Perceptual Effects of Room Reflections

[srrxr71]

All research points to an ideally linear response of such a mono speaker when evaluated with typical stereo recordings. And yet the sound is different when you listen to the same in phantom stereo setup. So then is Toole wrong?

How about with mono recordings?

Did he compare stereo condensed to mono recordings with one front speaker with and without that compensation?

Do we know if the preference changed?

Linear is an ideal goal. It’s hard to argue a speaker should be non linear and then the question comes in which way it should be.

He needed a way to test speakers so he used the Occam’s razor concept. He tested mono. He determined what sort of dispersion matters. He may not have bothered to fuss about the stuff we talk about. He wasn’t particularly studying tonality as much as he was about dispersion and trying to find other relevant factors.



Maybe someone can take that baton and fuss about it. Maybe we will learn something.

I’ve done casual experimentation with a summed mono signal from both speakers and from one speaker and turning my head etc.

I did not find much. I was very causal about it though. In fact in my setup I try to get the center image to sound as close to a mono speaker in the middle as I can. It’s a helpful reference. But beyond that it’s an opinion about tonality.

Why not input the curve in pEq and see if it sounds better?

 

[Thomas_A]

How about with mono recordings?

Did he compare stereo condensed to mono recordings with one front speaker with and without that compensation?

Do we know if the preference changed?

Linear is an ideal goal. It’s hard to argue a speaker should be non linear and then the question comes in which way it should be.

He needed a way to test speakers so he used the Occam’s razor concept. He tested mono. He determined what sort of dispersion matters. He may not have bothered to fuss about the stuff we talk about.

Maybe someone can take that baton and fuss about it. Maybe we will learn something.

No Toole admits it will sound different between mono vs stereo. He use multichannel instead. A center speaker solves it. You are thus left with the errors of stereo. Errors that will increase with less reflections and in near field.

 

[srrxr71]

No Toole admits it will sound different between mono vs stereo. He use multichannel instead. A center speaker solves it. You are thus left with the errors of stereo. Errors that will increase with less reflections and in near field.

Fair enough. It’s different. You can pEQ that out.

Tell me which system will derive the phantom center channel from the millions of stereo recordings in the world?

So back in the 50s they knew about this and proposed from mono we go to 3 channels. But the market wouldn’t have it.

What can we do?

If you know a system that derive that center channel from regular stereo then I will try it.

 

[Thomas_A]

Fair enough. It’s different. You can pEQ that out.

Tell me which system will derive the phantom center channel from the millions of stereo recordings in the world?

So back in the 50s they knew about this and proposed from mono we go to 3 channels. But the market wouldn’t have it.

What can we do?

If you know a system that derive that center channel from regular stereo then I will try it.

You cannot fix it, but only reduce errors. Meaning reflections to fill in the 1-2 kHz dip and a slight reduction of power 2-4 kHz.

 

[tuga]

Maybe you can sort of fix it:

 

[Thomas_A]

Maybe you can sort of fix it:

Yes, I would however not be long-lived if my wife catch me with that in our living room.

 

[sarumbear]

If you know a system that derive that center channel from regular stereo then I will try it.

I had one in 1970

Dynaquad - Wikipedia

en.m.wikipedia.org

 

[srrxr71]

I still say the mixer already fixed it. Unless the mixer was noseless.

 

[srrxr71]

You cannot fix it, but only reduce errors. Meaning reflections to fill in the 1-2 kHz dip and a slight reduction of power 2-4 kHz.

My early reflections are down more than 10dB.

I tried placing an acoustic panel in front of my face and really I could not tell much difference if any. I know people will say it’s not enough.

I’m still confused about what this dip applies to. It seems to apply to the small % of crosstalk only. What am I seeing wrong?

 

[srrxr71]

I had one in 1970

Dynaquad - Wikipedia

en.m.wikipedia.org

Yes and you speak in past tense. What happened?


In all seriousness how was it?

 

[Thomas_A]

I still say the mixer already fixed it. Unless the mixer was noseless.

How does the mixer fix cross-talk cancelation?

 

[sarumbear]

Yes and you speak in past tense. What happened?

Quadrophonic came and all sort of surround failed.

In all seriousness how was it?

Better than many up-mixers today.

 

[srrxr71]

How does the mixer fix cross-talk cancelation?

Because he too is suffering from the same effect while mixing and he compensates for it.

 

[srrxr71]

Quadrophonic came and all sort of surround failed.


Better than many up-mixers today.

So i’m very confused. Do you get a proper stereo image with speakers set up around the room like that?

 

[Thomas_A]

Because he too is suffering from the same effect while mixing and he compensates for it.

The cross-talk cancellation cannot be compensated for, especially if you are at a mixing console with a high direct:indirect sound ratio. Alll systematic EQs aimed to reduce cross-talk dips and peaks at the source would lead to a non-flat ideal on-axis frequency response curve for a speaker evaluated in mono. Which isn't the case, if you believe Toole's research.

 

[srrxr71]

The cross-talk cancellation cannot be compensated for, especially if you are at a mixing console with a high direct:indirect sound ratio. Alll systematic EQs aimed to reduce cross-talk dips and peaks at the source would lead to a non-flat ideal on-axis frequency response curve for a speaker evaluated in mono. Which isn't the case, if you believe Toole's research.

Ok but why compensate for it? Wouldn’t the mixer have mixed to what he hears?

Also this cancellation affects only the portion of the sound that from right speaker hits the left ear right?

Also what i’m saying is that he evaluated in mono because it takes out so many of these factors and said flat is good. Okay so? I think we all kind of knew that intuitively. It’s the other stuff he found about directivity that his research has contributed to the field.


Edit: okay now I see what you’re saying. Even if the mixer on some subconscious level tries to fix it he will likely make it worse. I just take the mix as the product of what the guy heard and leave it there.

For live space recordings it’s an issue.

For this I’d have to learn a lot more about microphone techniques and how they work to discuss it. I’d be way out of my depth if I tried.



ADDITIONAL THOUGHTS: I looked at that diagram again.

So figure C is an anechoic room. None of us have that.

The last figure is the ITU-R BS1116 complaint room.

It has taken me 40 panels to get my setup to show compliance per my Genelec GRADE report. In fact that is my target. For me it’s very simple. I just do what gets me the best GRADE reports and then play around and tweak to taste. I just got my first set of absorptive/diffusive panels today. I’ll get a chance to play with them see what they do.

 

[sarumbear]

So i’m very confused. Do you get a proper stereo image with speakers set up around the room like that?

That diagram as the speaker placement was not intended. I think the wiki authors wanted to simplify the layout due to the name. It was LCR at the front and single surround at the back. Similar to the original Dolby Surround, which was modified Dynaquad.

 

[srrxr71]

That diagram as the speaker placement was not intended. I think the wiki authors wanted to simplify the layout due to the name. It was LCR at the front and single surround at the back. Similar to the original Dolby Surround, which was modified Dynaquad.

Yeah you can do that with just the wires. I used to do it as a kid with wires out of a small amp and play with l+r and l-r. It was fun.

I bet it better than some of todays extractors. Because it’s simple and goes back to basics.

 

[sarumbear]

Yeah you can do that with just the wires.

It was done with wires!

I bet it better than some of todays extractors. Because it’s simple and goes back to basics.

True. Until DVD-Audio and SVCD brought discreet channels, all surround systems were based on the same principal.

 

[Thomas_A]

Ok but why compensate for it? Wouldn’t the mixer have mixed to what he hears?

Also this cancellation affects only the portion of the sound that from right speaker hits the left ear right?

Also what i’m saying is that he evaluated in mono because it takes out so many of these factors and said flat is good. Okay so? I think we all kind of knew that intuitively. It’s the other stuff he found about directivity that his research has contributed to the field.

I think this started as a question whether reflections are bad or good. They are both bad and good, discussed many times at ASR, good to smear cancelation errors of stereo which is flawed format from start, and bad if they are too early and too high in level (clarity and/or stereo imaging suffers).

The point of mono evaluating speakers is that it is simpler and it is easier to hear resonances etc. And yes, the mixer does what he can. But it still remains - a center mono source sounds different compared a center phantom image from stereo setup. Adding reflections help a bit to reduce the errors, as shown by Shirley et al.