• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

AdamG

Helping stretch the audiophile budget…
Moderator
Forum Donor
Joined
Jan 3, 2021
Messages
4,748
Likes
15,733
Location
Reality
The title of the thread is the use of centre speaker on stereo “music,” where the speaker is fed via a sound processor (up mixer). I think the use of a physical or phantom centre speaker used for a multi-channel film soundtrack is a different discussion.
Perhaps you missed where He mentions Movie and Tv use in his 2nd paragraph. Only until I asked him in post #31 did he define usage breakdown. So, I was under the initial impression that this was going to be a mixed usage setup. I only wish Thread Tittles were all accurate.
While planning my surround sound setup
This was his 1st sentence. Leading me to further think that this is not just a 2 channel stereo setup.
 

sarumbear

Master Contributor
Forum Donor
Joined
Aug 15, 2020
Messages
7,604
Likes
7,324
Location
UK
Uhm, I'm not sure what you're asking here. Extracting individual sounds from a stereo mix is exactly what upmixers do. Are you unaware of how they work?
My understanding of “individual” sound is separating the lead guitar, the piano, the drums and the singer. This is what was done in the film Get Back and the album Revolver 2022 Mix. They used AI and the process is not real-time.

On the other hand up mixers re-pan the stereo to multiple channels. The signal is separated into direct and uncorrelated ambient signals then re-pan to multiple channels. Their algorithms does not separate individual sounds, they just detect levels in frequency bands and reverb.

You may want to read this explanation.

This is how Dolby explains their upmixer.
The Dolby Surround upmixer is based on phase and gain relationships of elements in the signal, but importantly employs wideband functionality that analyzes and processes multiple perceptually spaced frequency bands in the signal.
 
Last edited:
OP
juliangst

juliangst

Addicted to Fun and Learning
Joined
Dec 11, 2021
Messages
976
Likes
1,000
Location
Germany
not just a 2 channel stereo setup
Correct. I will get an AVR anyways because of build in streaming, room correction and bass management.

I want to experience atmos music on apple music and also upmix my music to surround for that extra ambience.

But I don't know if I should also add a center channel.
It could help with the HRTF cancellation effects but it could also make the sound stage worse depending on the upmixer.

I also said that for dolby atmos music it makes absolutely sense to add a center channel but I have to withdraw that assumption. I extracted some dolby atmos tracks using blackhole and audacity and not all tracks use the center channel the same way.
Some don't use it at all, others only mix vocals withouts intruments to it (at a extremely reduced level) and on sometracks it's basically the same information as L+R.

I think it will be worth it to get the surround channels (and maybe heights) to experience atmos music but not really necessary to get a center just for the limited amount of center channel information in atmos music.

So in conclusion: I will stick to stereo unless a center channel with an upmixer can significantly improve the sound and get rid of those HRTF cancellations.
 

ocinn

Senior Member
Joined
Sep 14, 2022
Messages
377
Likes
926
Location
Los Angeles, CA
Yep. Although I don't think using only the differences between L&R, to L and Right would work well....unless the common signal was split into some high pass and low pass elements.

Here's an example of one of the matrices I employ for LCR . (snipped from the Elias Pekonen link in #26)
Yeah, it was something like this! I remember him talking about coefficients. The box he built to do it was all analog, custom PCBs, etc. I remember he explained how it had digitized pots controlling a resistor ladder to adjust settings and segment displays reading out the numbers. He had an AP Sys-2722 so I assume it was well-implemented. He had some sort of (MIDI?) interface in it as well so his computer would adjust the presets song-to-song for personal preference
haha too funny !!

I get accused of being a bit eccentric too.... thank providence for a beautiful wife that helps me defend this DIY mess to guests! :p
The combo of the realization you are doing this all in Q-Sys and the tools on the felt just about gave me a panic attack LOL.

Do you use the SH96HO style one as the center or is it the blue one above? I love syns, when I do freelance parties/raves I exclusively deploy them. Flawless coverage over the dancefloor.

IMG_7402.jpg

(during setup, it was aesthetically polished when done). Subs are beam-steered arc-array and there are SM80s flown 30ft down as delays, 2x Linea ASC48s.
 
Last edited:

Jon AA

Senior Member
Forum Donor
Joined
Feb 5, 2020
Messages
466
Likes
907
Location
Seattle Area
My understanding of “individual” sound is separating the lead guitar, the piano, the drums and the singer. This is what was done in the film Get Back and the album Revolver 2022 Mix. They used AI and the process is not real-time.
Of course they don't do that. They separate "individual" sounds based upon location. If they're mixed to the same location, the upmixer will put them in the same location. If the piano (or pipe organ, which is often the case) is behind you, it will place that individual sound there, where it should be. If most of the room reverb/ambiance is coming from behind you, they should place it there.

Nobody is talking about using an upmixer to break out individual stems/multitracks so we can re-mix the music/movie.
 

sarumbear

Master Contributor
Forum Donor
Joined
Aug 15, 2020
Messages
7,604
Likes
7,324
Location
UK
Of course they don't do that. They separate "individual" sounds based upon location. If they're mixed to the same location, the upmixer will put them in the same location. If the piano (or pipe organ, which is often the case) is behind you, it will place that individual sound there, where it should be. If most of the room reverb/ambiance is coming from behind you, they should place it there.
How do you locate a sound as behind you on a stereo track?
 

elvisizer

Active Member
Joined
Sep 16, 2018
Messages
264
Likes
212
It could help with the HRTF cancellation effects but it could also make the sound stage worse depending on the upmixer.
seems like an easy fix is to have the center for surround, then see how stereo sounds with it. if you don't like it, take it out of the upmixer config and just use it for surround.
 

gnarly

Major Contributor
Joined
Jun 15, 2021
Messages
1,037
Likes
1,472
Yeah, it was something like this! I remember him talking about coefficients. The box he built to do it was all analog, custom PCBs, etc. I remember he explained how it had digitized pots controlling a resistor ladder to adjust settings and segment displays reading out the numbers. He had an AP Sys-2722 so I assume it was well-implemented. He had some sort of (MIDI?) interface in it as well so his computer would adjust the presets song-to-song for personal preference

Awesome, it always sticks with me, the memory of encountering a master. He sounds like he was the real deal.
Analog masters like John Meyer, Tom D, etc amaze me. They see it from the roots up.

The combo of the realization you are doing this all in Q-Sys and the tools on the felt just about gave me a panic attack LOL.
I'd have no chance if it weren't for flexible DSP like Q-sys, and the relative simplicity of using linear phase Lol.

Do you use the SH96HO style one as the center or is it the blue one above? I love syns, when I do freelance parties/raves I exclusively deploy them. Flawless coverage over the dancefloor.

View attachment 292317
(during setup, it was aesthetically polished when done). Subs are beam-steered arc-array and there are SM80s flown 30ft down as delays, 2x Linea ASC48s.
Nice looking setup, i bet it sounds great.
Do you use the LIR xovers in the ASC48? I really like them...and found they didn't provide more delay than could go unnoticed, on a few live gigs.

In the pict I showed, the 3 blue horns are my LCR rig...all the same.
The white box is my current experimental prototype.
Whenever I think/feel I hear a quantum level improvement in my DIY syns, I usually stop and build a pair for stereo.
They blue guys earned a threesome haha. Although the newer white one is approaching quantum leap status.
 

ocinn

Senior Member
Joined
Sep 14, 2022
Messages
377
Likes
926
Location
Los Angeles, CA
I'd have no chance if it weren't for flexible DSP like Q-sys, and the relative simplicity of using linear phase Lol.
You could do it with software (Acourate, etc...) But I have always hated the thought of a Windows PC being the "heart" of the system. Q-Sys is way more elegant. Shame they are killing off their traditional I/O options and essentially forcing the consumer to buy networked amps or their under-performing I

Do you use the LIR xovers in the ASC48? I really like them...and found they didn't provide more delay than could go unnoticed, on a few live gigs.
Of course. Total pipeline latency needs to be kept at a minimum at these things because of DJ beatmatching and tight audience proximity. Festivals are easy to get away with FIR because the audience can't hear the stage monitors/iem feeds. As long as there aren't so many taps used that it starts to visually become an issue (but nowadays, with the IMAG screens, the video operator can just delay the live feeds and fool 99% of the crowd since almost everyone watches the screens)

I always wondered how the Marani "minimum phase" implementation works vs Linea. They both use the same marketing jargon so I was thinking Marani licensed it.
 

sarumbear

Master Contributor
Forum Donor
Joined
Aug 15, 2020
Messages
7,604
Likes
7,324
Location
UK
Largely phase manipulation. It can be done artificially or naturally. Watch this video with Dobly Surround Upmixer on:

The video is unavailable but you didn't understand my question. You said "If the piano (or pipe organ, which is often the case) is behind you, it will place that individual sound there, where it should be." How can the piano can be behind the listener on a stereo track?
 

Jon AA

Senior Member
Forum Donor
Joined
Feb 5, 2020
Messages
466
Likes
907
Location
Seattle Area
The video is unavailable but you didn't understand my question. You said "If the piano (or pipe organ, which is often the case) is behind you, it will place that individual sound there, where it should be." How can the piano can be behind the listener on a stereo track?
I understood your question perfectly. You didn't understand the answer. Hear it happen in the above video (just as one great example). I just tried the video on three different devices, it plays fine on all of them.
 

sarumbear

Master Contributor
Forum Donor
Joined
Aug 15, 2020
Messages
7,604
Likes
7,324
Location
UK
I understood your question perfectly. You didn't understand the answer. Hear it happen in the above video (just as one great example). I just tried the video on three different devices, it plays fine on all of them.
1686780646414.png

Maybe the video is geo-locked. I'm in the UK.

If the video is about a demo of an upmixer, I don't need to view it. My HT is based on a Trinnov Altitude-16 and has 16 speakers in a 9.3.4 layout. I have access to most upmixers.

My question is simple. You mention a piano sound located at the back on a stereo track. I am asking how can you hear a piano located at the back using just two speakers, which are in the front.
 

Jon AA

Senior Member
Forum Donor
Joined
Feb 5, 2020
Messages
466
Likes
907
Location
Seattle Area
Maybe the video is geo-locked. I'm in the UK.
A shame, that was a Chesky Records recording with David Chesky walking a circle around the mics playing a tom-tom. Very cool with an upmixer.
My question is simple. You mention a piano sound located at the back on a stereo track. I am asking how can you hear a piano located at the back using just two speakers, which are in the front.
No, your question was how it was located on a track. I told you. My whole point was it works dramatically better with an upmixer and more speakers than it does downmixed into two speakers. With only two speakers, depending upon your speakers and your room (because you're attempting to use the room as additional speakers) and how susceptible your ears/brain are to psychoacoustic tricks, you may kinda-sorta-maybe hear a similar effect. That's what the artist who mixed the track and put the decorrelated signal in there is counting on.

With an upmixer, it will identify those signals. In this case, DSU will extract that "individual sound" from the left and right speakers and place it in the speakers that correlates with its correct location--which works dramatically better than trying to trick your brain into hearing it in the correct location (and with the correct timing) with only two speakers.

Here's a very simplistic one, give this a try with DSU on:


If you have a 7.1 base layer, you'll see 100% of the sound is coming from the rear speakers (except for subs, of course). So that's how you'd mix the piano (if you knew your listeners would be using an upmixer--otherwise that may not be a wise choice). If you switch to stereo mode, you may kind of hear a similar effect (again, depending upon your speakers, room, etc) but the quality is degraded significantly even if you do perceive the sound as coming from behind you.
 

sarumbear

Master Contributor
Forum Donor
Joined
Aug 15, 2020
Messages
7,604
Likes
7,324
Location
UK
With only two speakers, depending upon your speakers and your room (because you're attempting to use the room as additional speakers) and how susceptible your ears/brain are to psychoacoustic tricks, you may kinda-sorta-maybe hear a similar effect. That's what the artist who mixed the track and put the decorrelated signal in there is counting on.
OK, I understand but you must admit that is a specialist example. How many stereo recordings rely on psychoacoustic tricks that upmixing will uncover? Besides, if you are mixing the stereo track specifically for upmixing, why not master in multi-channel in the first place?
 

dlaloum

Major Contributor
Joined
Oct 4, 2021
Messages
3,163
Likes
2,428
How do you do that? How can you extract individual sounds from a stereo mix? Unless you use an AI process like Peter Jackson used on the film Get Back or like Giles Martin used on Revolver 2012 remix all upmixes alter the tonality of the original sound as they use filters.
Listening to it in an environment other than that in which it was originally mastered, also alters the mix....

We impose our room, our speakers, our furniture and our gear on the original...

Various types of processing (speaker EQ, Room EQ) are used to try to reduce the impact of our environment on the original intent/recording.

A center speaker, and a matrix to take it from Stereo 2 channel to Stereo 3 channel - is just another step on this chain.

Done properly it has great benefits, as the stereo effect is no longer limited to the sweet spot, but becomes something that is discernible throughout the listening space - we end up with a Left, Center and Right, and the image no longer collapses into the closest speaker when we move around the listening space.

Dolby PLII and Lexicon (Harman) Logic 7 both did a very good job of this.

The current Dolby Surround decoder does a good job too - but from my (probably unreliable) auditory memory, PLII and L7 did better.
 

theREALdotnet

Major Contributor
Joined
Mar 11, 2022
Messages
1,202
Likes
2,077
How much do you watch Movies and Tv Shows ? The Center channel is the single most important speaker in a home theater setup (maybe tied with a Sub or two). From multiple articles written about home cinema audio mixes. Approximately 60% of the sound for a movie or tv show is directed to the Center channel speaker. Don’t skimp here or you will regret it.

After seeing this advice plastered all over the Internet for years, I actually went out and added a centre speaker to my 4.1 TV setup. I didn’t really have the space for a centre speaker, so I changed the way the TV was mounted and made room. I got a Paradigm centre that cost twice as much as the mains (Wharfedale). In the end, I’m sad to say that it was all a waste of money and effort. The centre didn’t improve anything audibly. Most of the time I can’t even tell whether it’s playing or not (I have to put my ear to it to be sure). For stereo material (non 5.1 movies or most of Youtube) I can get the AVR to upmix, but all this does is make everything a bit louder.

My conclusion is that a centre speaker only improves things when the mains are incapable of producing a good, stable phantom centre.
 

Chrispy

Master Contributor
Forum Donor
Joined
Feb 7, 2020
Messages
7,938
Likes
6,097
Location
PNW
Stereo isn't necessarily limited to 2ch.
 

Sancus

Major Contributor
Forum Donor
Joined
Nov 30, 2018
Messages
2,926
Likes
7,643
Location
Canada
Can upmixing to 3.x fix those cancellations?
That’s main reason why I wanted to get a center channel.
Yes, DSU and DTS upmixing both do a good job of extracting phantom center. The result is significantly clearer center vocals. Auro3D does some extraction but less. However they all do other things as well so whether or not that's better is up to you.

I do think an upmixing mode that only spreads the front soundstage over 3 speakers(no surround) would be interesting to people who are really picky about alteration of stereo sound, but unfortunately there isn't such a thing in normal AVRs. I'm not very picky because I think all stereo sounds quite bad so I think both Auro and DSU are big improvements to most tracks.

For Tv/movies I think DTS is best as it does a really good job of moving sound effects to the right place, much more aggressively than the others. For music this aggressive remixing produces weirdness more often than the others.
How many stereo recordings rely on psychoacoustic tricks that upmixing will uncover?
Phantom center is quite literally a psychoacoustic trick that upmixing can easily recognize. So, lots of them.
 
Last edited:
Top Bottom