SQ Decoding with Adobe Audition for New Converters - Start Here

[Groovy Daniel]

I meant the RIAA equalization!!!!

RIAA equalization isn't a tremendous obstacle. If you look at the difference between RIAA EQ at normal speed and RIAA EQ at half speed and adjust it toward the mean, it's within roughly +/- 2dB across the frequency spectrum. While 2dB isn't insignificant, it's within the tolerance of the frequency response of typical cartridge/tonearm combinations, let alone the compromises afflicted upon an audio signal by disk mastering and vinyl playback. Vinyl (i.e., analog disk reproduction) is a horrible medium. The fact that it was the best available medium for commercial distribution of recorded sound up to 1982 when the CD arrived doesn't mean it was ever a faithful medium by modern standards. (I'd be happy to tick off the limitations of disk mastering and vinyl playback in a separate posting.) What vinyl is very good at, however, is providing a theoretical duplicate of the master disk, which is one generation removed from the master tape (but modified to accommodate the limitations of analog disk reproduction), and may be the best available source for vintage recordings, especially quadraphonic recordings whose master tapes may be lost or suffer from the "sticky shed syndrome" that affects much of the 1970s tape stock. My point is, the sonic signature of a vinyl recording should not be accepted as the definitive or "correct" sonic signature, because it's inherently compromised. The +/- 2dB in RIAA reproduction caused by 1/2 speed playback is insignificant compared to the EQ required to blow past the limitations of vinyl and disk mastering.

 

[timalin]

I have been going through likely SQ- and QS-encoded CDs using scripts found here (SQ-Final Public Release 13 Aug 2013.scp and QS_dec_AA3_v1_.scp).

I'm brand new to decoding and I don't have great listening ears, so I was wondering whether anyone would be interested in having a listen and offering an opinion on whether it's working or whether it's quad.

The method I use is as follows:

I've used Adobe 3, upsampled to 32-bit float, 192 kHz, cut by 3dB when doing SQ (as per Kempfand's instructions on his script) to create the two-channel pairs of files.

I then load the two files into Audacity, split them into mono tracks and rearrange them into the correct order where necessary (1: LF; 2: RF; 3: Ls; 4: Rs).

Audacity does not support .w64 so to overcome the .wav file size limit I split the album into pieces not over 21 minutes and export each piece to a wav file.

Using Foobar2000 I create ReplayGain tags for the (usually three) .wav files. I then convert these files to a single .wav image at 24-bit (applying dither), apply RG album info at +20dB while preventing clipping according to peak (in order to normalize to the highest peak of the album), and apply the PPHS resampler at 48kHz in ultra mode. 24/48 would seem more than adequate when decoding from 16/44.1.

After that and still in Foobar, I split the image into individual .flac files using a .cue file created with Cue Tools using the original unprocessed files.

I'd be interested to know if this is fundamentally sound, and if there's anything I can do to improve the accuracy of the resulting files. Also, I'm sure there are simpler and more efficient ways to achieve the same results, so any pointers would be appreciated.

I've uploaded one-minute excerpts (flac) from the following albums to Dropbox (link here), and would be keen to hear anyone else's thoughts.

Ash Ra Tempel - Starring Rosi
The Awakening - Hear, Sense and Feel (1972)
The Cosmic Jokers - Gilles Zeitschiff
The Cosmic Jokers - Planeten Sit-In (1974)
Walter Wegmüller - Tarot (1973)

I'm also planning on attempting the following albums on the basis that they are likely to be matrix-encoded (CD unless noted).

Band Of Thieves - Band of Thieves (vinyl)
Cleveland Eaton - Plenty Good Eaton (1974)
The Cosmic Jokers - Sci Fi Party [1974]
Doug Carn - Spirit Of The New Land (1972)
Hot Tuna - King Biscuit Flower Hour_ 1974-09-26 [FM?]
Henry Franklin - The Skipper (1972)
Klaus Schulze - Cyborg (1973 RE 2006)
John Klemmer - Magic And Movement (vinyl)
Mythos - Dreamlab
New York Shakespeare Festival - Threepenny Opera (1976)
Popol Vuh - Seligpreisung [2004 Remaster]
Roland Haynes - 2nd Wave (1975)
Rudolph Johnson - The Second Coming (1973)
The Who - Tales From The Who (1973) (FM > vinyl > CD)
Wallenstein - Cosmic Century {1994 Reissue}
Wallenstein - Stories, Songs & Symphonies (1974)
Yatha Sidhra - A Meditation Mass [1974]

 

[Franklin]

I expect 192kHz is going to up the processing time without offering you much if any sound improvement. I'd be inclined to put two empty channels into the quad (3 = centre, 4 = sub IIRC) then rears are 5 and 6 then save as a six channel WAV. Personally I'd also split into tracks (but allowing a little time before the track starts) rather than 21 minute chunks + manual CUE file. If burning to DVD set the audacity timeline to the correct video format so you can get the edit on the frame. Others here are far more experienced at this than I am though.

 

[timalin]

I expect 192kHz is going to up the processing time without offering you much if any sound improvement. I'd be inclined to put two empty channels into the quad (3 = centre, 4 = sub IIRC) then rears are 5 and 6 then save as a six channel WAV. Personally I'd also split into tracks (but allowing a little time before the track starts) rather than 21 minute chunks + manual CUE file. If burning to DVD set the audacity timeline to the correct video format so you can get the edit on the frame. Others here are far more experienced at this than I am though.

Thanks for the response. I expect 192 kHz is probably overkill and I might otherwise be able to avoid splitting the image.

As I was making these files for my own personal use I chose 4.0 rather than 5.1, but I should probably go for 5.1 in case I redistribute them. I don't imagine there's any size overhead in the empty channels when compressed to flac.

I chose to split into 20-minute bits as it's quicker and I presumed lossless. However, I just compared two files, one unsplit and the other split then recombined, and have discovered they are not bit-identical, although any differences are inaudible. I conducted more tests, saving the audio in various ways and discovered that Audacity is brings a different, inaudible loss into the audio however I choose to do it. Even if I use identical methods to export the identical audio without making any changes, I end up with files that are not identical to the original or to each other. This has me concerned that I shouldn't be using Audacity if it is making arbitrary changes to the audio. Maybe there is a better program that will allow me to rearrange the channels and save them in a single .w64 file without adulterating the audio.

 

[Franklin]

Have you checked if it dithers on exporting and importing?

 

[timalin]

Have you checked if it dithers on exporting and importing?

Thanks. I just checked the guide and it says "exporting to 32-bit PCM WAV format is 100% lossless - no dithering is applied", so I doubt that can be the issue.

 

[timalin]

My understanding was that it doesn't dither unless there's a reduction in the bit-rate.

 

[Snood]

:bump: For Timalin

 

[EMR]

I have Adobe Audition 3.0 for Windows, which is running just fine, and I also have an Audition script called "SQ*Final Public Release 13th Aug 2013". I've tried to make it work, but the link to the script I created (in the Favorites menu) is always greyed out. Obviously, I've done something wrong. I've searched this thread, but I couldn't find a step-by-step guide as to how one should install this to make it work. In addition, it is unclear to me if the process is supposed to be entirely autonomous (like, opening an SQ-encoded WAV file and then running the script, which will generate one or more decoded WAVs) or one has to do some operations with Audition and the operating system (like copying the WAV file one or more times with particular names [front, rear, center or whatever] in a particular folder). Is there a detailed guide somewhere? Can someone provide a link or some pointers as to how to proceed? Many thanks!

 

[Franklin]

You've got to open the file to be processed first.

 

[EMR]

You've got to open the file to be processed first.

What do you mean by "the file"? An SQ-encoded WAV file? I've already done that. I need specific instructions:
1) Where should the SQ decoding script be copied?
2) Does it have any dependencies, like other scripts that are supposed to be there? If such dependencies exist, where do I get the extra scripts? How do I install them?
3) Does it require that the SQ-encoded WAV file be located in some specific folder? Which one?
4) Does it require a certain tree structure of folders somewhere (like C:\Temp, C:\SQ or the like)? Which one, specifically?
5) Other than opening the SQ-encoded WAV file, do I have to do something extra, like saving the file with some particular name in some particular, folder, or make one or more copies of said file? Which is the exact procedure?
6) Finally, how do I launch the script?

 

[Franklin]

I haven't done this for years, I'm not even on the laptop where the stuff is, so I'm quite rusty. The script text itself will tell you where the folders should be. I think they were in the root drive on drive C, but you can adjust this in the script. Let me see if I can fire it up and get back to you...

 

[Franklin]

I think you also needed a patch on the software (making it version 3.01?) to do the job.

 

[EMR]

Thank you for the info. I'll look into the 3.0.1 update and see if I can get it somehow. I was able to install Phasebug.dll to "C:\Program Files\VSTPlugIns" and edited the script so that it will point to "C:\SQ Temp" instead of "E:\SQ Temp", but the script is still greyed out in the AA Favorites menu.

 

[Franklin]

Right, what OS are you using? I was using Windows 7. The script was in C:\Program Files (x86)\Adobe\Adobe Audition 3.0

Your unencoded file needs to be saved identically as two files: Front.wav and Rear.wav and be sat in the C:\SQ temp folder.

This may not be strictly true as it rings a bell that I altered part of the script to adjust for file locations. It you read the script in a text editor you can see where it expects to see the files.

It rings a bell that I had a bit of an issue getting PhaseBug to be acknowledged and maybe getting the script to work too, but I have so little memory of it. Have you told Audition that it's there? You may have to create a nonsense script in Audacity and see where it saves and if it tells you to do anything to make it accessible, or a cheeky Help search.

Needless to say, I don't know if there are any copyright issues you may come across in Spain and I'm not suggesting you break any of them.

 

[EMR]

I have a Mac. I'm running AA 3.0 on a 32-bit Windows 10 virtual machine. The script is in "C:\Program Files\Adobe\Adobe Audition 3.0\Scripts".

I've just copied the SQ-encoded WAV file twice to "C:\SQ temp" as Front.wave and Rear.wav. The script is still greyed out.

I don't know how to "tell" AA that Phasebug.dll is in "C:\Program Files\VSTPlugIns". I've looked in Preferences and couldn't find anywhere that will let me specify an VST folder. I would think that is done automatically?

 

[Franklin]

Memory tells me that it wasn't automatic as expected but it's so long since I did it I can't remember, sorry.

 

[Franklin]

Try Effects from the top line and select VST Plug-In Manager... or maybe even Refresh Effects List.

 

[Franklin]

To run the script you need to click the script title next to the Run Script button first.

 

[EMR]

OK, I found the VST Plug-In Manager and selected the relevant folder. It found Phasebug, so I supposed it is now active, although it warned that some "instruments" weren't available unless I went to some panel. I have no idea what that may mean.

As for the "Run Script button", I don't know where that is. I can't see such a button.