[JohnnyReef]

I have a pair of Sennheiser HD650 which im very happy with, but there is something im confused about and many others as well. What is really the best between running 2 channel audio without any emulation technologies like CMSS3D and Dolby Headphone vs running 8 Channel  and Dolby Headphone or CMSS3D that downmixes into 2 channel with efffects ?
 
Personally i feel 2 channel stereo only sounds a bit weird, like you only hear sounds in your left ear if someone shoots on your left side, while surround emulation kinda gives you small audio portions of what is going on in your left side on your right side, and most audio in left, creating a kinda illusion that you have surround sound.
But ive read that doing 8 channel audio and then surround emulation, it downmixes 7.1 audio into 2 channel which looses out on alot of sounds? 
 
Would love to get your input on this, this forum is crawling with sound experts Personally i feel for gaming and movies on computer, doing 8 channel and dolby headphone is really the best and pure stereo 2 channel for music only. 

 

[PurpleAngel]

I believe when your listing to a 2-channel (stereo) source, there is no headphone surround sound.
So if you want headphone surround sound, you need 6-channels or 8-channels of audio to start out with.
Then use Dolby or SBX or CMSS-3D or Razor or other, to use all those channels (6 or 8) to creative 2-channels of headphone surround sound, which is sent to a headphones.

 

[ProtegeManiac]

  But ive read that doing 8 channel audio and then surround emulation, it downmixes 7.1 audio into 2 channel which looses out on alot of sounds? 

 

I wouldn't use the word "downmix" because that's not how it works, at least not in how "downmixing" traditionally works. That being, it just puts SL and SR on L and R, and anything on C ends up playing on both L and R.

 

Virtual surround simulation puts reverb on the sound to simulate depth and more precise directionality around the front arc (ie don't expect the helicopter test audio on surround sound programs to have a completely round spin around your head as it will be shaped more like an egg, with the pointier end to the front and what comes out as a flat bottom on a boiled egg to the rear). 

 

If you're looking for totally "hi-fi" with no processing then neither will work. If you just want stereo material to come out a little bit better, then use Crossfeed (soundcards might automatically use this on 2ch material once virtual surround is activated). If you want surround sound to come out with some hint of directionality and depth, then go with virtual surround - just note that given the fast paced action on games it's not like real life where not only would a shot from 20m away in-game not exactly sound like it did come from 20m away as you'll hear it but it won't be like how, in movies or comics for example, Deadpool or Snake Eyes in a pitch black room or blind folded Bruce Wayne will just be able to block an incoming shuriken. 

I personally use it more for immersion. Shots coming from all around just feels more "in it" rather than "it came 20deg to the right so I'll react immediately just from the sound localization and swing my rifle 20deg to the right automatically." In fact I use it more so when my 1st person cam is too focused on Total War at some point I'll just be surprised to hear hooves thundering behind me or to the flank, or when the situation is switched, my general/avatar is well past the isolated skirmishers but I'm still hearing clashing steel and crashing happening behind me as the rest of the cavalry squadron continues to trample on the hapless peasants their noble heavy infantry left unprotected.

 

[JohnnyReef]

Ok but crossfade requires the game to have a "headphone" option i guess? After alot of testing i feel doing 8 channel and then using dolby headphone is superior to just plain stereo. 

 

[obobskivich]

Ok but crossfade requires the game to have a "headphone" option i guess? After alot of testing i feel doing 8 channel and then using dolby headphone is superior to just plain stereo. 

You've hit on the bigger issue - what is the game (or other source) starting out with?

Some games have their own "headphone" mode which usually includes HRTF and positional audio features ("but I thought you needed an external processor?" not anymore, videogames have been using purely software-based audio for over a decade), and some do not (so stereo really is stereo, which will get you the "hard pans" you noted up above). So what some folks do, if their equipment supports it, is to try and treat the game like a movie with a surround sound track - you take 5.1 (or 7.1; you don't need "8 channels" for this to work - 5.1 will work just as well) and feed it into a headphone surround simulacra like Dolby Headphone (or DTS Headphone:X, or CMSS3D, or whatever you have) which both downmixes and processes it (using whatever proprietary HRTF + effects the given system uses) before outputting a stereo signal that is intended for headphones. Usually enabling something like SBX or CMSS3D or RSP or similar will set the internal output on your software to 5.1 (or 7.1), even though the hardware outputs remain 2ch (for headphones). You need the "surround sound output" from the software to allow the simulacra to do what it does.

  I wouldn't use the word "downmix" because that's not how it works, at least not in how "downmixing" traditionally works. That being, it just puts SL and SR on L and R, and anything on C ends up playing on both L and R.

There is actually downmixing going on, but its never a straight 1:1 map. Dolby and DTS decoders are required to use a pre-specified output matrix to convert x-channels input to y-channels output (where x>y). Most commonly this is probably done as 5.1 input to 2.0 output, but may be other maps (e.g. 5.1 to 3.1, or 6.1 to 4.0, etc). All channels are re-leveled for this downmix, however there's no HRTF or additional processing/effects being applied beyond that, like you would see with Dolby Headphone or Razer Surround. Depending on the input and output maps depends on whats actually done to each channel. Depending on the game this downmixing may be desirable over "straight stereo" (e.g. if there isn't any positional audio being worked out for stereo modes and there's no "headphone" option and there's no simulacra available on whatever is running the game).

Nothing should be lost with downmixing, and that's part of the other reason the downmix matrices exist - ideally everything that's coming from the source (e.g. the 5.1 bitstream from a DVD) is still played back, even if it isn't on the original output map that the source material intended. As we move into object-based surround (like Atmos) this will probably become even more streamlined, but as far as I'm aware object-based surround has not made it to video games yet.

+1 (or maybe even 2) to your description of headphone surround!

For the OP, here's a demo comparison of different simulacra in the same game:
https://www.youtube.com/watch?v=04yEtZJVpyY

You have to listen to it on headphones (if that wasn't obvious), but it should give you a good idea of differences between competing headphone surround simulacra. They are not all identical in their overall sound/performance, even though they're all attempting to do the same thing. Personally I'm a fan of Razer's Surround Pro because it is more or less hardware agnostic, and offers a lot of tweak-able options, while Dolby Headphone, CMSS, and SBX are tied to specific hardware packages (the much newer DTS Headphone:X is as well, and currently I think there's only one box that does it - the Turtle Beach Elite Pro - so you're limited by that).

 

[ProtegeManiac]

  Ok but crossfade requires the game to have a "headphone" option i guess?

 

What? Crossfade just has the end of one song overlapping with the intro of the next song. Don't you mean Crossfeed?

 

Either way that's different from whether the game has headphone option or not. If a game has a headphone specific audio program then it's basically virtual surround already embedded in the game's audio engine. No need to use any software nor hardware-based DSP tricks since it outputs 2ch with the channel crossfeed-ing to simualte both ears hearing all speakers and possibly also speaker in-room reverb simulation to simulate depth.

 

Now when the game doesn't have it what you need to do is set the game to 5.1 or 7.1, the soundcard DSP receives that, then does the 2ch conversion.
 

  After alot of testing i feel doing 8 channel and then using dolby headphone is superior to just plain stereo. 

 

Crossfeed, again, is not totally the same as virtual surround. While in some ways it's similar in how it distributes sound to both simulate hearing several speakers with both ears, it also adds a reverb to simulate depth.

Crossfeed only creates depth not from reverb simulating how speakers interact with a room but by simulating how both ears hear both drivers like how you hear both speakers in a 2ch set-up.

If the material you're listening to is 2ch then setting it to 8ch doesn't matter unless the computer somehow reprocesses it to eight, which is basically just more channels of whatever the heck that Dolby 4ch or 5ch stereo was (ie they clone L and R fronts to the rears, signals on both L and R come out of C, ie the vocals). Chances are what the DSP is more likely doing if you just set the game to 8ch is that it doesn't reprocess 8ch audio, it just takes 2ch audio and applies Crossfeed. If it doesn't revert to plain Crossfeed, it's still using Dolby Headphone to add reverb. Either way, I highly doubt that unless the material itself is 8ch (or 6ch) and without manually setting the DSP to make it 8ch (or 6ch) that the DSP would still bother to make a stereo program surround first when it can still apply the speaker to room interaction reverb simulation.

 

 

 

Nothing should be lost with downmixing...

 

Wasn't really my point. What I was saying was that virtual surround does more since it's not plainly making 2ch out of 5ch, but making a headphone less headphone-y.

 

+1 (or maybe even 2) to your description of headphone surround! 

 

I played Empire: Total War games with my brother and it all degenerates into both of us just quoting The Patriot and Waterloo (and I have my saber in the room). We thought we were really immersed into it until some guy posted a photo of his brother playing Rome II wearing a centurion's helmet.

 

[obobskivich]

Crossfeed, again, is not totally the same as virtual surround. While in some ways it's similar in how it distributes sound to both simulate hearing several speakers with both ears, it also adds a reverb to simulate depth.

+1. To add to this explanation, the idea behind crossfeed is to try and better recreate the experience of listening on stereo speakers. With stereo speakers in a room your left ear doesn't just hear the left speaker - it hears both speakers. And the same for your right ear. Whereas with headphones your left ear only hears the left driver, and the same for the right ear. So crossfeed attempts to "fix" that by "crossfeeding" the audio signal.

Headphone virtual surround (or whatever else we want to call it) is more like Dolby Virtual Speaker (to continue on the headphone v speaker analogy) - the goal is to try and recreate the experience of physical surround sound on your headphones, but still with only two output channels. It does this by applying HRTF and other DSP features (like reverb).

If the material you're listening to is 2ch then setting it to 8ch doesn't matter unless the computer somehow reprocesses it to eight,

This is what Dolby Headphone will do with 2ch input - it uses Pro Logic to generate the "surround sound" input for the Dolby Headphone feature. My understanding is that DTS Headphone:X does the same thing with DTS Neo to generate the input for its headphone surround feature from 2ch sources. Creative does something similar with CMSS. SBX and Razer Surround rely on the software/application outputting 5.1 (or 7.1), which is why you'll see the Windows audio settings change to 5.1 (or 7.1) when running either of those packages.

which is basically just more channels of whatever the heck that Dolby 4ch or 5ch stereo was (ie they clone L and R fronts to the rears, signals on both L and R come out of C, ie the vocals).

There is no "4 or 5ch stereo" in Dolby. There's Pro Logic and Pro Logic II, which apply transformation matrices to the 2ch input. Pro Logic I actually assumes that your source is matrix encoded (with Dolby Surround) and applies a matrix that would properly unfold that into 4ch (left, right, center, surround), while Pro Logic II is designed to take 2ch inputs and unpack them to 5.1 (or higher, depending on which revision of Pro Logic II is at play; e.g. IIx does 7.1). It does not just "clone" signals though - sends the signal through a matrix and band-pass equalization.

DTS Neo and CMSS offer similar functionality.

Chances are what the DSP is more likely doing if you just set the game to 8ch is that it doesn't reprocess 8ch audio, it just takes 2ch audio and applies Crossfeed. If it doesn't revert to plain Crossfeed, it's still using Dolby Headphone to add reverb. Either way, I highly doubt that unless the material itself is 8ch (or 6ch) and without manually setting the DSP to make it 8ch (or 6ch) that the DSP would still bother to make a stereo program surround first when it can still apply the speaker to room interaction reverb simulation.

Dolby Headphone will take 2ch input and run it through Pro Logic II in order to function, CMSS is similar. SBX and Razer Surround need a 5.1 input from the get-go (e.g. the game's output has to be 5.1). Otherwise yes its likely just crossfeed.

Wasn't really my point. What I was saying was that virtual surround does more since it's not plainly making 2ch out of 5ch, but making a headphone less headphone-y.

I was responding to part of the original question - I see it was in a confusing spot in my reply. Earlier in the thread it was asked if going from 8ch to headphones "loses data" so I was clarifying that, by design, it should not.

I played Empire: Total War games with my brother and it all degenerates into both of us just quoting The Patriot and Waterloo (and I have my saber in the room). We thought we were really immersed into it until some guy posted a photo of his brother playing Rome II wearing a centurion's helmet.

Good movies, good game, haha.

 

[ProtegeManiac]

 
Pro-Logic! There! Couldn't remember it and all the Googling for "stereo to surround" etc didn't have it come up for some reason.

 

[Swedindo]

Hi,
Imho, it really depends on how you want to use the sound image..
Here's my experience with a sennheiser Hd600 and Arctis 7 (gaming headset) which supports DTS 7.1.
I usually play the games in ps4. And no, I don't have any other 7.1 processor system other than my ps4.
I could only give you a clue on how they sound like.. I've compared them with their own cables even though Arctis 7 supports wireless too.

For gaming, I prefer the Arctis 7!
The sound quality is far ahead for gaming purposes. The most important part is the separation.. You can tell where a bullet is coming from, and if someone walk from where..
I could not clearly tell the difference on those events if I'm using the hd600.
Imho, you could save some $$ just by buying an Arctis 7 for example. And the great thing is even though I'm using the wireless connection, the separation is still there!
And don't worry about the sound quality, I still prefer the Arctis 7 compared to hd600 for base and sound stage.
All I can say is, try them for an audition in a shop or borrow them if you can.. Don't be amazed if you like the Arctis 7 more than your hd650 after that..

Cheers,


Sent from my SM-G935F using Tapatalk

 

[mindbomb]

When you set the windows mixer to 8 channel, windows will pad whatever the application outputs with silent channels if necessary to make it 8 channel. Dolby pro logic II requires 2 channel input, so it can't work when you have 8 channel pcm. Pro logic is also optional with 2 channel audio, you don't have to use it with dolby headphone. Matrix encoded content doesn't really exist anymore, so I don't see the point of pro logic. 
 
If you feed dolby headphone a stereo signal, it will try to recreate what things would sound like if it were playing through just the front two speakers of a surround sound system. This is functionally equivalent to a crossfeed. Both are trying to recreate what stereo speakers would sound like at that point.

 

[pinnahertz]

   Pro logic is also optional with 2 channel audio, you don't have to use it with dolby headphone. Matrix encoded content doesn't really exist anymore, so I don't see the point of pro logic. 
 

Legacy content.  The first 20 years of surround sound.

 

[obobskivich]

Pro-Logic! There! Couldn't remember it and all the Googling for "stereo to surround" etc didn't have it come up for some reason.

There's a lot of other competing solutions when you go "back in time" as well, and there actually is "Dolby Stereo" but it doesn't exist for consumers (if you're bored: https://en.wikipedia.org/wiki/Dolby_Stereo - if you read down enough it eventually does "grow into" Dolby Surround and later the consumer-facing Pro Logic).

When you set the windows mixer to 8 channel, windows will pad whatever the application outputs with silent channels if necessary to make it 8 channel.

If the application/source doesn't provide said channels, yes. It acts as a "straight thru" - so 2ch input with 7.1 output map you get 2ch to L/R and everything else is null. Microsoft doesn't pay to licence any matrix upmixing software, however many soundcards include such functionality in their drivers.

Dolby pro logic II requires 2 channel input

Pro Logic II can accept matrix encoded 2ch, "straight" 2ch input, or in the case of "extended" II (e.g. IIx, IIz) can accept 5.1 (or 7.1) inputs as well (to provide 6.1-9.1 output, as applicable). It will properly unpack matrix'd content through its matrix, and (per spec) should provide better channel separation while doing it (of course, it doesn't change that the source material's channel separation is whatever it is, so whether or not this is of real value is another question entirely).

Pro logic is also optional with 2 channel audio, you don't have to use it with dolby headphone.

My understanding is the Dolby Headphone package only accepts a multi-ch input for downmix/processing, so if you have a "straight" 2ch signal and want to enable Dolby Headphone, the decoder will run Pro Logic II (or on later iterations, IIx or IIz) internally to create that multi-ch source for the Dolby Headphone package to work on. Same applies for Lt/Rt mixed content, however in that case it will properly unpack the matrix encode before applying its DSP. This is how competing technologies work as well (or they exclusively require the source to originally be in multi-ch), and Dolby's marketing literature points to a similar cooperation between Pro Logic II and Headphone, although is quite short on specifics. If you have a "native" multi-ch signal (e.g. 5.1 output from a game or DVD) there is no Pro Logic II unless you're using IIx or IIz for "7.1" output.

In practice this does mean you can send 2ch into Dolby Headphone, and that its just taking care of everything behind the scenes, but if given the choice between starting out with 2ch and a multi-ch source, I'd take the multi-ch source on the assumption that it's either A) properly mixed/mastered for that (like a movie) or B) rendering the additional channels for that (like a game) as opposed to downmixing (or giving up outputs) and then trying to work back from that point.

Matrix encoded content doesn't really exist anymore, so I don't see the point of pro logic. 

"Its older than 5 minutes, it doesn't exist and never did!"

Surprisingly enough, matrix content is still very common even today, and you may not even notice it - any stereo downmix from a Dolby or DTS surround-sound source will generally be provided as Lt/Rt (matrix encoded) for legacy compatibility (this is a requirement for Dolby decoders, and very common, but not universal, for DTS-supporting equipment as well, although that doesn't mean it is enabled by default). Some games implement Pro Logic II internally as well, providing native Lt/Rt outputs (this is especially common on console games). And finally, some relatively newer movies even include a studio-mixed matrix stereo track on their home releases, primarily targeting headphone or late-night listening (Wall-E is an example). Pro Logic I can also be applied to "straight" stereo sources, although results will vary from source to source, setup to setup, and person to person in terms of "how good is it?" The basic Pro Logic matrix does a pretty good job with typical stereo though, especially TV/cinematic content that has more conventional mastering. For Dolby Headphone, matrix-encoded content will be properly handled through Pro Logic II internally.

If you feed dolby headphone a stereo signal, it will try to recreate what things would sound like if it were playing through just the front two speakers of a surround sound system. This is functionally equivalent to a crossfeed. Both are trying to recreate what stereo speakers would sound like at that point.

Crossfeed is simply attempting to "spoof" or "recreate" the ITD, with no consideration of HRTF or other DSP to try and recreate a given acoustic setting. Dolby Headphone relies on HRTF and its own proprietary FIR filters (and these are never "defeated") which are based on a few acoustic models (I think they're up to 3 or 4 now) that try to simulate sound in a specific environment (Dolby says they're based on in-room response of some idyllic speaker system). They are not "functionally equivalent" as a result. Also, to wit, there is no "universal crossfeed" - different manufacturers/designers implement crossfeed in different ways on different amps or in different software packages, with some including equalization as well, while Dolby Headphone is meant to be consistent across implementations as its a pre-packaged DSP routine.

 

[pinnahertz]

There's a lot of other competing solutions when you go "back in time" as well, and there actually is "Dolby Stereo" but it doesn't exist for consumers (if you're bored: https://en.wikipedia.org/wiki/Dolby_Stereo - if you read down enough it eventually does "grow into" Dolby Surround and later the consumer-facing Pro Logic).
If the application/source doesn't provide said channels, yes. It acts as a "straight thru" - so 2ch input with 7.1 output map you get 2ch to L/R and everything else is null. Microsoft doesn't pay to licence any matrix upmixing software, however many soundcards include such functionality in their drivers.
Pro Logic II can accept matrix encoded 2ch, "straight" 2ch input, or in the case of "extended" II (e.g. IIx, IIz) can accept 5.1 (or 7.1) inputs as well (to provide 6.1-9.1 output, as applicable). It will properly unpack matrix'd content through its matrix, and (per spec) should provide better channel separation while doing it (of course, it doesn't change that the source material's channel separation is whatever it is, so whether or not this is of real value is another question entirely).
My understanding is the Dolby Headphone package only accepts a multi-ch input for downmix/processing, so if you have a "straight" 2ch signal and want to enable Dolby Headphone, the decoder will run Pro Logic II (or on later iterations, IIx or IIz) internally to create that multi-ch source for the Dolby Headphone package to work on. Same applies for Lt/Rt mixed content, however in that case it will properly unpack the matrix encode before applying its DSP. This is how competing technologies work as well (or they exclusively require the source to originally be in multi-ch), and Dolby's marketing literature points to a similar cooperation between Pro Logic II and Headphone, although is quite short on specifics. If you have a "native" multi-ch signal (e.g. 5.1 output from a game or DVD) there is no Pro Logic II unless you're using IIx or IIz for "7.1" output.

In practice this does mean you can send 2ch into Dolby Headphone, and that its just taking care of everything behind the scenes, but if given the choice between starting out with 2ch and a multi-ch source, I'd take the multi-ch source on the assumption that it's either A) properly mixed/mastered for that (like a movie) or B) rendering the additional channels for that (like a game) as opposed to downmixing (or giving up outputs) and then trying to work back from that point.
"Its older than 5 minutes, it doesn't exist and never did!"

Surprisingly enough, matrix content is still very common even today, and you may not even notice it - any stereo downmix from a Dolby or DTS surround-sound source will generally be provided as Lt/Rt (matrix encoded) for legacy compatibility (this is a requirement for Dolby decoders, and very common, but not universal, for DTS-supporting equipment as well, although that doesn't mean it is enabled by default). Some games implement Pro Logic II internally as well, providing native Lt/Rt outputs (this is especially common on console games). And finally, some relatively newer movies even include a studio-mixed matrix stereo track on their home releases, primarily targeting headphone or late-night listening (Wall-E is an example). Pro Logic I can also be applied to "straight" stereo sources, although results will vary from source to source, setup to setup, and person to person in terms of "how good is it?" The basic Pro Logic matrix does a pretty good job with typical stereo though, especially TV/cinematic content that has more conventional mastering. For Dolby Headphone, matrix-encoded content will be properly handled through Pro Logic II internally.
Crossfeed is simply attempting to "spoof" or "recreate" the ITD, with no consideration of HRTF or other DSP to try and recreate a given acoustic setting. Dolby Headphone relies on HRTF and its own proprietary FIR filters (and these are never "defeated") which are based on a few acoustic models (I think they're up to 3 or 4 now) that try to simulate sound in a specific environment (Dolby says they're based on in-room response of some idyllic speaker system). They are not "functionally equivalent" as a result. Also, to wit, there is no "universal crossfeed" - different manufacturers/designers implement crossfeed in different ways on different amps or in different software packages, with some including equalization as well, while Dolby Headphone is meant to be consistent across implementations as its a pre-packaged DSP routine.

Great post!  
 
On Dolby Stereo...a bit of detail.  Because the base encoding matrix resulted in only 3dB of adjacent channel separation (but much greater non-adjacent channel separation), Dolby Stereo (LtRt, optical SVA soundtrack on film) in the theatrical version employed a steering logic similar to ProLogic, but less adept, to artificially improve adjacent channel sep.  Think of it as adaptive level gating that was matrix-aware.  It was created by using Tate Audio chip sets intended to apply similar steering in quadraphonic systems, but the use was modified for a LCRS channel plan.  In a Dolby Cinema Processor, they were on the Cat. 150 card.  The Tate chips had limited ability for adjacent channel separation of about 20dB at high levels, worse at low levels.  L and R to Surround crosstalk was further reduced by delaying S by enough time that the precedence effect (Haas) reinforced the L and R channels as dominant.  The surround channel was also limited to 7kHz and had modified Dolby B noise reduction applied. 
 
Dolby Surround was consumer targeted, and had no steering logic so adjacent channel separation was 3dB, but it did provide surround delay, and the requisite surround channel noise reduction (modified Dolby B).  It also appeared in network TV shows, billed as such, and those mixes were intended to be decoded by consumer Dolby Surround units.  
 
ProLogic, the first analog version, was similar in function to the Cinema Cat.150 card, but targeted at consumers and therefore permitted the creation of custom chips in quantity, which actually beat the performance of the original Dolby Cat. 150 cards by quite a bit, but still followed the same matrix.  There were some very expensive consumer surround products then, and Lexicon developed an early digital implementation of ProLogic that included Home THX processing as well.
 
The rest of the ProLogic flavors were improvements, DSP implementations, 6.1, 7.1 and height extractions, 2 channel music compatibility, etc., etc, but the encoded/decode matrix rules remain the basically same as the original Dolby Stereo (there are some minor differences).  So a Dolby Stereo track from 1980 will play through ProLogic II today even better than it did using a professional Dolby Cinema Processor of it's day.  
 
A key side-benefit of Dolby Stereo in theaters was that the installation of the processor also entailed 1/3 octave equalization of the LCRS speakers, and the application of Dolby A type noise reduction on the optical film soundtrack.  Those two features extended the usable frequency response and dynamic range of optical film sound almost another octave and another 10dB, presenting the semblance of high fidelity.  It was dramatic when compared to the typical Academy Mono tracks that were more popular at the time.  
 
And a bit of trivia: Because of a few anomalies of using dynamic steering logic, Dolby Stereo tracks were mixed while monitoring through a reference decoder.  One of the issues was that the steering logic took time to steer a sound to the right channel.  The time constant chosen was roughly syllabic.  That meant that short impulsive sounds would be mis-steered because the logic didn't respond fast enough.  These errors (called Toddisms after Craig Todd, the Dolby engineer who developed the system) were compensated for by mixing in an early low level sound that the logic could respond too in advance of the real desired sound.  Pre-steering, sort of.  Another issue was the tendency of the system to reduce moderately separated stereo to mono in the center channel.  Music was mixed for films with a somewhat wider than normal stereo spread.  
 
I once stood in a line at a theater waiting to buy a ticket.  A woman behind me was heard saying, "If Star Wars is sold out, do you want to see that other film, "Dolby Stereo?".  Confusing times.

 

[mindbomb]

There's a lot of other competing solutions when you go...

- snip -

 


I remember dolby pro logic for surround sound for analog television broadcasts. But that doesn't have relevance today. And there are so many codecs and interfaces that support surround sound, so matrix decoding is not necessary for surround sound. Dolby pro logic and dolby headphone are separate technologies, you don't have to use them together, even with 2 channel input. There actually were minidisc players that just had dolby headphone, and used it as a stereo crossfeed. A crossfeed is anything that at a minimum, mixes elements on a delay to represent itd. Since dolby headphone does that, it is a crossfeed.

 

[obobskivich]

Great post!  

On Dolby Stereo...a bit of detail.  Because the base encoding matrix resulted in only 3dB of adjacent channel separation (but much greater non-adjacent channel separation), Dolby Stereo (LtRt, optical SVA soundtrack on film) in the theatrical version employed a steering logic similar to ProLogic, but less adept, to artificially improve adjacent channel sep.  Think of it as adaptive level gating that was matrix-aware.  It was created by using Tate Audio chip sets intended to apply similar steering in quadraphonic systems, but the use was modified for a LCRS channel plan.  In a Dolby Cinema Processor, they were on the Cat. 150 card.  The Tate chips had limited ability for adjacent channel separation of about 20dB at high levels, worse at low levels.  L and R to Surround crosstalk was further reduced by delaying S by enough time that the precedence effect (Haas) reinforced the L and R channels as dominant.  The surround channel was also limited to 7kHz and had modified Dolby B noise reduction applied. 

Dolby Surround was consumer targeted, and had no steering logic so adjacent channel separation was 3dB, but it did provide surround delay, and the requisite surround channel noise reduction (modified Dolby B).  It also appeared in network TV shows, billed as such, and those mixes were intended to be decoded by consumer Dolby Surround units.  

Great info. Just to clarify myself: I'm using "Dolby Surround" to broadly refer to non-consumer/cinema formats, while "Pro Logic" encompasses the consumer-facing stuff of the late 1980s (1987 sounds right but may not be). I know that, technically speaking, Dolby has re-used "Dolby Surround" as a brand for many things, even including digital multi-channel, and frankly I find it unnecessarily confusing on their part. I do know that on VHS, for example, some are labeled as "Dolby Surround" to refer to Pro Logic compatible matrix encodes, but that audio track is not necessarily what was in theaters in the 1970s or 1980s, nor is the equipment decoding it directly comparable to what was in theaters either.

ProLogic, the first analog version, was similar in function to the Cinema Cat.150 card, but targeted at consumers and therefore permitted the creation of custom chips in quantity, which actually beat the performance of the original Dolby Cat. 150 cards by quite a bit, but still followed the same matrix.  There were some very expensive consumer surround products then, and Lexicon developed an early digital implementation of ProLogic that included Home THX processing as well.

Yamaha had their own go at this too, which implemented acoustic simulacra and grew into "CinemaDSP." Excluding some of the most recent receivers/processors, this all ran on custom ASICs. Per Yamaha's self-written history (so take that for its worth), they were the only ones to get Dolby's blessing for "enhanced" modes of Pro Logic (and later AC-3 5.1) that combined the Dolby decoder with their own DSP and channel steering. I think (but may be mistaken) they were also the first do do "height" channels (their literature calls it "presence"), but Lexicon/Harman may have been there at or around the same time.

The rest of the ProLogic flavors were improvements, DSP implementations, 6.1, 7.1 and height extractions, 2 channel music compatibility, etc., etc, but the encoded/decode matrix rules remain the basically same as the original Dolby Stereo (there are some minor differences).  So a Dolby Stereo track from 1980 will play through ProLogic II today even better than it did using a professional Dolby Cinema Processor of it's day.  

Pro Logic II is, as I understand it, an entirely distinct development, designed by Fosgate Labs and licenced over to Dolby; you're 100% that it will take the Dolby Surround/Pro Logic I matrix input and unpack it "correctly" for backwards compatibility, but the actual DPL-II matrix looks a bit different, one of the biggest changes is that the surround channels are discrete (it actually provides 5 channels out). And as you pointed out, channel separation is meant to be higher. However for actual "1980s content" I'm not sure if this makes much of an appreciable difference.

A key side-benefit of Dolby Stereo in theaters was that the installation of the processor also entailed 1/3 octave equalization of the LCRS speakers, and the application of Dolby A type noise reduction on the optical film soundtrack.  Those two features extended the usable frequency response and dynamic range of optical film sound almost another octave and another 10dB, presenting the semblance of high fidelity.  It was dramatic when compared to the typical Academy Mono tracks that were more popular at the time.  

Aye.

And a bit of trivia: Because of a few anomalies of using dynamic steering logic, Dolby Stereo tracks were mixed while monitoring through a reference decoder.  One of the issues was that the steering logic took time to steer a sound to the right channel.  The time constant chosen was roughly syllabic.  That meant that short impulsive sounds would be mis-steered because the logic didn't respond fast enough.  These errors (called Toddisms after Craig Todd, the Dolby engineer who developed the system) were compensated for by mixing in an early low level sound that the logic could respond too in advance of the real desired sound.  Pre-steering, sort of.  Another issue was the tendency of the system to reduce moderately separated stereo to mono in the center channel.  Music was mixed for films with a somewhat wider than normal stereo spread.  

Interesting.

I once stood in a line at a theater waiting to buy a ticket.  A woman behind me was heard saying, "If Star Wars is sold out, do you want to see that other film, "Dolby Stereo?".  Confusing times.

Ha!

  I remember dolby pro logic for surround sound for analog television broadcasts. But that doesn't have relevance today. And there are so many codecs and interfaces that support surround sound, so matrix decoding is not necessary for surround sound. Dolby pro logic and dolby headphone are separate technologies, you don't have to use them together, even with 2 channel input. There actually were minidisc players that just had dolby headphone, and used it as a stereo crossfeed. A crossfeed is anything that at a minimum, mixes elements on a delay to represent itd. Since dolby headphone does that, it is a crossfeed.

I'm not here for a semantic argument, sorry.

I provided some examples of real-world scenarios where matrix content may be found; just because your usage scenario doesn't align with that doesn't mean nobody else's does.