PGGB Offline Remastering

Dec 5, 2022 at 8:57 PM Post #106 of 346
I don't have the newest PGGB software so I can't comment on higher bitrate (128, etc) conversion.
It's not higher bitrate, it's higher precision. The noise floor is reduced because it’s a more accurate reconstruction. The reason I brought it up was that higher precision provides great insight into system transparency. I wonder if that transparency would be retained with the additional processing you perform on these files.
 
Dec 5, 2022 at 11:19 PM Post #107 of 346
I'm afraid that's a question you can answer if you have both the newer PGGb software and HQPlayer Pro. And yes, I stand corrected, it is higher precision indeed. I'm not sure what happened here b/c I got blue death when using HQPlayer the other day and since then, the software just won't work anymore. I can launch it, but that's it. Asked around at Audiophilestyle but no responses yet. Long story short, I can't make more comparisons until I figure out what had happened to HQPlayer.
 
Dec 7, 2022 at 12:49 AM Post #108 of 346
I’ve gone back to 64b precision. The processing time and file size negated the sound improvement in my system.
In case you missed it, the PGGB developer said this on Audiophile Style: “The PGGB-256 version I am currently will do all the computation at 256 bits, but is much faster and also has much reduced memory requirements.”
 
Dec 7, 2022 at 7:10 AM Post #109 of 346
In case you missed it, the PGGB developer said this on Audiophile Style: “The PGGB-256 version I am currently will do all the computation at 256 bits, but is much faster and also has much reduced memory requirements.”
Thanks, I did see that and I look forward to the upgrade.
 
Dec 27, 2022 at 4:43 PM Post #111 of 346
Just processed an album at 128 insane setting and made the album 40GB bigger.
I have never compared sizes after changing precision but this doesn’t sound right to me. Are you sure that the only change was precision?
 
Dec 27, 2022 at 5:21 PM Post #113 of 346
I should have said compared to the Redbook album.
Yes of course the upscaled files will be larger. High res files are always larger.
 
Jan 16, 2023 at 2:05 PM Post #115 of 346
Yay, I can finally talk about PGGB 256. Let me start by saying that it 256 bit precision brings with it quite a bump in sound quality. So much so I felt it worth the effort to reprocess everything. It’s a lot faster, so reprocessing at P256 took a fraction of the tIme than it took to reprocess at P192 with PGGP AP. I understanding that processing at P64 is faster too but I’ve not tried it.

PGGP 256 is much better at memory handling. There were albums I would have never been able to process with PGGP AP at higher precisions because PGGP would have displayed an out of memory message and crashed (or it would have just locked up my Mac. Those days are over providing that your computer has at least the minimum required memory. I’m still using a 2010 Mac Pro with 128 GB of RAM - and there is nothing I can’t process with this machine. Likewise for the 2019 iMac Core i9 with 128 GB that I also use.
 
Jan 16, 2023 at 5:38 PM Post #116 of 346
Just upgraded myself. There is definitely a sound improvement especially at 256. My iMac 2020 with 40 gb ram has allowed the 256 processing but I did receive the warning. The speed is incredible. I had one album - Vangelis Juno to Jupitor that tracks 13 and 14 would not play through hqplayer no matter how I processed them. Now they play even at 256. I’m not sure I will reprocess everything but it is a long winter.
 
Jan 17, 2023 at 8:49 AM Post #117 of 346
Thanks chaps for the updates - I look forward to getting the update. Even with a hefty Windows machine there have been tracks that at 192bit with eq convolution applied would just fail out with resource errors after many hours. Even when successfully processing, some whole albums have taken 3 or 4 days to complete. Hopefully those days are going to be consigned to history.
 
Jan 17, 2023 at 11:49 AM Post #118 of 346
Thanks chaps for the updates - I look forward to getting the update. Even with a hefty Windows machine there have been tracks that at 192bit with eq convolution applied would just fail out with resource errors after many hours. Even when successfully processing, some whole albums have taken 3 or 4 days to complete. Hopefully those days are going to be consigned to history.

For sure, give it a whirl. I think you'll be pleased. I was on the verge of ordering a hefty server, with a hefty price tag, when ZB suggested we could solve it in software. Now, my 128GB W10 desktop can tackle even my torture tracks, like 30+ min DSD and DXD files.
 
Jan 20, 2023 at 2:55 PM Post #119 of 346
FA0DEC5B-0DD0-449E-8788-77A8FE470506.png


Glorious in 256b processing!
 
Jan 20, 2023 at 6:11 PM Post #120 of 346
I just gave two well-recorded jazz tracks a careful listen with the new PGGB compared to their original "1FS" counterparts (here's one of them).

I like it (I mean, I still like it!). The three or four most noticeable things to me are:
  • Leading edges are more "rounded" and nuanced. Less "square-wave" kind of a feeling. I like this very much.
  • Treble is less splashy (obviously related to point 1), and smoother. Upper-mids, too.
  • Instrument images are a shade more substantial / defined / solid.
  • Bass in particular is more well defined.
This was using:
  • Bartok > AHB2 > Utopia 2022 and
  • Bartok > HA-300 tube amp > Verite Closed.
The specific DAC being used matters a ton, of course. Whether the Bartok molests the signal less when fed at its full 352/384k bitrate versus 44/48k is a mystery, made even more so by the so-called "Ring DAC" architecture and the novel way it generates its analog signal from the final bitstream internally.

Regardless, I prefer the PGGB versions of tracks while using the Bartok ("8fs", 24bits, P256). I have not bothered to compare the new P256 implementation against my old P192 files from the 'Insane Edition', but the improvements from the new version do seem more pronounced and easier to pick out.
 
Last edited:

Users who are viewing this thread

Back
Top