• QuadraphonicQuad welcomes you and encourages your participation! Treat all members with respect. Please keep all discussions civil, even when you have a strong opinion on a particular topic.

    Do not offer for free, offer for sale, offer for trade, or request copies or files of copyrighted material - no matter how rare or unavailable to the public they might be. We do not condone the illegal sharing of music. There are many places on the internet where you can participate in such transactions, but QuadraphonicQuad is not one of them. We are here to encourage and support new multichannel releases from those companies that still provide them and as such the distribution of illegal copies of recordings is counter-productive to that effort. Any posts of this sort will be deleted without notification.

    Please try to avoid discussions that pit one format against another. Hint for new users: make liberal use of the search facilities here at QuadraphonicQuad. Our message base is an incredibly rich resource of detailed information on virtually all topics pertaining to surround-sound. You will be surprised at what you can find with a little digging!

Can We Talk About Upmixes?

Help Support QuadraphonicQuad:

boondocks

1K Club - QQ Shooting Star
Joined
Apr 19, 2006
Messages
1,256
Location
boondocks of NC Florida
Long and boring. Do NOT read.
As someone who has done hundreds of upmixes, I have had some satisfying successes that were very, very discreet. Being discreet isn't the be-all end-all in listening quality, even so. I know Aart (EoH or Eye of Horus on the net) was a big advocate for Ambisonics - blended methods to remove that aggravating "swooshy" sound & other artifacts that sometimes inserted themselves, and for general smoothness.

At one time there were folks using a voice removal/karaoke method in Nero software circa Nero 7 to attain separation of stereo. I suspect many have heard those, but I don't know exactly who they were, just their avatar names.

dts350z (Glenn) worked hard on SPEC, which is more than a simple upmix engine but with the Plogue Bidule version allows a wide latitude of adjustments.

The early upmix routines conceived by Aart and implemented by Kempfand with Plogue Bidule, as well as others that invented ways to use existing vst's to separate sound, (you too Mr. SMaCS ) were my early indoctrination into the upmix gig. Whew, that's three countries right there!

DKA, while not a software inventor type, attacked upmixing with a mania like no one else, turning out prolific (and often quite good!) upmixes sometimes by the dozens it seemed. I have heard from him occasionally, but he's on to other things now as far as I know......and hell, I mention him because I miss him, I guess, or at least miss those days and times......but life moves on.
Many others were active in upmixing as well.

Oh, I used to have conversations with John (Penteo) before he had his inventions yanked out from under him. I thought his mechanical/software methods were not really all that good on average, and once we swapped upmixes purely to show each other what we could do. lol. Although I thought he was on the wrong track, I admired his work and had to grudgingly admit that his way eliminated artifacts in the sound.
I don't know what, if anything, has changed with Penteo; could be it is I that have changed, but I have had a chance, thanks to a fine fellow, to listen to some Penteo upmixes lately that were, if not on par with actual multitracks, very very enjoyable and worthy of repeat listening without a doubt in my non miltary mind...

The End (ps the software engineer did it)
 

J. PUPSTER

2K Club - QQ Super Nova
QQ Supporter
Joined
May 30, 2017
Messages
4,249
Location
CALIFORNIA (CENTRAL)
Long and boring. Do NOT read.
As someone who has done hundreds of upmixes, I have had some satisfying successes that were very, very discreet. Being discreet isn't the be-all end-all in listening quality, even so. I know Aart (EoH or Eye of Horus on the net) was a big advocate for Ambisonics - blended methods to remove that aggravating "swooshy" sound & other artifacts that sometimes inserted themselves, and for general smoothness.

At one time there were folks using a voice removal/karaoke method in Nero software circa Nero 7 to attain separation of stereo. I suspect many have heard those, but I don't know exactly who they were, just their avatar names.

dts350z (Glenn) worked hard on SPEC, which is more than a simple upmix engine but with the Plogue Bidule version allows a wide latitude of adjustments.

The early upmix routines conceived by Aart and implemented by Kempfand with Plogue Bidule, as well as others that invented ways to use existing vst's to separate sound, (you too Mr. SMaCS ) were my early indoctrination into the upmix gig. Whew, that's three countries right there!

DKA, while not a software inventor type, attacked upmixing with a mania like no one else, turning out prolific (and often quite good!) upmixes sometimes by the dozens it seemed. I have heard from him occasionally, but he's on to other things now as far as I know......and hell, I mention him because I miss him, I guess, or at least miss those days and times......but life moves on.
Many others were active in upmixing as well.

Oh, I used to have conversations with John (Penteo) before he had his inventions yanked out from under him. I thought his mechanical/software methods were not really all that good on average, and once we swapped upmixes purely to show each other what we could do. lol. Although I thought he was on the wrong track, I admired his work and had to grudgingly admit that his way eliminated artifacts in the sound.
I don't know what, if anything, has changed with Penteo; could be it is I that have changed, but I have had a chance, thanks to a fine fellow, to listen to some Penteo upmixes lately that were, if not on par with actual multitracks, very very enjoyable and worthy of repeat listening without a doubt in my non miltary mind...

The End (ps the software engineer did it)
I did read; and not boring at all thank you 😊
 

Wagonmaster_91

300 Club - QQ All-Star
Since 2002/2003
Joined
Mar 11, 2002
Messages
372
Location
Dallas,TX
If you are getting 'swooshy" or "watery" results in the rear channels when using SPEC, then you are most likely using the ArcTan method and have the options controls pushed too radically in order to increase discreteness. I find the Slice method has much better rear channel fidelity and less artifacts, plus gives a result much closer to discrete quad.
I understand ArcTan is designed to give a more 'wrap-around' effect than Slice so if that is your preference then use that, but be aware that getting the 'water' out of the rear channels is a bigger challenge.
 

J. PUPSTER

2K Club - QQ Super Nova
QQ Supporter
Joined
May 30, 2017
Messages
4,249
Location
CALIFORNIA (CENTRAL)
If you are getting 'swooshy" or "watery" results in the rear channels when using SPEC, then you are most likely using the ArcTan method and have the options controls pushed too radically in order to increase discreteness. I find the Slice method has much better rear channel fidelity and less artifacts, plus gives a result much closer to discrete quad.
I understand ArcTan is designed to give a more 'wrap-around' effect than Slice so if that is your preference then use that, but be aware that getting the 'water' out of the rear channels is a bigger challenge.
I would be interested in knowing what command line options you use; the Spec program still vexes me?
 
Last edited:

Jim the Oldbie

800 Club - QQ All-Star
Joined
Jun 6, 2015
Messages
879
Location
Midwest USA
I got excited awhile back when I found a version of The Wall that was allegedly posted by the late lamented EoH, assuming it would be one of his upmixes. But it turned out to be just the edited movie soundtrack as described by Jon above.
 
Last edited:

boondocks

1K Club - QQ Shooting Star
Joined
Apr 19, 2006
Messages
1,256
Location
boondocks of NC Florida
If you are getting 'swooshy" or "watery" results in the rear channels when using SPEC, then you are most likely using the ArcTan method and have the options controls pushed too radically in order to increase discreteness. I find the Slice method has much better rear channel fidelity and less artifacts, plus gives a result much closer to discrete quad.
I understand ArcTan is designed to give a more 'wrap-around' effect than Slice so if that is your preference then use that, but be aware that getting the 'water' out of the rear channels is a bigger challenge.
Most methods have drawbacks. But, it always depends on the actual source music, and knowing ways to sometimes avoid severity of the artifacts. My personal preference is to almost always use the Plogue Bidule version of SPEC, (or even the ambisonics infused ones) as it usually allows for greater flexibility. Plus there are VST's that can be used to helpful effect sometimes. As I've aged I confess that "all out separation" would not be my goal now, but work harder to get a more listenable , if milder, outcome.
But you are definitely not alone in preferring SLICE as I've seen that comment often over the years (years? wow I guess so).
Anyway my current project is to take a stereo dvd and insert a Penteo upmixed group of tracks in place of the stereo. I've managed this with BD's before, by taking for example two Tangerine Dream BD's that were released in stereo, while an identical DVD was released in each case with DD 5.1! Pulling the DD track out of the DVD's and inserting into the BD's turned out to be fairly easy for me, somehow!
Anyway for some reason this particular DVD and multiAVCHD are kicking my butt, (and it is a little out of spec.). I'll keep at it and quit writing these damn books now.
Forgive an old man's reminiscing.
 

humprof

1K Club - QQ Shooting Star
Joined
Jun 10, 2016
Messages
1,400
Location
NoCal
I got excited awhile back when I found a version of The Wall that was allegedly posted by the late lamented EoH, assuming it would be one of his upmixes. But it turned out to be just the edited movie soundtrack as described by Jon above.
I think Jon may be describing a different edited movie version, but I could be wrong. With help from some friends, EoH did in fact complete what he called The Wall 2.0 not long before his death in 2018. It's Frankensteined together from three different audio sources--mostly a "Dr. Rob" rip of the 2016 vinyl remaster--and two video sources. Comes with a booklet where EoH goes into some detail about his Ambisonics-based S2S upmix method.
 

DistantEarth

Member
Joined
Jan 27, 2020
Messages
18
Location
Charleston, SC
I'm jumping back in because I was thinking about speaker placement and my home theater is a Vizio 5.1.4 Soundbar which sounds amazing for movies but stereo music is a little lacking. The upmixes fill the room much better, (I have only done this with ambient music, so far...). I gotta play with some levels but I could see having a little bit of the faux surround might actually work out. I hope I can make a preset so if I want to listen in the theater, I, could hit a button and have the rear channels slightly play the stereo mix. I mean, proper stereo speakers set 5 feet apart sound a whole lot different than 12 feet apart. I might not be so against fuax surround as I was before. Anyone doing something similar? I need to start following, "if it sounds good to you, then it's right". I can't stand by official releases anyway since I got the LA Woman 5.1 Mix (This is how you ruin a perfectly good record).
 

Arconada

Member
Joined
Feb 20, 2007
Messages
33
I used SpecWeb to upmix a few Robert Rich albums and they came out great! (I wouldn't doubt the official 5.1 mixes did something similar because the few samples I got sound the same). I compared the waveforms in Audacity and there are differences but listening, I couldn't tell what they were... I don't know the SpecWeb secret sauce but it does an amazing job. I have the button on my remote to upmix on my Nvidia Shield but that just plays left and right in the rear channels. SpecWeb seems to pull stuff out and move it to the rear channels but in the original resolution. I was interested in it because there are so many subtle sounds in Ambient music and specweb made those come out a little more. I now prefer the upmixes, I made. It also does not reduce the quality like the on demand upmix from the shield.
I

OK here's something I just threw together, working in Audacity that may have some potential.
I used the Audio Fidelity 2010 SACD rip of Miles, Someday My Prince Will Come of the track Old Folks; (it started as a 3 channel -5.0 with a couple silent channels. I deleted the 5th channel, moved the Front Left channel (piano) to both rears (Quad), then moved Mile's trumpet in the center to the Right Front and Coltrane's Sax from the Right Front to the Left Front. I think it sounds pretty cool that way, but might want to do something else with the rear piano channels.
Any thoughts?

View attachment 46810
View attachment 46811
View attachment 46812
I use Audtiion, I don't have any experience with Audacity but I doubt it is better.
 

holland123

1K Club - QQ Shooting Star
Joined
Feb 1, 2010
Messages
1,048
Location
detroit, michigan.usa
The recent 5.1 dvd-a iso or mch flac of Pink Floyds - The Wall that is being circulated has a dubious origin, not sure if it is upmixed or actually real escaped multi's. the one thing I can say for certain is that it's not a DKA, EoH, or Area 51 work.(Jorge's was DTS, AArts was mostly unremarkable, and Area 51's was dull and lifeless) this one however is quite outstanding and is the best 5.1 version of that album that I have ever heard. it is remarkably discrete, sonicly crisp and has a bottom end that will rattle your windows.
 

HomerJAU

Moderator: MCH Media Players
Staff member
Joined
Jun 13, 2013
Messages
2,981
Location
Melbourne, Australia
Garry (HomerJAU) has started processing and capturing his own Specweb upmixes for playback through a media player.
Nah. I bought Penteo 16 after I played with a trial version in late November. It never leaves any artefacts. Great source in equals great sound out. I’m very happy with the results but it’s very dependant on how wide the original stereo image is. If it’s near mono in, it’s not too immersive out. When it’s at its best it’s better than many commercial 5.1 releases.

I was pretty sensitive to the SpecWeb ‘swish’, so I stopped using it a long time ago now. It wasn't in all my UMs but it in enough for me to stop using it.

I’m in full scale production mode with upmixing all my stereo albums to quad or 5.1. I’m about to release a new update to MMH that has an updated tool to semi automate retagging on UMed file’s by copying tags and art from the source stereo files, locating missing artwork on your computer drives and/or network (fuzzy search) and downloading album and disc images art from a web service, remixing quad UMs into a 5.0 or 5.1 file format (AVR compatibility). This will work with SpecWeb and Penteo.
 

boondocks

1K Club - QQ Shooting Star
Joined
Apr 19, 2006
Messages
1,256
Location
boondocks of NC Florida
Nah. I bought Penteo 16 after I played with a trial version in late November. It never leaves any artefacts. Great source in equals great sound out. I’m very happy with the results but it’s very dependant on how wide the original stereo image is. If it’s near mono in, it’s not too immersive out. When it’s at its best it’s better than many commercial 5.1 releases.

I was pretty sensitive to the SpecWeb ‘swish’, so I stopped using it a long time ago now. It wasn't in all my UMs but it in enough for me to stop using it.

I’m in full scale production mode with upmixing all my stereo albums to quad or 5.1. I’m about to release a new update to MMH that has an updated tool to semi automate retagging on UMed file’s by copying tags and art from the source stereo files, locating missing artwork on your computer drives and/or network (fuzzy search) and downloading album and disc images art from a web service, remixing quad UMs into a 5.0 or 5.1 file format (AVR compatibility). This will work with SpecWeb and Penteo.
Could be handy. what is this web service you speak of?
 

Sonik Wiz

1K Club - QQ Shooting Star
QQ Supporter
Joined
May 30, 2005
Messages
1,056
Location
Kansas City
As someone who has done hundreds of upmixes, I have had some satisfying successes that were very, very discreet. Being discreet isn't the be-all end-all in listening quality, even so.
I, for one, would be very interested to hear what methods you’ve used & how they’ve may have evolved over time.

Indeed you have presented a sort of history of approaches to S2S up mixing. In the beginning us old Quaddies were quite happy with hardware implementation such as the S&IC, Sansui QS/Synth & the Fosgate Tate 101A. As surround sound systems “improved” they seemed to move farther away from good music surround to movie surround. I pegged my hope on the new DPL II via a rather pricey Tag McLaren unit & that failed to surpass the earlier ways.

So about 15 years ago I built my first PC with the intent of capturing my Moody Blues R2R collection to dts. Reading on QQ there were several methods of decoding SQ/QS on the PC. I was more interested in S2S & the preceding scripts & method seemed overly complicated & not optimum for decoding stereo.

My approach that I use to this day (tho not much of a need to anymore) actually consisted of 3 steps.

1: After capturing/ripping/DL’ing the audio open the 2 ch stereo up in AA 3. This is sort of a pre-step that lets you check & correct level balance, perhaps some EQ, hard limit the occasional wild transient, etc. It also is the chance to pre-sytnh the waveform with some out of phase blending that simulates the Sansui QS Synth mode. You can use a little, you can use a lot & you can do this by at least two different ways. Then it is ready to decode.

2: Previous scripts for QS decode relied on +- 90 deg phase shifts for the core decoding & cross talk elimination between front L/R & rear L/R. In the Sansui method all decoding & separation enhancement was done using L, R, L+R and L-R. The phase shift was left to the final output to correct for the phase shifts in encoding. Since there was no guaranty that a stereo recording would comply to this I felt no need to deal 90 with deg phase shifts at all. Also since cross talk in the QS/RM decoding is on the diaganol (that is for a Lf sound the cross talk is Rf/Lb) I simply used a channel mixer to decode pairs of Lf/Rb & Rf/Lb. Then using the AA 3 Center Ch Extractor The cross talk between the diagonals is eliminated. It works pretty slick: The amount of final separation is directly proportional the amount of center ch reduction that is set. The decoding & separation enhancement is now complete.

3: The laborious part now that can’t be done with scripts is to copy/paste the wave files into matching front L/R & rear L/R. Put these into AA 3 Multi track view & what I think of as post production can be done. Another chance to EQ, balance levels, etc. I find since this method puts such wide L/R seperation in the rears I frequently use AA 3 Pan/Expand to open up the width of the front chs. And even tho I am a big fan of high DR you’d be surprised how a gentle bit of multi-band compression can actually make it sound better.

After all of this the files can go through the AA 3 Surround Exporter to make separate files for dts, etc. Most always now I just export at 5.1 FLAC.

One of my fave movies from the late 80’s was The Wizard of Speed and Time.
I had it on LD so the video quality is much better than you see on the YouTube clip. But the whole soundtrack was barely stereo & it needed some TLC. I captured the video in uncompressed AVI through a Black Magic Intensity video card & the audio output was kept in the 16/44.1 digital domain via SPDIF to my Delta 1010 sound card. I added many Foley effects such as thunder, office noises, this & that & even lifted the swirling gunfight scene from The Shadow dts DVD to enhance a scene. It really complimented the SFX which done on the cheap is still fun to watch today.

I have used the above method to add full surround to all of 14 Chisato Moritaka’s DVD, various collection of live & music video. These are all done in 16/48 LPCM so it’s clean to begin with. I demux the vob files with absolutely no transcoding of audio or video, work with the audio in AA 3 & mux them back again in Pegays TMPGEnc. Sound is in perfect synch & the dts is the high bit rate core such as used in Blu-Rei.

Dozens of CD’s etc, etc. I have also used SpecWeb with some very good results but I don’t have quite the same control. In any event, I would always use this approach for something special but for everything else the SM v2 is my go to for S2S.
 

boondocks

1K Club - QQ Shooting Star
Joined
Apr 19, 2006
Messages
1,256
Location
boondocks of NC Florida
The initial release will use Fanart.tv which has a fairly good collection. Led Zeppelin as an example:


The more popular the artist the more images. Some fairly obscure artists don’t have any artwork.
Thank you. I've seen the name come on Google images occasionally and just thought it was one of those clubs that charge for images.
 

boondocks

1K Club - QQ Shooting Star
Joined
Apr 19, 2006
Messages
1,256
Location
boondocks of NC Florida
I, for one, would be very interested to hear what methods you’ve used & how they’ve may have evolved over time.

Indeed you have presented a sort of history of approaches to S2S up mixing. In the beginning us old Quaddies were quite happy with hardware implementation such as the S&IC, Sansui QS/Synth & the Fosgate Tate 101A. As surround sound systems “improved” they seemed to move farther away from good music surround to movie surround. I pegged my hope on the new DPL II via a rather pricey Tag McLaren unit & that failed to surpass the earlier ways.

So about 15 years ago I built my first PC with the intent of capturing my Moody Blues R2R collection to dts. Reading on QQ there were several methods of decoding SQ/QS on the PC. I was more interested in S2S & the preceding scripts & method seemed overly complicated & not optimum for decoding stereo.

My approach that I use to this day (tho not much of a need to anymore) actually consisted of 3 steps.

1: After capturing/ripping/DL’ing the audio open the 2 ch stereo up in AA 3. This is sort of a pre-step that lets you check & correct level balance, perhaps some EQ, hard limit the occasional wild transient, etc. It also is the chance to pre-sytnh the waveform with some out of phase blending that simulates the Sansui QS Synth mode. You can use a little, you can use a lot & you can do this by at least two different ways. Then it is ready to decode.

2: Previous scripts for QS decode relied on +- 90 deg phase shifts for the core decoding & cross talk elimination between front L/R & rear L/R. In the Sansui method all decoding & separation enhancement was done using L, R, L+R and L-R. The phase shift was left to the final output to correct for the phase shifts in encoding. Since there was no guaranty that a stereo recording would comply to this I felt no need to deal 90 with deg phase shifts at all. Also since cross talk in the QS/RM decoding is on the diaganol (that is for a Lf sound the cross talk is Rf/Lb) I simply used a channel mixer to decode pairs of Lf/Rb & Rf/Lb. Then using the AA 3 Center Ch Extractor The cross talk between the diagonals is eliminated. It works pretty slick: The amount of final separation is directly proportional the amount of center ch reduction that is set. The decoding & separation enhancement is now complete.

3: The laborious part now that can’t be done with scripts is to copy/paste the wave files into matching front L/R & rear L/R. Put these into AA 3 Multi track view & what I think of as post production can be done. Another chance to EQ, balance levels, etc. I find since this method puts such wide L/R seperation in the rears I frequently use AA 3 Pan/Expand to open up the width of the front chs. And even tho I am a big fan of high DR you’d be surprised how a gentle bit of multi-band compression can actually make it sound better.

After all of this the files can go through the AA 3 Surround Exporter to make separate files for dts, etc. Most always now I just export at 5.1 FLAC.

One of my fave movies from the late 80’s was The Wizard of Speed and Time.
I had it on LD so the video quality is much better than you see on the YouTube clip. But the whole soundtrack was barely stereo & it needed some TLC. I captured the video in uncompressed AVI through a Black Magic Intensity video card & the audio output was kept in the 16/44.1 digital domain via SPDIF to my Delta 1010 sound card. I added many Foley effects such as thunder, office noises, this & that & even lifted the swirling gunfight scene from The Shadow dts DVD to enhance a scene. It really complimented the SFX which done on the cheap is still fun to watch today.

I have used the above method to add full surround to all of 14 Chisato Moritaka’s DVD, various collection of live & music video. These are all done in 16/48 LPCM so it’s clean to begin with. I demux the vob files with absolutely no transcoding of audio or video, work with the audio in AA 3 & mux them back again in Pegays TMPGEnc. Sound is in perfect synch & the dts is the high bit rate core such as used in Blu-Rei.

Dozens of CD’s etc, etc. I have also used SpecWeb with some very good results but I don’t have quite the same control. In any event, I would always use this approach for something special but for everything else the SM v2 is my go to for S2S.
You mentioned copying R2R to dts? I think some of the guys used to use those Creative DTS-610 units. (I have one myself). the input or output can be (5.1) mini RCA or toslink IIRC. I actually have a heavy duty toslink cable 50 ft long running from where my pc is located to the back of the HT in the living room. I used it as a quick and dirty way to listen to my upmixes on the HT instead of on 5 pc speakers which don't have optimal placement.
 

Arconada

Member
Joined
Feb 20, 2007
Messages
33
1: After capturing/ripping/DL’ing the audio open the 2 ch stereo up in AA 3. This is sort of a pre-step that lets you check & correct level balance, perhaps some EQ, hard limit the occasional wild transient, etc. It also is the chance to pre-sytnh the waveform with some out of phase blending that simulates the Sansui QS Synth mode. You can use a little, you can use a lot & you can do this by at least two different ways. Then it is ready to decode.

2: Previous scripts for QS decode relied on +- 90 deg phase shifts for the core decoding & cross talk elimination between front L/R & rear L/R. In the Sansui method all decoding & separation enhancement was done using L, R, L+R and L-R. The phase shift was left to the final output to correct for the phase shifts in encoding. Since there was no guaranty that a stereo recording would comply to this I felt no need to deal 90 with deg phase shifts at all. Also since cross talk in the QS/RM decoding is on the diaganol (that is for a Lf sound the cross talk is Rf/Lb) I simply used a channel mixer to decode pairs of Lf/Rb & Rf/Lb. Then using the AA 3 Center Ch Extractor The cross talk between the diagonals is eliminated. It works pretty slick: The amount of final separation is directly proportional the amount of center ch reduction that is set. The decoding & separation enhancement is now complete.

3: The laborious part now that can’t be done with scripts is to copy/paste the wave files into matching front L/R & rear L/R. Put these into AA 3 Multi track view & what I think of as post production can be done. Another chance to EQ, balance levels, etc. I find since this method puts such wide L/R seperation in the rears I frequently use AA 3 Pan/Expand to open up the width of the front chs. And even tho I am a big fan of high DR you’d be surprised how a gentle bit of multi-band compression can actually make it sound better.
This is more or less the method I use.
- I record stereo material in Audition. Stereo Expand the two channels.
- Decode the stereo in a QS-like routine, except I also produce a Front-Back pair
- Reduce the crosstalk with the channel-extraction routine
- Phase shift the Back Center channel 90 degrees and spead that over the LB-RB pair
- Adjust the levels to compensate for the increased volume of the back channels due to the expansion in the first step
- So now I have 5.0 signals.
- I don't take the trouble of converting it to DTS, just save it to a FLAC-file.
That's it. Some material comes out very impressive, some music shows a bubly effect. If anyone knows why that is please let me know.
 
Last edited:

J. PUPSTER

2K Club - QQ Super Nova
QQ Supporter
Joined
May 30, 2017
Messages
4,249
Location
CALIFORNIA (CENTRAL)
I, for one, would be very interested to hear what methods you’ve used & how they’ve may have evolved over time.

Indeed you have presented a sort of history of approaches to S2S up mixing. In the beginning us old Quaddies were quite happy with hardware implementation such as the S&IC, Sansui QS/Synth & the Fosgate Tate 101A. As surround sound systems “improved” they seemed to move farther away from good music surround to movie surround. I pegged my hope on the new DPL II via a rather pricey Tag McLaren unit & that failed to surpass the earlier ways.

So about 15 years ago I built my first PC with the intent of capturing my Moody Blues R2R collection to dts. Reading on QQ there were several methods of decoding SQ/QS on the PC. I was more interested in S2S & the preceding scripts & method seemed overly complicated & not optimum for decoding stereo.

My approach that I use to this day (tho not much of a need to anymore) actually consisted of 3 steps.

1: After capturing/ripping/DL’ing the audio open the 2 ch stereo up in AA 3. This is sort of a pre-step that lets you check & correct level balance, perhaps some EQ, hard limit the occasional wild transient, etc. It also is the chance to pre-sytnh the waveform with some out of phase blending that simulates the Sansui QS Synth mode. You can use a little, you can use a lot & you can do this by at least two different ways. Then it is ready to decode.

2: Previous scripts for QS decode relied on +- 90 deg phase shifts for the core decoding & cross talk elimination between front L/R & rear L/R. In the Sansui method all decoding & separation enhancement was done using L, R, L+R and L-R. The phase shift was left to the final output to correct for the phase shifts in encoding. Since there was no guaranty that a stereo recording would comply to this I felt no need to deal 90 with deg phase shifts at all. Also since cross talk in the QS/RM decoding is on the diaganol (that is for a Lf sound the cross talk is Rf/Lb) I simply used a channel mixer to decode pairs of Lf/Rb & Rf/Lb. Then using the AA 3 Center Ch Extractor The cross talk between the diagonals is eliminated. It works pretty slick: The amount of final separation is directly proportional the amount of center ch reduction that is set. The decoding & separation enhancement is now complete.

3: The laborious part now that can’t be done with scripts is to copy/paste the wave files into matching front L/R & rear L/R. Put these into AA 3 Multi track view & what I think of as post production can be done. Another chance to EQ, balance levels, etc. I find since this method puts such wide L/R seperation in the rears I frequently use AA 3 Pan/Expand to open up the width of the front chs. And even tho I am a big fan of high DR you’d be surprised how a gentle bit of multi-band compression can actually make it sound better.

After all of this the files can go through the AA 3 Surround Exporter to make separate files for dts, etc. Most always now I just export at 5.1 FLAC.

One of my fave movies from the late 80’s was The Wizard of Speed and Time.
I had it on LD so the video quality is much better than you see on the YouTube clip. But the whole soundtrack was barely stereo & it needed some TLC. I captured the video in uncompressed AVI through a Black Magic Intensity video card & the audio output was kept in the 16/44.1 digital domain via SPDIF to my Delta 1010 sound card. I added many Foley effects such as thunder, office noises, this & that & even lifted the swirling gunfight scene from The Shadow dts DVD to enhance a scene. It really complimented the SFX which done on the cheap is still fun to watch today.

I have used the above method to add full surround to all of 14 Chisato Moritaka’s DVD, various collection of live & music video. These are all done in 16/48 LPCM so it’s clean to begin with. I demux the vob files with absolutely no transcoding of audio or video, work with the audio in AA 3 & mux them back again in Pegays TMPGEnc. Sound is in perfect synch & the dts is the high bit rate core such as used in Blu-Rei.

Dozens of CD’s etc, etc. I have also used SpecWeb with some very good results but I don’t have quite the same control. In any event, I would always use this approach for something special but for everything else the SM v2 is my go to for S2S.
Hey Scott, maybe you could expand more on this statement of yours ( pun intended )
And even tho I am a big fan of high DR you’d be surprised how a gentle bit of multi-band compression can actually make it sound better.

I’ve noticed that although using the Pre-synth method often increases DR for MC it also seems to take some of the desired “punch” or presence out of the music, so is there a way to retain that without just upping the amplitude?
 
2
Top