3 Speakers For Stereo Content?

[SadMonster]

Hi, this has been on my mind alot. Anyways...

My understanding is that the precedence effect or "stereo" effect of having 2 speakers is not as good as actually having a dedicated center channel.

But the problem is that a huge amount of content both movie and music (especially music) is available only in stereo.

But couldn't we use some sort of DSP or something to playback stereo content on 3 speakers (left, right, center) with content that is identical to both channels playing on the center speakers?

I actually don't think this would work at all with my setup because a center speaker would have to be mismatched from the left and right because of the size and it wouldn't be positioned well because the TV would be blocking the position it should be in so I think any potential benefits from a setup like this in my situation would be drowned out by the negatives.

But still I'm curious if this is a good idea and if there are any products that let you do this and if anyone is doing it...

THANKS FOR YOUR TIME!!!

 

[Kal Rubinson]

But couldn't we use some sort of DSP or something to playback stereo content on 3 speakers (left, right, center) with content that is identical to both channels playing on the center speakers?

Here you go: https://www.sst.audio/new-products/trinaural-processor

 

[Ron Texas]

Here you go: https://www.sst.audio/new-products/trinaural-processor

Only $3k for that baby.

 

[Kal Rubinson]

Only $3k for that baby.

Yeah, sure. One can also do it with a little DSP manipulation depending on what tools are available.

 

[abdo123]

Most AVRs have dolby pro logic.

 

[Kal Rubinson]

Most AVRs have dolby pro logic.

That's too easy.

 

[daftcombo]

What is an AVR or a sub with "left", "right", "center" outputs doing, exactly?

 

[KSTR]

Here you go: https://www.sst.audio/new-products/trinaural-processor

That's the design from the late James Bongiorno.
Here is another commercial offering (probably out of production, as is the Bongiorno?): https://www.milestech.com/m44.htm

 

[KSTR]

What is an AVR or a sub with "left", "right", "center" outputs doing, exactly?

Typically, those are "steering" devices, Dolby stuff, with some intelligence to extract the center channel.

The trinaural matrix is simple and static and can be realized very easily.
All of this is based on Michael Gerzon's "Trifield" process : http://decoy.iki.fi/dsound/ambisonic/motherlode/source/7038.pdf
There is a Trifield plugin for Foobar2000 : https://www.foobar2000.org/components/view/foo_dsp_trifield, but don't recall if I've tried it or not.

 

[daftcombo]

Typically, those are "steering" devices, Dolby stuff, with some intelligence to extract the center channel.

The trinaural matrix is simple and static and can be realized very easily.
All of this is based on Michael Gerzon's "Trifield" process : http://decoy.iki.fi/dsound/ambisonic/motherlode/source/7038.pdf
There is a Trifield plugin for Foobar2000 : https://www.foobar2000.org/components/view/foo_dsp_trifield, but don't recall if I've tried it or not.

Thanks. What would be a cheap DAC working with it? (at least 4 channels simultaneously I guess)

 

[abdo123]

That's too easy.

less time having a mental break down is more time spent listening to music.

 

[KSTR]

Actually, you need only two channel for new L and R channels (L'=L-R/2, R'=R-L/2), and the Center is the sum of them (or the average, need to check) and the summing can be done passively (resistors).
The new L and R can be set up with a simple matrix plugin: https://www.kvraudio.com/product/channel_matrix_2x2_by_rs_met.
Or directly in the DAC if it's an RME ADI-2 DAC or ADI-2 Pro.
Some headroom issues must be adressed (as L-R/2, etc, could be 1.5x digital full scale, worst case).

 

[Kegemusha]

Something like this?

 

[abdo123]

Actually, you need only two channel for new L and R channels (L'=L-R/2, R'=R-L/2), and the Center is the sum of them (or the average, need to check) and the summing can be done passively (resistors).
The new L and R can be set up with a simple matrix plugin: https://www.kvraudio.com/product/channel_matrix_2x2_by_rs_met.
Or directly in the DAC if it's an RME ADI-2 DAC or ADI-2 Pro.
Some headroom issues must be adressed (as L-R/2, etc, could be 1.5x digital full scale, worst case).

I remember the last time you mentioned this a year ago i was very interested so I tried it.

Apparently L' and R' are basically the same thing with the phase inverted.

If you are sitting EXACTLY in the front of the speaker, you can listen something close to the original stereo picture "expanded" on the three speakers, but this is not a valid way to derive the channels because nobody has the speaker perfectly tuned and centered. Furthermore if you tilt or rotate your head you'll listen very strange Phase effects because L=-R and vice versa.

I tried something very close, but a little bit better for the phase:

L'=L-0.5*R
C=(L+R)*.5
R'=R-0.5*R

this reduces the phase inversion and preserve center separation and stereo picture.
It's not perfect but a good compromise IMO.

 

[KSTR]

L'=L-0.5*R
C=(L+R)*.5
R'=R-0.5*L

Yep, that's exactly the correct matrix equations.

 

[SadMonster]

So does anyone have any experience on if this actually sounds better though?

 

[KSTR]

@SadMonster,
It does, but probably not for everyone. Soundstage is different (larger, better layered) , sweetspot is larger and less prone to collapse with head movement/rotation, mono center is "nailed" (and doesn't have the usual elevation effect), the room and the speaker "disappear" better, etc. Basically, it exchanges the roles of real and phantom images. In normal stereo, center is a phantom and fully panned left or right are real sources, with trinaural it's the other way round. One needs to get used to it for a while, certainly.
It's a different but nonetheless pretty compatible projection mechanism for two-channel content. Some content based on HRTF trickery like Q-sound does not fully render "correctly" as it does with 2 speakers, though.

Technically, it does things that are somewhat "sketchy", for example the summings for the three speakers can generate "comb filtering" right in the speakers that don't exist in normal 2-speaker setup, but that is not a penalty per se as it gives rise to HRTF-triggered "false" image height impressions and such, but I really don't see this as a penalty as it seperates the various image sources in a mix better by this. Part of the effect is that pure X/Y content (intensity level difference between channel) are rendered noticably different than A/B content (time delay between channels, carrying almost the same signal, level-wise). Only the A/B content has those comb-filter effects. Further, decorrelated content like reverb has a more 3D-quality.

Additionally, it renders mixing deficiencies better, the typical one is panning a lot of sources (vocals, bass drum, bass guitar) 100% dead center (very common in digital mixing as it is so precise) and most of the rest hard-left/right and little content on the whole in-between area. This I found much more cleary rendered in trinaural as the center image is so solid.

Stereo playback is an arbitrary "illusion engine" after all, and everything that creates a better, more palpable "life-like" illusion for you is "better" and "more correct" by definition. But, as I said, YMMV.

 

[KSTR]

The biggest problem for implementation is three 100% equal speakers, especially wrt phase response. DRC (full correction of frequency response, including phase, at the listening position) comes to rescue in this case but then 3 independent DAC channels are required. I would recommend DRC anyway, it is always a big improvement.

 

[SadMonster]

The biggest problem for implementation is three 100% equal speakers, especially wrt phase response. DRC (full correction of frequency response, including phase, at the listening position) comes to rescue in this case but then 3 independent DAC channels are required. I would recommend DRC anyway, it is always a big improvement.

A little off topic but the improvement in clarity from Dirac is so big that I imagine if your budget for a stereo is 1000 or less, you would be better off with half your budget going to Dirac and the other half going to cheaper speakers versus just getting 1000 dollar speakers which might be a little better than the 500 dollar speakers.

I think I would rank the improvement on par with 2 subwoofers instead of 1.

More on topic, would Dirac be able to calibrate a 3 speaker stereo system? Kind of a Dirac noob still.

 

[ob1]

I find it certainly worth checking! May I kindly ask for your guidance, notably KSTR and Abdo123?

1. Initially, Abdo123 quoted the matrix, but I believe there is a typo. I am not nitpicking, I would like to make sure I am doing right since I really want to try it. Correct me if I'm wrong, but I believe that KSTR corrected the typo and that the matrix should be as follows, if would kindly confirm?
L'=L-0.5*R
C=(L+R)*.5
R'=R-0.5*L (and not R'=R-0.5*R, right?)

2. I am not familiar at all with multichannel routing, having always played back 2 channel content on a stereo system. My understanding was RME ADI-2 DAC had only two independent channels, how would the C be assigned then? Please excuse my dumb questions. I have a Gustard X16 DAC which also has xlr and rca outs, would I be able to do it with a 4 channel amp (say a 4 channel Buckeye, although only 3 channels are required)? I am currently using JRiver and Acourate, but if JRiver could handle it, that would be awesome! I could definitely try Foobar, but I thought it was worth asking.

Thank you in advance for your time and thank you all for pointing me to this very interesting rendering!

Olivier