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

[Miska]

17 minutes ago, Zaphod Beeblebrox said:

If you process the whole album as one then you are assuming all the tracks are truly continuous in time which is rarely true and we don't even know if it is an album or a playlist with unrelated material.

 

One reason why HQPlayer makes a difference between album and playlist. Most albums are gapless these days. If you are reading a CD you can actually know if it's a case or not. But I have only very few CDs where it is not the case.

 

19 minutes ago, Zaphod Beeblebrox said:

Then you would have to accept that you are reconstructing your new samples borrowing samples from unrelated material when you are near the end of a track or beginning of the next one, that could not be possibly good either.

 

This is what MScaler does too, since it cannot know where a track boundary is. Or where someone seeked or skipped a track unless sampling rate changes. This is advantage of player applications since they can have this information.

 

21 minutes ago, Zaphod Beeblebrox said:

I have tried both approaches, and settled on processing on a per track basis based on both subjective listening tests and a over all better user experience.

 

Yeah, then the trade-off is that you need to make the track filter length amount of longer than it originally was to avoid truncating transients near beginning and end. Longer the filter, longer that goes.

 

22 minutes ago, Zaphod Beeblebrox said:

This thread wont exists if everyone is happy to just settle for  'typical' filter 😏

 

Yes, for me, the interesting thing is to have as few taps as possible to minimize amount of ringing and transient spread and still have optimal frequency domain behavior too without leakages. IOW, optimize both time and frequency domain at the same time, although the have 1/x relationship.

 

[Zaphod Beeblebrox]

2 minutes ago, Miska said:

Yes, for me, the interesting thing is to have as few taps as possible to minimize amount of ringing and transient spread and still have optimal frequency domain behavior too without leakages. IOW, optimize both time and frequency domain at the same time, although the have 1/x relationship.

Different strokes for different folks. Our approaches are different, I may not necessarily agree that one approach is better than the other, but do appreciate you chiming in. 

[happybob]

11 hours ago, chrille said:

Hi I agree regarding macrodynamics, very loud big dynamic peaks are not what electrostats thrive at.

And yes I  too used to own a pair of huge HORN coffins which could play much louder than my current electrostats can. But like all horns I have heard including the truly big  Avantgarde Trios,they are just  a bit too colored for my "acoustic music only please" ears.

One of the best speakers I have heard with really  large scale acoustic music where the Gryphon Pendragons,but they are way beyond both my wallet and room capacities.

My compromise  electrostats both economically and dynamically sound quite nice with PGGB and I can play my PGGB  test tracks a bit louder than normal hi res.

And at 1 metre 85 centimetres, taller than me I can stand at my listening position and conduct along without having to be a "couch potato" all day.

Cheers Chrille

 

 

I've found that the best "traditional driver" systems excel at not just macro dynamics but also micro dynamics. For me the panel speakers don't to either quite as well as the best "traditional" speakers I've heard, but panels are still great. As to horns, I think they really excel over almost anything I've ever heard in the realm of dynamics, but they don't need to be big - the best ones I've heard are actually quite small (and often have separate bass units of some sort). Big horns are wonderful, but mostly only work well IMO in very very large rooms where we sit at quite a distance from the drivers.

 

There is another issue with large panels - and that is (with some exceptions like Quad ESLs) that they basically provide a wave of sound but that means that different portions of the large panel's sound arrives at our ear at different times. This does create a huge and often very desirable sound, but does give up some accuracy in imaging. So certain music will work better than others with panels. 

 

Always tradeoffs! 

[happybob]

@Fourlegs - this is my new challenge too - that of comparing 32 bits vs 24 bits and noise shaping on or off as it goes through the MScaler. And then comparing this to the SRC-DX (which requires 24 bit). I'd talked to ZB earlier about this issue and he initially thought that 32 bit with noiseshaping by PGGB was not optimal if going through the MScaler, but now considering this "no noiseshaping and 32 bit" option for PGGB files is intriguing. 

 

One issue I have with these variants though is that as I process more and more music (with quite a lot left to do), I'd rather not have different (each large) files for different ways that I do my playback. i.e. don't want to need a different version for MScaler playback vs SRC-DX vs perhaps some other future option that goes to other systems I have (not Chord DAC based). But maybe these variants are unavoidable for the optimal sound... 

[Zaphod Beeblebrox]

4 hours ago, happybob said:

@Fourlegs - this is my new challenge too - that of comparing 32 bits vs 24 bits and noise shaping on or off as it goes through the MScaler. And then comparing this to the SRC-DX (which requires 24 bit). I'd talked to ZB earlier about this issue and he initially thought that 32 bit with noiseshaping by PGGB was not optimal if going through the MScaler, but now considering this "no noiseshaping and 32 bit" option for PGGB files is intriguing. 

 

One issue I have with these variants though is that as I process more and more music (with quite a lot left to do), I'd rather not have different (each large) files for different ways that I do my playback. i.e. don't want to need a different version for MScaler playback vs SRC-DX vs perhaps some other future option that goes to other systems I have (not Chord DAC based). But maybe these variants are unavoidable for the optimal sound... 

Ideally, you are able to test the combination that works best for you and then remaster your library in that format. However, it will be hard to predict what the future holds, so we need to take comfort in the fact that whatever causes you to remaster your library in a different format is something  that improves over what we have now and I think that is a good thing.

 

If you had asked me a couple of months back if I would ever use 24bits into DAVE, I would have said absolutely not, yet here I am still remastering my library at 24bits to work with SRC-DX a, nd for me the difference is not subtle and is totally worth it and I get to hear my library all over again.

 

With what you have right now, do a simple comparison. 32bit without any noise shaping direct to Mscaler vs 32bit noise shaped in to DAVE's USB (same as @hanshopf). Pick what you prefer and remaster your library. If SRC-DX arrives, compare it to noise shaped to 24bit  in SRC-DX. Maintaining two versions is not needed as it will still take double the time and double the space.

 

 

[Gavin1977]

PGGB-RT + Leedh volume algorithm?

[Zaphod Beeblebrox]

15 minutes ago, Gavin1977 said:

PGGB-RT + Leedh volume algorithm?

PGGB-SRC + PGGB-EQ + PGGB-NS = PGGB-RT

 

PGGB-RT incorporates three independent modules:

• PGGB-SRC: Samplerate conversion
• PGGB-EQ:  Convolution engine
• PGGB-NS: PGGB real time digital volume control with noise shaping.

 

PGGB-EQ can be placed before or after SRC but needs to be before PGGB-NS

 

PGGB-NS is designed to be a transparent real-time digital volume control with noise shaping, so the gain can be adjusted interactively during playback. It uses the same noise shaping algorithms as PGGB desktop, so guarantees the same low noise floor in audible range. 

 

This is still a SDK, it is based on C++, so light weight compared to PGGB desktop. But it is not a music player, it is meant for OEM/licensing and is available for PC, Mac and Arch Linux.

[Fourlegs]

7 hours ago, happybob said:

@Fourlegs - this is my new challenge too - that of comparing 32 bits vs 24 bits and noise shaping on or off as it goes through the MScaler. And then comparing this to the SRC-DX (which requires 24 bit). I'd talked to ZB earlier about this issue and he initially thought that 32 bit with noiseshaping by PGGB was not optimal if going through the MScaler, but now considering this "no noiseshaping and 32 bit" option for PGGB files is intriguing. 

 

One issue I have with these variants though is that as I process more and more music (with quite a lot left to do), I'd rather not have different (each large) files for different ways that I do my playback. i.e. don't want to need a different version for MScaler playback vs SRC-DX vs perhaps some other future option that goes to other systems I have (not Chord DAC based). But maybe these variants are unavoidable for the optimal sound... 

I will try to accept the challenge. When I have settled on my favoured playback/processing method it is my intention to just use that for my PGGB library.

 

Live streaming and other albums that did not get processed (I only have 12TB local storage 🤣) will be played with HQPlayer or possibly the Mscaler depending on which streamer I am using (Antipodes K50 for HQP and Innuos Zenith for HMS).

[happybob]

53 minutes ago, Fourlegs said:

I will try to accept the challenge. When I have settled on my favoured playback/processing method it is my intention to just use that for my PGGB library.

 

Live streaming and other albums that did not get processed (I only have 12TB local storage 🤣) will be played with HQPlayer or possibly the Mscaler depending on which streamer I am using (Antipodes K50 for HQP and Innuos Zenith for HMS).

This is a tough challenge we have!

 

In my case, I have 3 separate systems that I'll want to play content on. Currently everything is sourced from Roon, but I'm going to head eventually to a highend server setup like what you have (not sure, maybe Innuos, maybe Taiko, maybe something else), but even then I'll need Roon for some of my "multizone" listening modes. I generally in the evenings have music playing on 3 systems simultaneously so it gives a "fill the house with sound" effect - quite nice for dinner music, etc. But 2 of my systems (one Chord DC4 Dave/HMS with my Abyss headphones and HPA, and another living room system currently with a Devialet integrated amp/DAC but I'm going to upgrade that soon to a TBD DAC, maybe a Holo May or Tambaqui or dCS or maybe another Dave) are for "highest quality" listening. The problem here is that one version of PGGB tracks is not necessarily optimal for every system - and maybe not even playable on every system. The Devialet can only handle up to 192K/24 input and so I currently have Roon downsample the PGGB processed files to 192K, but if I get a dedicated server, it might not be able to do that on the fly....

 

I hope SSD drives come down in price quickly! I don't even have 12 TB😀.

 

Such first world problems we have!

[HeeBroG]

For those with multiple server playback options; is PGGB a leveller or do high-end servers still come out on top?

[Fourlegs]

9 minutes ago, HeeBroG said:

For those with multiple server playback options; is PGGB a leveller or do high-end servers still come out on top?

 

If I just say that many of the guys who helped develop PGGB have Taiko Extremes and I haven’t noticed any of them selling those for anything cheaper afterwards!!

 

[davide256]

Has anyone used dual BNC with the Hugo 2? The analog section is supposedly better with it than the Qutest but using the mini coax to dual BNC adaptor

raises a concern that the BNC connection would degrade any advantage gained in the analog section over Qutest. Mojo also states its supports

768 " special operation" on mini coax  but there is no content out there to support that it works/ is in use by anyone.

[Progisus]

4 hours ago, davide256 said:

Has anyone used dual BNC with the Hugo 2? The analog section is supposedly better with it than the Qutest but using the mini coax to dual BNC adaptor

raises a concern that the BNC connection would degrade any advantage gained in the analog section over Qutest. Mojo also states its supports

768 " special operation" on mini coax  but there is no content out there to support that it works/ is in use by anyone.

H2 with dual coax cable works splendidly with PGGB. I gave used 16fs 32b and also 16fs 24b with an SRC-DX. I use a Moon Audio dual bnc to h2 adapter cable. I used to use this setup with mscaler but that is looong gone. 

[jkelly]

Quick question on Hyperthreading.

I am using one of my older Flight Simulator computers with a overclocked 4790k, water cooled and 32G ram.

I have gotten better performance flying with Hyperthreading off.  What does PGGB like? TY!

[Zaphod Beeblebrox]

2 hours ago, jkelly said:

Quick question on Hyperthreading.

I am using one of my older Flight Simulator computers with a overclocked 4790k, water cooled and 32G ram.

I have gotten better performance flying with Hyperthreading off.  What does PGGB like? TY!

PGGB does not care if you have HT on or off, as long as you set the number of workers to the physical cores (which is the default).  But, yes the performance will degrade if you use virtual cores with PGGB too as it also increases the memory pressure. Overclocking definitely helps with performance and you will get even more performance if you can bump your RAM up by 32G more. I have a silent workstation, and it will shutdown due to overheating if I overclock with PGGB running, especially when gargle-blasting DSD.

 

[romaz]

16 hours ago, HeeBroG said:

For those with multiple server playback options; is PGGB a leveller or do high-end servers still come out on top?

 

If anything, I'm finding PGGB widens the gap among the best servers and the lower end servers and this is what you might expect with files that are more transparent, they scale to equipment that can better reveal their qualities.

 

I'm currently testing a prototype server from a certain manufacturer and so part of my testing process has been to compare this server against other servers including a basic NUC, an Innuos Zen Mk3, and the Extreme.  Playing standard Redbook files, the differences among the servers are easily apparent but with PGGB files, the superiority of the Extreme becomes much more apparent.  

 

Where PGGB has the potential to be a leveler is with DACs, especially DACs incapable of high sample rates.  For example, I found the $1,500 Gustard X26 Pro + PGGB to outperform (at least to my ears) another much more expensive DAC without PGGB.

 

PGGB (especially PGGB EQ) has the potential to also be a leveler with amplifiers.  I am still seriously astounded by how much better my bass has gotten with PGGB EQ.  I feel like I have seriously upgraded my amplifiers.

[ted_b]

31 minutes ago, romaz said:

If anything, I'm finding PGGB widens the gap among the best servers and the lower end servers and this is what you might expect with files that are more transparent, they scale to equipment that can better reveal their qualities.

+1  I'm finding the Holo May to be quite up to task on discerning PGGB files from their standard PCM counterparts, even upsampled.  It's all part of the double edged sword of transparency, you get all the warts too, much more discernible.  And this with a run of the mill Audiophile-optimized Win10 server but nowhere near state of the art. 

[ASRMichael]

I know I’ve already talked about PGGB + EQ but I feel compelled to write about it again & encourage people to try. Hopefully I manage to do this and other users hear what I’m & others are hearing. 
 

Also to point out my TAD DAC is 8fs. Yet I’m still hearing all the changes are hearing. Maybe it’s not just what DAC you have, could be AMP & Speakers also. Maybe it’s everything that matters! We know that already I suppose. 
 

I’m going out on a limb here. Out off all the upgrades I’ve made recently PGGB + EQ has the most profound affect. Then you take into account the cost. It’s an absolute no brainer. What upgrade have you done the has such a profound affect on transparency, Bass. Separation, depth, image etc etc. I’m not talking about having to sit down and have long listening sessions to find these things. It’s just there from the get go. Instant hit. 
 

Another case of what you put in is what you get out! 
 

Taiko ATX + new cables

Unregulated SJ LPS. 
PGGB+EQ

 

I really encourage people to find tune your room frequency. Mitch at Acourate sound does amazing job. Then use PGGB EQ. It will enhance everything. 
 

I’m pretty sure Nenon’s V2 or V3 will take Taiko to another level. 
 

 

[jkelly]

Can someone suggest a starting bit rate for the original Holo Spring Dac (L3) ?

I have read from 19 to 23 bit.

[Zaphod Beeblebrox]

1 hour ago, jkelly said:

Can someone suggest a starting bit rate for the original Holo Spring Dac (L3) ?

I have read from 19 to 23 bit.

I suggest starting with 19

[Fourlegs]

On 7/3/2021 at 6:25 AM, romaz said:

I'm currently testing a prototype server from a certain manufacturer and so part of my testing process has been to compare this server against other servers including a basic NUC, an Innuos Zen Mk3, and the Extreme. 

  Playing standard Redbook files, the differences among the servers are easily apparent but with PGGB files, the superiority of the Extreme becomes much more apparent. 

Can you tell me how you are playing 705/768 PGGB files on the Zen? I don’t know of any way to do this with its current software. However when Innuos 2.0 is released for the Zen that will play 705/768 PGGB files (and sounds very good).

[romaz]

2 hours ago, Fourlegs said:

Can you tell me how you are playing 705/768 PGGB files on the Zen? I don’t know of any way to do this with its current software. However when Innuos 2.0 is released for the Zen that will play 705/768 PGGB files (and sounds very good).

 

I never specified sample rates.  This prototype server is making the rounds to different systems and not all servers, players, and DACs being tested are capable of 16fs.

[romaz]

On 7/2/2021 at 11:00 PM, ted_b said:

It's all part of the double edged sword of transparency, you get all the warts too, much more discernible. 

 

I have some music that I enjoy that is better left as mp3s and listened to on my car's audio system.  Sometimes, transparency can be a bad thing, lol.

[ASRMichael]

On 7/4/2021 at 10:49 AM, romaz said:

 

I have some music that I enjoy that is better left as mp3s and listened to on my car's audio system.  Sometimes, transparency can be a bad thing, lol.

That raises an interesting question. If we hear things that the artist/producer didn’t hear! Are we now listening to something other than the artist wanted us to hear? 
 

If I was patient enough I’m sure my library/individual song would have different PGGB settings, and some maybe have EQ & some may not! Glad I don’t have the patience for that! 😂

[romaz]

8 hours ago, ASRMichael said:

That raises an interesting question. If we hear things that the artist/producer didn’t hear! Are we now listening to something other than the artist wanted us to hear? 

 

In my case, this has nothing to do with hearing things that the artist/producer didn't hear.  Some performances / recordings / masterings are just so bad that they do not sound good in a transparent and resolving system by today's standards.  You actually want a veil to make then listenable.  Growing up in the 80s, I enjoyed bands like Duran Duran, The Cure, The English Beat, Depeche Mode, The Smiths, China Crisis, etc.  Yes, the 2nd British invasion.  I saw these bands live on many occasions and I enjoy listening to them even today but only in my car.  Back then, for many of us, our audio systems were slow, not as resolving, treble was rolled off, etc.  Perhaps those albums were mixed for systems like that.  If I play that type of music in my system now, they sound thin and strident.  Certainly, all of my system building efforts are not for music like that.

What's interesting is that with my main system at home, the large majority of what I listen to (at least 75%) is unamplified acoustical music (chamber, choral, symphonic).  Among my favorite instruments is the organ.  These are the live concerts I most enjoy attending today and what I endeavor to recreate to scale in my main listening room.  This type of music does not not translate nearly as well in my car, in the shower, or on my Sonos system in the kitchen.  I don't even bother because I find very little engagement and so when I'm not listening in my main system, I listen to almost everything else.  This means that over the course of a day, since I'm often listening to music in the car and at work, the music I most enjoy listening to is the music I intentionally listen to the least.  It's quite the irony.