A Philosophical Issue (matrix synthesizing surround sound)

[kfbkfb]

Matrix synthesizing surround sound - with or without logic.

Having both (logic directed) DPL2 and (no logic) DynaQuad, I'm tending to favor DynaQuad because it doesn't change the content of stereo L and R at all, although the surround channel has poor channel separation from L and R, it sometimes sounds "discrete".

I'm considering rewiring my DPL2 system to provide surround sound from DPL2 movie [for videotaped TV shows and movies] and DynaQuad [for stereo content] and not using DPL2 music.


Kirk Bayne

 

[par4ken]

You are touching on two issues here, the first is that a simple matrix decoder can sound very good. The second is that Dolby sucks on music. My home built (Dyna style) decoder sounds very nice and is fully adjustable, info is in the thread about the Akai SS-1. The Dyna speaker connection is simply the poor mans way to surround.

I recommend Sansui Vario-matrix for stereo to surround enhancement (notice that I didn't say synthesis). Involve works very well but lacks the surround mode, which IMHO is really needed to pull the stereo sound apart.

Another cheap alternative is the Gemini SP-1 (Circle Surround Decoder). The rear effect control on that unit is essential in that it can be adjusted to throw more sound to the rear. After Dyna Quad it's the best bang for the buck.

 

[Sonik Wiz]

Matrix synthesizing surround sound - with or without logic.

Having both (logic directed) DPL2 and (no logic) DynaQuad, I'm tending to favor DynaQuad because it doesn't change the content of stereo L and R at all, although the surround channel has poor channel separation from L and R, it sometimes sounds "discrete".

I'm considering rewiring my DPL2 system to provide surround sound from DPL2 movie [for videotaped TV shows and movies] and DynaQuad [for stereo content] and not using DPL2 music.


Kirk Bayne

The charmer for me in simple no logic decoding is the humble Electro-Voice EVX-4. Elsewhere long ago you said that unit succeeded at its goal playing stereo because it gave people what they expected: quite a wide stereo front soundstage & good depth in the rear. Now if I wanted to enjoy this EV decoding, or demo it to someone, I would probably just do it on my PC it is so very simple in just about any audio editing app.
My first encounter with DPL II was with the purchase of a (rather expensive) Tag McLaren AV32R pre-pro. It seemed an improvement over other methods but no where near as good as decoders from the Golden Age of Quad that was made specifically for music.

The Tag developed early Alzheimer's & did strange things so I moved to an Anthem AVM 30 pre-pro. It has worked perfect for years. It's nice the DPL II is adjustable & can be used for QS but it still fails to impress me. I was surprised when I slowly noticed that DPL II is rather sluggish in reacting to fast changing music & tends to muddy fine detail. A good example is massed strings in an orchestra. The DPL II tends to congeal the instruments into one big (for example) violin sound. Properly done in stereo or surround you should get the feel that there are separate instruments playing. Maybe you can't exact count 5 violins but there should be enough detail to make you aware of multiple instruments.

So I fully appreciate your interest in best decoding for music vs movies. I simply just don't use DPL II at all anymore & I am super happy with the Surround Master for decoding with music or movies. Of course the high res formats are decoded by my Oppo player. But the SM works magic on TV shows like Netflix where even some 5.1 audio is better down mixed to stereo & fed to the SM.

 

[kfbkfb]

Using matrix "decoding" to create/synthesize surround sound from stereo - the stereo mix is the what the artist(s) want, extracting surround sound from this mix by (dynamically) reducing the L to R channel separation seems to be a step away from hearing the stereo mix as it was intended to be heard.

More thought - perhaps I'll go back to DPL(1) [movie] for listening to my videotaped Dolby Surround encoded content, since it is the intended decoder for that type of content.

One experiment (I can't do this because the DPL decoder is built in to my receiver) - feed stereo into a matrix decoder (w/or wo/logic), only use the back/surround outputs from the decoder and send the pure stereo (unprocessed by the matrix decoder) to the stereo/front channels, see if the pure stereo + created/synthesized surround sound sounds better than using the matrix decoder processed front channels.


Kirk Bayne

 

[Sonik Wiz]

You are touching on two issues here, the first is that a simple matrix decoder can sound very good. The second is that Dolby sucks on music. My home built (Dyna style) decoder sounds very nice and is fully adjustable, info is in the thread about the Akai SS-1. The Dyna speaker connection is simply the poor mans way to surround.

The Dyna or Scheiber speaker hook up is indeed the simplest (and not too shabby) way to advance from stereo. The down side is unless you use L pads at strategic points it is not adjustable level control or decoding. Your DIY is a great way to go if you have 4 separate amps to use. And @MidiMagic has various circuits quite elaborate that works at the speaker level.

I recommend Sansui Vario-matrix for stereo to surround enhancement (notice that I didn't say synthesis). Involve works very well but lacks the surround mode, which IMHO is really needed to pull the stereo sound apart.

I have run across a surprising number of posts from QQ members that listen to surround content in surround but listen to stereo as stereo as a rule. I'm probably like you, I listen to everything in surround. Stereophiles will say that's not the way the content was intended. But when you consider the significant phase/amplitude relationships in a stereo recording representing width & depth, it's not just about "synthesizing" fake quad. It's more correct to think of this as a stereo decoder.

Of course that term may confuse more than help.... when I was a kid "stereo decoder" referred to a add on multi-plex decoder so you could play stereo from your mono FM receiver.

The semantics of synthesized quad, stereo to enhanced surround, or stereo decoding is one that hasn't been settled on for 50 years. I have used the tag S2S for Stereo to Surround; a nice all enveloping term that no one has adopted. So I Find the longevity of the Sansui terms most acceptable. Hall mode for an up front decode of stereo with pleasant ambience in the rear. Basic QS Vario-Matrix gives a nice 180 deg soundfield from stereo. And Synthesizer mode still has validity because I think it is commonly accepted this refers to a 270 deg wrap around effect synthesized from stereo.

Another cheap alternative is the Gemini SP-1 (Circle Surround Decoder). The rear effect control on that unit is essential in that it can be adjusted to throw more sound to the rear. After Dyna Quad it's the best bang for the buck.

Check this out:

https://www.ebay.com/itm/224532086123
For that price I was actually tempted. But to paraphrase @gene_stl topic title: I need another decoder like I need another hole in the head. Maybe Kirk will purchase & he can report back!

 

[par4ken]

Using matrix "decoding" to create/synthesize surround sound from stereo - the stereo mix is the what the artist(s) want, extracting surround sound from this mix by (dynamically) reducing the L to R channel separation seems to be a step away from hearing the stereo mix as it was intended to be heard.

And that is exactly what you still hear via QS Surround. The stereo is simply magnified, spread around the room 270°, no left to right separation loss. It's like listening to stereo through a magnifying glass! You can make out fine details of the mix much more clearly that with only two speakers. Even those lack luster narrow stereo mixes actually produce more normal sounding stereo (from the front) when listened to that way. That might not be true to the original but only if the intent was a mono sounding mix!

If the stereo sound field is too large/wide for you you can move the back speakers more to the sides or even a bit to the front. Either way it beats the crap out of two speaker stereo!

Also I could care less what the original intent of the producer/artist was, I want my surround.

 

[par4ken]

The Dyna or Scheiber speaker hook up is indeed the simplest (and not too shabby) way to advance from stereo. The down side is unless you use L pads at strategic it is not adjustable level control or decoding. Your DIY is a great way to go if you have 4 separate amps to use. And @MidiMagic has various circuits quite elaborate that works at the speaker level.

That is why a decoder like the Akai or the one that I proposed does the the very same thing but with the advantage of full adjustability. Why play around with passive techniques when you have extra amplifier channels available? The adjustable decoder can be set up the same as the EV-4 decoder or better yet tweaked to your own taste.

 

[ar surround]

I really liked Dynaquad when I first started using it. I had four speakers connected to a Dynaquad equipped Dyanco SC80Q integrated amplifier. It was my first experience with surround sound and I was definitely impressed. I've never cared for PLII for synthesizing surround from stereo. It never produced any "wow" moments, and I thought that Dynaquad was better. PLII sounded like an SQ decoder struggling with straight stereo signals. (However, PLIIx seemed to be a different animal. It was much more effective than PLII, and they should have given it some other name.)

I love the Surround Master Involve mode, but would like to hear this QS Sythesizer mode. IIRC, one of you jerry-rigged something to attach to the front end of the SM to provide QS Synthesizer capabilities.

 

[par4ken]

The Dyna or Scheiber speaker hook up is indeed the simplest (and not too shabby) way to advance from stereo. The down side is unless you use L pads at strategic points it is not adjustable level control or decoding. Your DIY is a great way to go if you have 4 separate amps to use. And @MidiMagic has various circuits quite elaborate that works at the speaker level.



I have run across a surprising number of posts from QQ members that listen to surround content in surround but listen to stereo as stereo as a rule. I'm probably like you, I listen to everything in surround. Stereophiles will say that's not the way the content was intended. But when you consider the significant phase/amplitude relationships in a stereo recording representing width & depth, it's not just about "synthesizing" fake quad. It's more correct to think of this as a stereo decoder.

Of course that term may confuse more than help.... when I was a kid "stereo decoder" referred to a add on multi-plex decoder so you could play stereo from your mono FM receiver.

The semantics of synthesized quad, stereo to enhanced surround, or stereo decoding is one that hasn't been settled on for 50 years. I have used the tag S2S for Stereo to Surround; a nice all enveloping term that no one has adopted. So I Find the longevity of the Sansui terms most acceptable. Hall mode for an up front decode of stereo with pleasant ambience in the rear. Basic QS Vario-Matrix gives a nice 180 deg soundfield from stereo. And Synthesizer mode still has validity because I think it is commonly accepted this refers to a 270 deg wrap around effect synthesized from stereo.



Check this out:

https://www.ebay.com/itm/224532086123
For that price I was actually tempted. But to paraphrase @gene_stl topic title: I need another decoder like I need another hole in the head. Maybe Kirk will purchase & he can report back!

Twenty dollars! the case alone is worth more than that! All you need are to add some XLR to RCA adaptors as it has balanced inputs and outputs! Only ships to the US pity!

 

[ar surround]

Check this out:

https://www.ebay.com/itm/224532086123For that price I was actually tempted. But to paraphrase @gene_stl topic title: I need another decoder like I need another hole in the head. Maybe Kirk will purchase & he can report back!

Twenty dollars! the case alone is worth more than that! All you need are to add some XLR to RCA adaptors as it has balanced inputs and outputs! Only ships to the US pity!

Someone snagged it. I hope it was a QQ person.

 

[J. PUPSTER]

Someone snagged it. I hope it was a QQ person.

I used to think (many years ago now) that CS sounded pretty good (decoded off my Marantz AVR), until I got an Involve Surround Master, blows it out of the water!

 

[Sonik Wiz]

I love the Surround Master Involve mode, but would like to hear this QS Sythesizer mode. IIRC, one of you jerry-rigged something to attach to the front end of the SM to provide QS Synthesizer capabilities.

That would be me. I posted an extensive article on a remote control pre-synth some time ago. I can dig it up if you want. I think Ken has this capability too.

For grins if you want you can download this folder with two audio files in it. Both are just white noise panned left to right. One is regular stereo & the other has what I call pre-synth phase blending. Using the Involve QS mode, or QS on a Sansui unit, the stereo file should start as center left, move to center front and then finish at center right. The latter enhanced track should start left back and do a 270 deg clock sweep and finish at right back. That is, for those that have levels and speakers set up properly haha etc.

https://www.dropbox.com/s/lxyajn2ycuufje7/TEST PAN L to R.zip?dl=0

 

[LB-V]

Gentlemen, if you don't mind, I always get confused between the difference of an Involve SM stereo to surround and the synthesized surround (i.e. via a Tate or QSD-1) is. I've read it a few times, usually in a @Sonik Wiz post, and get it for a moment, but then forget the premise again. Do you mind explaining again? How is it that a Tate, for example, will give a 270 degree surround, and the Involve doesn't? Or more accurately, what precisely is the difference? Apologies, it intrigues me, but collectively, you've probably explained it a thousand times.

 

[par4ken]

The Involve SM is a QS decoder. With that mode of operation using an unencoded stereo signal a hard left will decode equally from both left front and left back speakers. Same thing for a hard right. Center front will decode from the front only (i.e. vocals). Out of phase signals will decode from the rear (center back).

What Sansui QS Surround mode does is add some out of phase blend to a stereo signal to replicate an actual QS encoded signal. Now a hard left is decoded from the left back only the same goes for a hard right decoded as right back. The Cf and Cb signals decode much as before, with Cf slightly reduced in level and Cb slightly increased.. Signals panned in between will decode from the front speakers. A paned signal will move around the room as an inverted U, 270° .

The Audionics S&IC does a very similar thing with the stereo enhance control which is fully adjustable. The signal will pan in as an inverted V , 270° . There is no attenuation of Cf or increase in Cb level. The Fosgate decoders have a switch setting but it is not adjustable.

Scott's original circuit adds both in phase blend and out of phase blend, which works to create fully adjustable Hall/QS/Surround effects from a Sansui or Involve decoder.

 

[Sonik Wiz]

The Involve SM is a QS decoder. With that mode of operation using an unencoded stereo signal a hard left will decode equally from both left front and left back speakers. Same thing for a hard right. Center front will decode from the front only (i.e. vocals). Out of phase signals will decode from the rear (center back).

What Sansui QS Surround mode does is add some out of phase blend to a stereo signal to replicate an actual QS encoded signal. Now a hard left is decoded from the left back only the same goes for a hard right decoded as right back. The Cf and Cb signals decode much as before, with Cf slightly reduced in level and Cb slightly increased.. Signals panned in between will decode from the front speakers. A paned signal will move around the room as an inverted U, 270° .

The Audionics S&IC does a very similar thing with the stereo enhance control which is fully adjustable. The signal will pan in as an inverted V , 270° . There is no attenuation of Cf or increase in Cb level. The Fosgate decoders have a switch setting but it is not adjustable.

Scott's original circuit adds both in phase blend and out of phase blend, which works to create fully adjustable Hall/QS/Surround effects from a Sansui or Involve decoder.

Thanks Ken. A extensive discussion. I was too exhausted to revisit this again. Well done.

Edit: make this a sticky for future enquires.

 

[kfbkfb]

From now on, I plan to use the phrase surround sound creation rather than synthesize.

Probably will also be specific as to whether the creation was due to matrix decoding [w/ or wo/ preprocessing (a la CBS SQ stereo enhancement) and logic assistance] or upmixing.


Kirk Bayne

 

[LB-V]

Thank you, much appreciated. Sorry to wear you guys out!

Edit: make this a sticky for future enquires

I will definitely bookmark for myself and future reference. Thanks again.

 

[kfbkfb]

Maybe a drawing (hand/computer) with the mapping of sound locations in stereo to the sound locations in created surround sound (matrix decoding with and without preprocessing [CBS et al.]) would be helpful in explaining what the various matrix decoding systems do to create surround sound.


Kirk Bayne

 

[ar surround]

What Sansui QS Surround mode does is add some out of phase blend to a stereo signal to replicate an actual QS encoded signal. Now a hard left is decoded from the left back only the same goes for a hard right decoded as right back. The Cf and Cb signals decode much as before, with Cf slightly reduced in level and Cb slightly increased.. Signals panned in between will decode from the front speakers. A paned signal will move around the room as an inverted U, 270° .

Does anyone know if the phase blend can be introduced using a DAW such as Audacity? I've heard a track with a 10 degrees phase blend played through the Surround Master IA 4.0 mode. It was extremely impressive.

I also played that same track through 11 speaker Dolby Surround and it was in-f%@$ing-credible. But I suspect that with Dolby Surround, certain tracks may sound overblown* while the SM, as usual, will do no harm.

*Dolby Surround sometimes exhibits too much rear reverberation for my tastes even with tracks that haven't been messed with.

 

[Sonik Wiz]

Does anyone know if the phase blend can be introduced using a DAW such as Audacity? I've heard a track with a 10 degrees phase blend played through the Surround Master IA 4.0 mode. It was extremely impressive.

I also played that same track through 11 speaker Dolby Surround and it was in-f%@$ing-credible. But I suspect that with Dolby Surround, certain tracks may sound overblown* while the SM, as usual, will do no harm.

*Dolby Surround sometimes exhibits too much rear reverberation for my tastes even with tracks that haven't been messed with.

Read till your eyes bleed:

https://www.quadraphonicquad.com/forums/threads/stereo-pre-synthesis.25299/#post-446553
Adobe Audition 3 has gone from being simply "long in the tooth" to ancient. How ever I continue to cling to it as it works great from Win XP > Win 10. It also has some features & polish that is not found on any of the current freebie/low cost audio editors. If some day a Windows update breaks it, I would probably move to Reaper.

So your on your own to figure out how to do this pre-synth in specific on another editing app. But once you know the basics it should be easy to figure out.

Also there are VST plug ins & even free standing apps that will do this but none in the same way as AA 3 Pan/Expand (process) function.

Yet another shout out to @chucky3042 that a 270 deg synth mode really needs to be integral to the next version of the Surround Master.