime, I just tell you to buy fast enough computer.

 

Lot of the native DSD content you can buy online has been mastered with HQPlayer Pro (which is a professional mixing / mastering and a file converter tool), which has exactly same algorithms but can process offline to files.

 

[Zaphod Beeblebrox]

2 hours ago, littlej0e said:

I discovered that DACs w/true native DSD are very few and far between. In addition to the other DACs mentioned, I believe the Wandala also has native DSD capability, but only up to DSD 256 IIRC. My personal recommendation would be the T+A 200 or the Holo May KTE. I've owned both and prefer the T+A for the extra utility and slightly more engaging sound, but I believe the May is a bit cheaper. YMMV. 

You can add APL and some AKM DACs, there is a good list here: 

 

[littlej0e]

1 hour ago, Miska said:

 

Why? HQPlayer is not holding back any algorithms or making any shortcuts because it is performed in realtime. I have never restricted any algorithms because of something happening in realtime, I just tell you to buy fast enough computer.

 

Lot of the native DSD content you can buy online has been mastered with HQPlayer Pro (which is a professional mixing / mastering and a file converter tool), which has exactly same algorithms but can process offline to files.

 

Not necessarily because of anything HQPlayer does or doesn't do under-the-hood. Rather it can make the evaluation process a bit more prone to user error (see stupid users doing stupid things). For example, I spent a good week comparing DSD 64, 128, 256, and 512 before deciding on 128. I then converted my entire library, but soon discovered the converted files sounded very different than what I heard during the evaluation process. I eventually realized why: I didn't have "DirectSDM" enabled in settings, meaning I was evaluating files with HQPlayer filters, shapers, etc. enabled. Because stupid.
 

Again, not really an HQPlayer problem and more of a "you must have xx number of chromosomes to operate this software" type of thing. Also, from a simplicity and system utilization perspective, testing with converted files as opposed to in-line upscaled files would likely be more representative of the final performance. 

TBH, your post was bombshell for me personally as I never knew HQPlayer Pro does offline file conversion, similar to dBpoweramp, and the upcoming PGGB DSD. Shows what I know. 

[littlej0e]

On 3/19/2024 at 1:21 PM, Zaphod Beeblebrox said:

That is a great question. PGGB DSD is built on the same principles as PGGB PCM which you have indicated you like:

1. (Algorithm) Maximizing reconstruction accuracy using all of the information in the source track and upsampling to the required output rate: In the case of PCM the maximum output rate was up to 64fS. In the case of DSD, the maximum output rate can be 128fS, 256fS, 512fS or 1024fS. So, the upsampling algorithms remains the same as before, just that the output rate for DSD is a lot higher.
2. (Precision) Maximizing reconstruction accuracy using high precision: PGGB PCM supports 64/128/256 bit precision as higher precision helps improve reconstruction accuracy in the audible range. The reconstruction accuracy in the audible range is only limited by bit depth and sample rate. For PCM, for most bit depths higher than 16 bits and sample rates higher than 8fS, 256bit precision helps to deliver the highest reconstruction accuracy in the audible range. With DSD, things are not as clear cut. Since DSD has a bit depth of 1, the amount of information that can be retained in the audible range depends on the DSD rate and also the modulator. Higher order modulators help retain more information in the audible range by reducing the quantization noise, higher DSD rates also provide a larger bandwidth to push the quantization noise out. So, the choice of precision becomes a function of the output sample rate and also the modulator order. There is no point in using higher precision for reconstruction if that information cannot be retained. Which is why for DSD, PGGB DSD uses adaptive precision. From what I have analyzed, 128bit precision works best for DSD rates. 256bit precision is an overkill other than for DSD1024 using 9th order modulator.
3. (Modulators) Minimizing the quantization noise in audible range: In the case of PGGB PCM, I use noise shapers to keep the quantization noise in the audible range as low as possible.  The noise floor is limited by the precision used for computing and also the quality of the noise shaper. For 64bit precision, the limit is about -320dB. If a noise shaper is capable of achieving a quantization noise floor that is lower than -320dB for the given bit depth and sample rate, then it makes to go for higher precision which is 128bit. The limit for noise floor for 128bit (quad precision) is about -690dB. If this noise floor can be breached, only then does 256bit computing makes sense. For PCM, the noise shapers I use can achieve a quantization noise floor as low as -750dB in the audible range, which is why I provide the option of 256bit precision. For PGGB DSD, for DSD 256 and higher, the -320dB limit for 64bit computing is easily breached and for DSD rates 512 and 1024, the quantization noise in the audible range is -400dB or lower with 9th order modulators, which justifies the use of 128bit precision.

There is a cheesy cake analogy I like to give, (forgive me in advance):

1. Algorithm: How good the reconstruction algorithms are, i.e. how well you can predict the intermediate samples. This determines the quality of the cake we baked.
2. Precision: How fine/precise you can get in reconstructing. If you start with a bad reconstruction algorithm, it does not really matter how precise the flawed intermediate samples are! This is like the icing on the cake. If the cake is bad, a good icing does no good!
3. Noise shaping/Modulators: This is the container where you store the cake, it decides how much of that icing you get to keep. If the container is too small, you end up mushing the cake or losing the icing. Too big a container may make the cake wobbly (i.e., there is a limit to how aggressive the noise shaper can be) 

Edit: Since not all DACs/Players support 48k DSD rates, I have also added an inter-rate conversion option, i.e. it will be possible to convert a 48kHz/96kHz/293kHz source track to DSD rates taht are multiples of 44.1kHz. I have tried to do this in a transparent fashion, the Jury is still out on if it is truly transparent compared to 48kHz DSD rates of the same tracks. PCM conversions will also benifit form this for some DACs that are capped at a maximum 44.1kHz rate.

Thanks so much for the detailed and comprehensive answer (as well as the outstanding cake analogy)! The main thing I noticed when converting PCM to DSD with the other software is how much of the noise, harmonics and tone seem to shift depending on DSD 64, 128, etc., It is fascinating to me how this can result in a "better" listening experience without necessarily adding or subtracting any musical information (pedantically, I suppose you could technically be adding and subtracting info in some circumstances, but you get my point).
 

So glad there are folks like you who understand these things at such a deep level and make turnkey solutions for people like me. Speaking of which, PM Inc...

[Zaphod Beeblebrox]

43 minutes ago, littlej0e said:

It is fascinating to me how this can result in a "better" listening experience without necessarily adding or subtracting any musical information (pedantically, I suppose you could technically be adding and subtracting info in some circumstances, but you get my point).

 

You are right, it is not possible to add new information to the musical content present in the source track, all you can do is process it carefully and not throw away information. Continuing the cake analogy, consider the source track as the ingredients needed to bake the cake. Better quality ingredients do help improve the quality of the baked cake, but you will have to work with what you have got! The hope is the upsampler makes it easy for the DAC to bake a better cake.

[Zaphod Beeblebrox]

Just a PSA as I got some enquires and wanted to avoid confusions:

 

Going forward, starting with v6, I plan to simplify and do away with the intermediate PGGB128 version because it looks like the middle child nobody wants, most of current license holders are using PGGB 256.

 

When I release PGGB v6 that will include DSD, there will only be two tiers (the actual names I use may change):

1. PGGB 64: Will allow you to do PCM upsampling to 64fS rates at a 64bit precision and DSD upsampling to DSD256 at an adaptive rate. 
2. PGGB 256: Will allow you to do PCM upsampling to 64fS rates at up to 256bit precision and DSD upsampling up to DSD1024 at an adaptive rate. 

 

Note: Those with current PGGB 128 licenses will be automatically upgraded to v6 (PGGB 128), they will retain PCM 128bit precision and will be able to do DSD rates up to DSD256. We will continue to support it, but no new licenses will be issued.

 

Choosing higher precision or higher DSD rates will still work but will be limited to two tracks at a time.

[genefruit]

1 hour ago, Zaphod Beeblebrox said:

Note: Those with current PGGB 128 licenses will be automatically upgraded to v6 (PGGB 128), they will retain PCM 128bit precision and will be able to do DSD rates up to DSD256. We will continue to support it, but no new licenses will be issued.

Do you know if that will be the same for PGGB-IT?  If so, I need to rush off to buy the upgrade.  To me, the middle tier could provide the best balance for DSD levels. Nevermind, I was recalling this which must have changed:

 

• PGGB 128bit: Will be available on Windows both as PGGB 128 from remastero and as PGGB-IT! 128bit from Audiowise. Mac version will be available as PGGB 128 through remastero. PGGB 128bit will support DSD128, DSD256 and DSD512 with up to 9th order modulators.

[Zaphod Beeblebrox]

8 minutes ago, genefruit said:

Do you know if that will be the same for PGGB-IT?  If so, I need to rush off to buy the upgrade.  To me, the middle tier could provide the best balance for DSD levels.

Yes, it will be the same for PGGB-IT!, just to be clear, both PGGB 64 and PGGB 128 will be able to do DSD256 at the same quality with PGGB 128 being able to do PCM at 128bit precision while PGGB 64 stops at 64bit precision for PCM. 

 

If your interest is DSD, I don't see any advantage in going with the middle tier unless you also need 128bit PCM for perhaps additional processing by the other software.

[Zaphod Beeblebrox]

13 minutes ago, genefruit said:

PGGB 128bit: Will be available on Windows both as PGGB 128 from remastero and as PGGB-IT! 128bit from Audiowise. Mac version will be available as PGGB 128 through remastero. PGGB 128bit will support DSD128, DSD256 and DSD512 with up to 9th order modulators.

That was an error on my part, and I lost the ability to fix it and hence the new post as I got a few enquires such as yours. Sorry again for the confusion.

[genefruit]

8 minutes ago, Zaphod Beeblebrox said:

Yes, it will be the same for PGGB-IT!, just to be clear, both PGGB 64 and PGGB 128 will be able to do DSD256 at the same quality with PGGB 128 being able to do PCM at 128bit precision while PGGB 64 stops at 64bit precision for PCM. 

 

If your interest is DSD, I don't see any advantage in going with the middle tier unless you also need 128bit PCM for perhaps additional processing by the other software.

Keeping files below 4gb for flac and DSD512 is the limit of Roon native playback, so these constraints are part of my decision making process.  While I can and do use HQPlayer for playback, any files that are outside of Roon's ability to library make them non-negotiable for me.  Until I either do the math or some file size examples are provided, I'm in a holding pattern.

[Zaphod Beeblebrox]

21 minutes ago, genefruit said:

Keeping files below 4gb for flac and DSD512 is the limit of Roon native playback, so these constraints are part of my decision making process.  While I can and do use HQPlayer for playback, any files that are outside of Roon's ability to library make them non-negotiable for me.

The 4GB limit is really not a Roon limit but the wav format, PGGB-IT! supports 64bit wav header that should allow Roon to play them even if they are larger than 4GB. Also, PGGB will split large Wav files to slightly smaller than 4GB and renumber the track so that Roon will play them in correct order.

 

DSD has no such length limits, Roon should play them fine even if they exceed 4GB.  Roon however will not read DSD files if their base rate is 48kHz and not 44.1khz, a bug Roon has not fixed, and you can vote for the feature. As a way to circumvent this issue, PGGB DSD will allow conversion from 48kHz rates to DSD rates that is a multiple of 44.1khz. The process is transparent, but I do not have subjective feedback on it yet.

 

None of the above has anything to do with precision, precision is only used for computing, the size of output tracks will be the same for a given rate and bit depth (be it DSD or PCM) irrespective of the precision you choose. Precision applies only for processing, noise shapers (for PCM) or modulators (for DSD) are responsible to create the final output at the requested bit depth (16, 24, 32 bits etc., for PCM) and 1 bit for DSD.

 

I hope that helps. 

 

Edit: I agree about the DSD512 and 768kHz rate limit of Roon, another feature we have been asking for years! But I see the lack of support for 48 x DSD rates even more troubling!

 

[genefruit]

Very helpful.  I'm looking forward to trying the release.  If someone has already used PGGB to convert a number of files from their native rate to 16fs/24bit to use those converted files as a basis for DSD conversion?  Or should one resort to the native rate files and process from there?

[Zaphod Beeblebrox]

3 minutes ago, genefruit said:

.  If someone has already used PGGB to convert a number of files from their native rate to 16fs/24bit to use those converted files as a basis for DSD conversion?  Or should one resort to the native rate files and process from there?

There may be a very small advantage in terms of saved time. However, the higher rate input track will require a lot more memory and will negate any benefits.

[Atriya]

22 hours ago, Zaphod Beeblebrox said:

Edit: I agree about the DSD512 and 768kHz rate limit of Roon, another feature we have been asking for years! But I see the lack of support for 48 x DSD rates even more troubling!

 

On this note, many of us have been asking for various things from Roon - including things as basic as supporting the latest version of the FLAC format! - all without success. In fact, they seem antagonistic and dismissive of anything cutting edge, experimental, or niche. The whole voting system they have is a joke. They seem to care very little about technological advancement in digital audio, and seem to be just software development people who happen to work in digital audio streaming.

[Atriya]

@Zaphod Beeblebrox: An excerpt from the guide:

 

"Dither (Set ‘Dither Only’ in Configuration): No noise shaping is done, only a Gaussian dither is applied, not recommended for output rates lower than 352.8kHz."

 

So, you would recommend noise shaping for output rates lower than 8fS? I thought noise shaping was only effective at higher sample rates?

 

With some further close listening, I realized I cannot really tell between 8fS and 4fS upsampling for my Weiss DAC (which converts 8fS -> 4fS internally). I always feel like 8fS sounds a little better, but then I fail an A/B test.

 

If I just upsample to 4fS to save space then, the above quote seems to rule out using "dither only"? The Weiss will further convert from 192Khz -> 195Khz, so it is doing some further processing whether I like it or not.

[Zaphod Beeblebrox]

8 hours ago, Atriya said:

@Zaphod Beeblebrox: An excerpt from the guide:

 

"Dither (Set ‘Dither Only’ in Configuration): No noise shaping is done, only a Gaussian dither is applied, not recommended for output rates lower than 352.8kHz."

 

So, you would recommend noise shaping for output rates lower than 8fS? I thought noise shaping was only effective at higher sample rates?

 

With some further close listening, I realized I cannot really tell between 8fS and 4fS upsampling for my Weiss DAC (which converts 8fS -> 4fS internally). I always feel like 8fS sounds a little better, but then I fail an A/B test.

 

If I just upsample to 4fS to save space then, the above quote seems to rule out using "dither only"? The Weiss will further convert from 192Khz -> 195Khz, so it is doing some further processing whether I like it or not.

That was a typo, it should have said:  "not recommended for output rates higher than 352.8kHz."

[Atriya]

2 hours ago, Zaphod Beeblebrox said:

That was a typo, it should have said:  "not recommended for output rates higher than 352.8kHz."

 

That's what I thought!

[Zaphod Beeblebrox]

@genefruitI saw a discussion on a different forum (and someone was responding to you and making references to long filters) regarding what PGGB does. In spite of me having repeated this multiple times here and elsewhere. Currently PGGB does not use filters or taps in the same sense as Chord, it is not a longer version of Chord's filters. Any discussion along these lines is simply not true hence any conclusions reached from the discussion along those lines are also incorrect and misleading. 

 

The approach currently used by PGGB is completely different. PGGB looks at the whole track, it keeps the original samples intact and unchanged (assuming no HF filtering was engaged, which are by default off for CD and 2fS rates). Then PGGB reconstructs the intermediate samples all at once by time shifting the original samples. It takes the same approach be it upsampling to 16fS PCM or DSD1024.

Edit: For clarity

[genefruit]

16 minutes ago, Zaphod Beeblebrox said:

@genefruitI saw a discussion on a different forum regarding what PGGB does. In spite of me having repeated this multiple times here and elsewhere. Currently PGGB does not use filters or taps in the same sense as Chord, it is not a longer version of Chord's filters. Any discussion along these lines is simply not true hence any conclusions reached from the discussion along those lines are also incorrect and misleading. 

 

The approach currently used by PGGB is completely different. PGGB looks at the whole track, it keeps the original samples intact and unchanged (assuming no HF filtering was engaged, which are by default off for CD and 2fS rates). Then PGGB reconstructs the intermediate samples all at once by time shifting the original samples. It takes the same approach be it upsampling to 16fS PCM or DSD1024.

Thank you for the explanation.  It was not my intent to misrepresent PGGB.  I'm not sure which forum, as I believe I've mentioned PGGB in two locations other than here.  I don't believe in either case I stated that PGGB doesn't look at the whole track.  I'm suspecting it was a quick reference to/comparison with Chord.  In the future, I'll simply reference that people may be interested in PGGB and provide this link - https://www.remastero.com/pggb.html

[Zaphod Beeblebrox]

10 minutes ago, genefruit said:

Thank you for the explanation.  It was not my intent to misrepresent PGGB.  I'm not sure which forum, as I believe I've mentioned PGGB in two locations other than here.  I don't believe in either case I stated that PGGB doesn't look at the whole track.  I'm suspecting it was a quick reference to/comparison with Chord.  In the future, I'll simply reference that people may be interested in PGGB and provide this link - https://www.remastero.com/pggb.html

@genefruit I edited my original post, someone was responding to your question, since I am unable to respond on that headphone forum (due to restrictive polices requiring to be paid sponsor), I was responding here. Apologies if it was a coincidence (in the user handle), and I certainly was not suggesting you were making the reference.

[genefruit]

12 minutes ago, Zaphod Beeblebrox said:

@genefruit I edited my original post, someone was responding to your question, since I am unable to respond on that headphone forum (due to restrictive polices requiring to be paid sponsor), I was responding here. Apologies if it was a coincidence (in the user handle), and I certainly was not suggesting you were making the reference.

Ah, now I have context.  Same user handle and same person. 

[Zaphod Beeblebrox]

6 minutes ago, genefruit said:

Ah, now I have context.  Same user handle and same person. 

Regarding the context, when using another upsampler or playback software, if you do not plan to do any additional processing such as EQ, it is best to put it in pass through mode, turn off all filtering and dither and set the output rate to match the PGGB rate, this is true for PCM. 

 

When PGGB-DSD is released, all DSP such as volume control should be turned off and if there is an option for 'Direct SDM' or 'Native DSD' or similar, then that should be turned on. If the DAC supports 48k DSD rates and if there is an option in the playback software to check, then 48k DSD should be checked too.

 

One of the main reasons for me developing PGGB DSD is to eliminate the need for a hybrid approach which I think is burdensome on the user and significantly complicates playback. 16fS PCM @32bits is the same size as DSD512, so there is very little need to do that if the constraint is space.

[chelseafc]

I have tested PGGB samples with the DAVE and they compared favorably to both the M Scaler and HQPlayer, so I am considering purchasing a license.

 

Few newbie questions (forgive me if they've been asked many times before)

 

1. Any reason to buy PGGB-IT! vs. the original distribution from Remastero?

I tried both and found only one annoying limitation with Remastero (would refuse to generate FLACs on 16Fs and x256)

 

2. Any consideration for the OS (Mac OS vs Windows)?

 

3. How much I need to spend for a Mac/PC that would do PGGB256 album in under 30 min?

[kennyb123]

5 hours ago, genefruit said:

Ah, now I have context.

Always bad form to bash a competitor’s product, but what makes it even worse is that he violated the terms of service on that site as well.  (“A Member of the Trade may not post subjective assessments or negative comments about the products/services of competitors or other Members of the Trade.”).  

 

6 hours ago, Zaphod Beeblebrox said:

I saw a discussion on a different forum (and someone was responding to you and making references to long filters) regarding what PGGB does. In spite of me having repeated this multiple times here and elsewhere. Currently PGGB does not use filters or taps in the same sense as Chord, it is not a longer version of Chord's filters. Any discussion along these lines is simply not true hence any conclusions reached from the discussion along those lines are also incorrect and misleading.

 

Not cool, but he did provide a great example showing the use of a straw man fallacy (setting up a dummy version of an opponent’s stance just to knock it down).

[Zaphod Beeblebrox]

40 minutes ago, chelseafc said:

I have tested PGGB samples with the DAVE and they compared favorably to both the M Scaler and HQPlayer, so I am considering purchasing a license.

Thanks for your interest and feedback.

 

40 minutes ago, chelseafc said:

Few newbie questions (forgive me if they've been asked many times before)

 

1. Any reason to buy PGGB-IT! vs. the original distribution from Remastero?

I tried both and found only one annoying limitation with Remastero (would refuse to generate FLACs on 16Fs and x256)

It is really up to you to see what fits your needs best. Audiowise licenses PGGB and the processing is the same whether you get it from remastero or Audiowise, as you found the difference are really in how the interface and output formats differ. PGGB-IT! has some additional bells and whistles such as multi-channel processing, allowing multiple tracks to be processed (if there is enough RAM and CPU cores).

You can move from PGGB-IT! to PGGB from remastero at no cost anytime at the same level of license. Moving from PGGB from remastero to PGGB-IT! will incur a small fee. That may make the decision easier for you.

 

47 minutes ago, chelseafc said:

2. Any consideration for the OS (Mac OS vs Windows)?

PGGB-IT! is Windows only, so if you plan to use Mac, the choice is remastero. But if you are open to both Mac and Windows, I strongly suggest Windows as Windows is more flexible (compared to Mac) when it comes to memory allocation. Windows will let you get the most out of your Hardware in terms of processing speed. If you were to go with Mac, PGGB does not run natively on Apple Silicon, so will require Rosetta and is less efficient running on emulation. 

From performance: Windows > Mac (Intel) > Mac (M1/M2/M3)

 

53 minutes ago, chelseafc said:

3. How much I need to spend for a Mac/PC that would do PGGB256 album in under 30 min?

It is better to think in terms of 'per minute of track' because albums can be of varying lengths. Then there are additional factors:

1. What precision do you plan to use?
2. What is the typical sample rate of the albums (CD/Hires/DSD)?
3. What is the typical length of tracks?

Windows will work cheaper most of the time, a used Mac Pro running Intel may also work.