dynaquad diamond

[Quadrobineg]

Pro Logic I does what I tell it to. I turn the pan pot and set f/b and the image goes wherever I turn the pot.

nah, it just goes to "somewhere there".

in prologic I you have a gap of 150 degrees between left front and surround, and since direction is recognized by filtering and runtime difference between the ears it is quite hard to place a sound source at 75 degrees when there is no speaker somewhere near.

if the - basically only 3 - speaker of those systems at least had equal distances between the speakers... but even then quad setups would offer more precision when it comes to direction.

Remember that all of this digital sound stuff did not exist for the general public when Dolby Surround debuted in Star Wars.

most digital filters are much older than that, it was probably more a matter of cost what they have chosen.

but of course the digital implementations for phase shifting were quite bad in the 80ies, too, and beeing affordable for as many households as possible was the aim of these products.

I will always disagree with those that claim that sounds can't be properly panned in quad. There are many discussions about this in other threads but if the speakers are placed close enough together around the listener then PWM works just fine.

sure, if you compare it against other setups then you need to be at least so fair and use equal distances and consider only the center to be the hotspot.

you are still at 90 degrees then, which is a lot.

however, one should not only compare systems against each other in a professional studio. in a livingroom with lots of room reflections having 4 speakers compared to more than 4 is even worse.

given that eventually one´s main target is a binaural 2-channel downmix (and/or immersive formats are listened to using earphones, which is technically the same), you should also do a comparison on that domain.
to my ears, the difference between 4 and 10 speakers is huge.

maybe i should better move that to one of the threads you mentioned and add some sound examples. my encoder question has nothing do with that.

(it should be added that i am experimenting with synthesis with that, i.e. i modulate positions of complex soundfields at signal rate. no mixing desk or DAW can do anything of that. i first have to get used to think in traditional mixing terms when it comes to surround.)

With modern surround they always talk about speaker placement in terms of degrees. In the seventies speaker placement was never described in that way. Many different speaker arrangements were once suggested, none right or wrong. No hard and fast rules (I hate rules) it was always suggested to arrange the speakers dependant on your particular listening room and your own personal preference. It would seem that today you require a special home theatre room with the proper Dolby approved speaker placement!

having ceiling speakers is not a bad idea at all, and dolby´s suggestion to use 7.4.1. at a minimum might be reasoned, but i tend to ignore this modern development, too.

there is still enough to explore in a 2-dimensional world.

 

[Quadrobineg]

So are you wanting a huge number of tracks in a recording so you can use any playback method you want by choosing the correct tracks?

hm, no, what i tried to say is that i see no reason to take care about stereo and mono compatibility. i would just make 2 CDs, and for stereo and one to paly with surround decoders, and ignore how it is done normally. a second CD is 20 cents more and if you ignore stereo compatibility you have more options.

I am sorry to say that a 2-channel analog recording (or even a digital recording) is limited to the actual recording it was recorded with. You can't put multiple systems into a single recording.

in conjunction with 4:2:4 matrix conversions you would usually record 4 channels, no? neumann SM?

 

[kfbkfb]

https://quark.cykik.com/^^^
A free Sansui QS matrix encoder (QS can be decoded fairly well with the Hafler/DynaQuad passive speaker matrix and Dolby Pro-Logic 2 music mode).


Kirk Bayne

 

[MidiMagic]

nah, it just goes to "somewhere there".

in prologic I you have a gap of 150 degrees between left front and surround, and since direction is recognized by filtering and runtime difference between the ears it is quite hard to place a sound source at 75 degrees when there is no speaker somewhere near.

if the - basically only 3 - speaker of those systems at least had equal distances between the speakers... but even then quad setups would offer more precision when it comes to direction.

All the theory in the world does not counter the result of experiment.

I connected my encoding mixer to my monitoring system set for Dolby Pro Logic I, and I fed several sound sources into the mixer. As I panned each source around the room, the image I heard was where I panned the sound. It moved smoothly around the room as turned the panpot.

I did this for several people and all but one heard the image where I panned it. That one had hearing damage.

The listener had to be facing front.

I think the trick in doing this is the delay in the back channels.

When I tried QS settings, the imaged cogged between the LF and LB and between the RF and RB.

most digital filters are much older than that, it was probably more a matter of cost what they have chosen.

but of course the digital implementations for phase shifting were quite bad in the 80ies, too, and beeing affordable for as many households as possible was the aim of these products.



sure, if you compare it against other setups then you need to be at least so fair and use equal distances and consider only the center to be the hotspot.

you are still at 90 degrees then, which is a lot.

however, one should not only compare systems against each other in a professional studio. in a livingroom with lots of room reflections having 4 speakers compared to more than 4 is even worse.

given that eventually one´s main target is a binaural 2-channel downmix (and/or immersive formats are listened to using earphones, which is technically the same), you should also do a comparison on that domain.
to my ears, the difference between 4 and 10 speakers is huge.

maybe i should better move that to one of the threads you mentioned and add some sound examples. my encoder question has nothing do with that.

(it should be added that i am experimenting with synthesis with that, i.e. i modulate positions of complex soundfields at signal rate. no mixing desk or DAW can do anything of that. i first have to get used to think in traditional mixing terms when it comes to surround.)

hm, no, what i tried to say is that i see no reason to take care about stereo and mono compatibility. i would just make 2 CDs, and for stereo and one to paly with surround decoders, and ignore how it is done normally. a second CD is 20 cents more and if you ignore stereo compatibility you have more options.

in conjunction with 4:2:4 matrix conversions you would usually record 4 channels, no? neumann SM?

I do not make any 4-channel master recording and then matrix it down. More errors occur in matrixing a 4-channel tape down to a matrixed 2-channel recording using the specified "4-corners" matrix encoder. That is where the errors occur.

I create surround mixes directly in the encoding mixer, mixing the parts together and panning them where they go all at once, and hearing where they are panned in the Dolby Surround monitor system.

And I design the mix so it can be listened to in mono, stereo, Dolby Surround, and any RM system. Single inventory for all.

 

[Quadrobineg]

All the theory in the world does not counter the result of experiment.

as well as perception won´t beat the laws of physics.

to find out which of two systems works better it might not be enough to try out only one of them, rather you would have to compare them against each other.

I connected my encoding mixer to my monitoring system set for Dolby Pro Logic I

i am of course most sceptical about the prologic I setup with its wide angels, yet i have to admit that it gives you results better than you would expect from it in theory, and is somehow a clever system.

however, i notice the difference between 4 and 10 speakers easily, probably because i dont do surround techniques with it.

while the conversion of channel based soundfields to traditional surround formats will be lossy, it could be an interesting experiment to try to get the best of both worlds together.

 

[Quadrobineg]

how would the circle surround matrix look like? compared to pro logic II? i might add those two even they are not "4.0" anymore.

CS patend is expired i guess? and also does not have steering, right? like PL II.

 

[par4ken]

how would the circle surround matrix look like? compared to pro logic II? i might add those two even they are not "4.0" anymore.

CS patend is expired i guess? and also does not have steering, right? like PL II.

They are similar, I would have to check exactly how close they are. Circle Surround does use steering logic. Here is one example, the Gemini SP-1. The Gemini can be found for sale on eBay for bargain prices. IMHO the best deal in a surround decoder!

 

[Quadrobineg]

moment, i thought circle surround is always 6.x?

 

[ar surround]

Let me ask you guys something about extracting out-of-phase information. The simple Hafler circuit extracts out-of-phase information and routes it to the rear channels. Is there any way of using a DAW such as Audacity to extract this difference signal? I know that it is easy to separate the center (mono L-R material) and remove the center using the "Vocal Reduction and Isolation" effect in Audacity, but it is not the same as extracting out-of-phase infomation.

 

[par4ken]

Let me ask you guys something about extracting out-of-phase information. The simple Hafler circuit extracts out-of-phase information and routes it to the rear channels. Is there any way of using a DAW such as Audacity to extract this difference signal? I know that it is easy to separate the center (mono L-R material) and remove the center using the "Vocal Reduction and Isolation" effect in Audacity, but it is not the same as extracting out-of-phase infomation.

Yes it is the same. You can extract the difference signal as well.

 

[ar surround]

Yes it is the same. You can extract the difference signal as well.

Now that I see you state this in two simple sentences, I say to myself, "Of course it. Doh!"

 

[par4ken]

moment, i thought circle surround is always 6.x?

There are many different versions of Circle Surround including six channel. There is even a Cinema mode that uses the 7 Khz filter. The drawing that I showed shows 4.0 (L R C S) as well as 5.1. There is another version with variable encode coefficients. The White Paper about 5,2,5 from Discord (Ty Chamberlain) can be found here. Sadly Ty passed some time ago.

https://issuu.com/disclord/docs/circle_surround_technology/2​

 

[par4ken]

Coefficients for Circle Surround 4:2:4 (L,R,C,S) would be as follows, (j represents a 90° phase shift.)

Lt=L+0.707C-J0.707S
Rt=R+0.707C+j0.707S

Coefficients for Circle Surround 5:2:5 (L,R,C,Ls,Rs) would be as follows (j represents a 90° phase shift.)

Lt=L+0.707C-j(0.707Rs+0.549Ls)
Rt=R+0.707C+j(0.707Ls+0.549Rs)

From wikipedia, Dolby surround (L,R,C,S) looks almost the same but shows the j terms reversed.

Lt=L+0.707C+j(0.707S)
Rt=R+0.707C-j0.707S)

And Dolby PL II (L,R,C,Ls,Rs), the Wikipedia article (I believe) is incorrect about the C coefficient of PL II, it should be the same as for Dolby Surround (which is 0.707). Corrected below,

Lt=L+0.707C+j(0.866Rs+0.5Ls)
Rt=R+0.707C-j(0.866Ls+0.5Rs)

According to OD, Dolby PLII has the coefficients of 0.76 and 0.24 instead, The -j is on the left side instead of the right. More recently in his blog he stated that those given by Dolby were incorrect! He had the evidence but was not willing to share it.

Lt = L-j(0.76 Ls+0.24Rs)
Rt = R+j(0.76Rs+0.24Ls)

https://en.wikipedia.org/wiki/Matrix_decoder#Dolby_Stereo_and_Dolby_Surround_(matrix)_4:2:4https://www.quadraphonicquad.com/forums/threads/sq-encoder.16681/page-2

 

[Quadrobineg]

thanks again. so i was completly on the wrong path.

it is not that i would not have tried to google it. but where dolby seem to have the habit to give out incomplete info after patent protection is gone, for DTS you can´t even find some.

your assumption that the C should be sqrt(1/2) is well reasoned and i will follow it.

what remains open is how CS 6.0 would look like. if a single surround channel is the side signal and two surround speakers are the side signal and and the side signal inverted, it is not too obvious what the rear center would be in a 6.0 setup with three surround speakers.

 

[par4ken]

i would think that with 6.0 you are just adding a centre back speaker. It would be mixed equally into the surrounds 3dB down (0.707). Same as is done the front.

 

[kfbkfb]

One note - AFAIK, (original) Dolby Pro-Logic is no longer offered in new Audio Video receivers (since about 2014?), instead something called Dolby Surround Upmix is included which doesn't decode (original) Dolby Surround encoded content.


Kirk Bayne

 

[MidiMagic]

as well as perception won´t beat the laws of physics.

to find out which of two systems works better it might not be enough to try out only one of them, rather you would have to compare them against each other.

I did. I just reported the Dolby Surround results because it worked so well.

- Hafler Diamond - Behaved a lot like Dolby Surround. The panned sound moved smoothly around the listener.
- Dynaquad - Has a small cog effect just behind each front speaker.
- Stereo-4 - Cogged between the front and back speaker.
- QS - Cogged between the front and back speaker.
- SQ* - Cogged between front and back speakers for side pans and diagonal splits.
- Dolby Surround - smooth pan all around. Back has treble cut.
- Dolby Pro-Logic I - Smooth pan all around.
- Dolby Pro-Logic II - Minor cogging effect with certain sounds.
- Discrete (2 channel strip pans and a crossfader) - Cogging between front and back.

*Used 2 SQ decoders connected as an encoder fed by my discrete setup above.

Systems not listed were because I had no encoder or decoder.

 

[MidiMagic]

One note - AFAIK, (original) Dolby Pro-Logic is no longer offered in new Audio Video receivers (since about 2014?), instead something called Dolby Surround Upmix is included which doesn't decode (original) Dolby Surround encoded content.


Kirk Bayne

Sounds like Microsoft designed it. No backward compatibility.

 

[kfbkfb]

https://www.lifewire.com/dts-neo-6-1846892^^^
...with a soundtrack that only provides two channels of information, DTS Neo:6 can expand the sound field to 6.1 channels.


With Dolby Pro-Logic apparently disappearing from A/V receivers, DTS Neo:6 seems like the only widely available matrix based decoder (& fake surround sound system) in A/V receivers - how does DTS Neo:6 sound decoding QS, SQ, H and some of the other less used matrix surround sound encoding systems?


Kirk Bayne

 

[Quadrobineg]

I did. I just reported the Dolby Surround results because it worked so well.

- Hafler Diamond - Behaved a lot like Dolby Surround. The panned sound moved smoothly around the listener.
- Dynaquad - Has a small cog effect just behind each front speaker.
- Stereo-4 - Cogged between the front and back speaker.
- QS - Cogged between the front and back speaker.
- SQ* - Cogged between front and back speakers for side pans and diagonal splits.
- Dolby Surround - smooth pan all around. Back has treble cut.
- Dolby Pro-Logic I - Smooth pan all around.
- Dolby Pro-Logic II - Minor cogging effect with certain sounds.
- Discrete (2 channel strip pans and a crossfader) - Cogging between front and back.

*Used 2 SQ decoders connected as an encoder fed by my discrete setup above.

Systems not listed were because I had no encoder or decoder.

to find out how 6, 10 or 24 speakers will sound compared to 4 it will not be enough to only compare systems with 4 speaker against each other.

pro logic I is surpringly good (and i am not saying that one would need 7th order ambisonic to record pop music and place 3 guitars left and right) but if you want an exact reproduction of two-dimensional direction, you cant have enough speakers.

the only alternative is object based formats, which means having filters in the decoder, and even then 7.4.1 reproduces stuff better than 4.0 or mono.

beside the direction/number of speakers thing (which is a bit a matter of stereo matrix vs. channel based discussion after all) i can assure you that for certain applications you will also love when a system is uniform im most aspects - starting with the setup geometry - because that makes controlling and programming it a bit simpler.


regarding PL II, which repeats the wrong ideas of SQ (i am trying to avoid terms like "ripoff" or "buyout" here), we probably share the same opinion.

left/right symetry is the absolute minimum i want to see and it is debunking that no DAW supports those three different types of pan/upmix/encoder for PL II in the way they existed (in theory) for SQ, it would be asked too much from the average user to take about this stuff.