The Advantages of Floor Coupled Up-Firing Speakers

[markus76]

I agree, no wavefiled synthesis is needed but speakers and the room and the decoder in our heads as a system

This is something that I tend to believe too but in the 80 years of audio reproduction nobody was yet able to describe the necessary framework in an objective way.

 

[el`Ol]

What do you mean by "turned 90°"? This is not how multichannel audio works. Upmix algorithms had and have all sorts of problems and limitations, that's why multichannel audio like we know it today was invented (see Toole for more details). The problem is that the industry that would benefit the most from multichannel audio didn't adopt it (yet): the audio industry. The technical platform is there but instead of using it, guys like Key "invent" a "new" upmix algorithm. This will not change things to the better.

With turning 90° I mean that I have one of my ears turned to the middle between the two speakers and the other looking away from it, so that the speakers of a stereo setup become front and rear of the same side. With multichannel I meant 5.1. And what I wanted to say is that mixing between front and rear of the same side has to be very different from mixing between the two front cannels.

 

[markus76]

That's has something to do with how phantom source localization works. Furthermore encoding 4 channels into two results in a loss of information and bad channel separation. Wait for Toole's book to arrive. It should then become much more clearer.

 

[el`Ol]

Just one question:
What is the current standard for panning between front and rear of the same side?

 

[markus76]

We can't localize phantom sources on the sides very well so it should not be tried. Only additional loudspeakers would help. Left and right surround channels should be used for envelopment. Something 2 channel stereo can't provide.

 

[el`Ol]

Listen to Andrea von Salis recordings. Speakers stand in 4 x 90° and ideally front and rear speakers are the same type and driven by the same stereo amp. My percussion CD meanwhile is sold for €69 used at Amazon, but there are demo downloads at their website.
And if Key had a mixing algorithm that really works and would licence it to manufacturers of mixers, that would also be fine.

I heard that in cinemas the center is used for dialogues only. So is panning usually only done between left and right?

 

[markus76]

Listen to Andrea von Salis recordings. Speakers stand in 4 x 90° and ideally front and rear speakers are the same type and driven by the same stereo amp. My percussion CD meanwhile is sold for €69 used at Amazon, but there are demo downloads at their website.
And if Key had a mixing algorithm that really works and would licence it to manufacturers of mixers, that would also be fine.

Doesn't sound like a realistic scenario. Multichannel is here, why try to push a new standard? The music industry is further nullifying itself with that kind of behavior.

I heard that in cinemas the center is used for dialogues only. So is panning usually only done between left and right?

There is no "usual". You would have to differentiate between cinema sound and music. Cinema is somewhat standardized (e.g. dialogs mostly from the center channel to get rid of the "stereo seat"), music is just pure chaos.

 

[Paul W]

The music industry is further nullifying itself with that kind of behavior.... music is just pure chaos.

So sad, but so true. Optimize a system for one recording only to have a completely different mix in the next recording.

 

[markus76]

Someone that actually agrees with what I say? That's new

 

[el`Ol]

Doesn't sound like a realistic scenario. Multichannel is here, why try to push a new standard? The music industry is further nullifying itself with that kind of behavior.

Of course a new standard like the AVS quadrophonia will never happen. Maybe it would be the standard today if it had been invented in 1960. But it is an experience I didn't want to miss.

What concerns new mixing methods: They are possible within the existing standard.

 

[markus76]

We don't need any new mixing methods. We just need to have people start using what's already there.

 

[el`Ol]

You just said under the common mixing standard it is better to use the rear speakers as ambient speakers only. Would be a waste of money if that stayed that way.

 

[markus76]

That's not what I said.

 

[el`Ol]

What is the current standard for panning between front and rear of the same side?

...it should not be tried. ... Left and right surround channels should be used for envelopment.

 

[markus76]

Guess you misunderstood what I wrote. We were talking about lateral summing localization, e.g. one speaker to the front left and one to the back left. This is something that does not work well with any system that relies on summing localization. If you want good lateral phantom image localization, you would need to introduce additional speakers. Phantom images between back left and back right surround speakers work similar good as between front left and right.
Does that make it any clearer?

 

[el`Ol]

In the AVS recording I tested the front and rear phantom localization is very wide and the ambience closes the gap smoothly. Although it will never be a standard it shows one can get very far without dozens of extra speakers.

Graaf, if you want to test: You will need two amps and the balance between front and rear is critical. And use the Kefs in front and the flooders in the rear.

 

[markus76]

Don't understand what you're trying to say. Multichannel that is derived from a downmixed signal will always be inferior to discrete multichannel signals.

 

[el`Ol]

I wanted to say that it is too early to give up.
"downmixing" at it is understood today can be very different from a more sophisticated way of panning.

 

[markus76]

I wanted to say that it is too early to give up.
"downmixing" at it is understood today can be very different from a more sophisticated way of panning.

Wait for the Toole to arrive. There's much more to it than what we've discussed and what some people proposed.

 

[Key]

A lot of questions I'd like to answer but just don't have the time. Been really busy lately with real world.

I will say that markus is missing something here. My standard that I am about to explain will be adopted one way or the other because it's the objective truth. I know he and others probably think I am full of it and I really can't blame them. But I am going to prove it to a certain extent soon. It will have to be rigorously proven over time but really that isn't my job or interest. I want to just get back to making music honestly.

Okay that first "out of speaker" localization demo was really spur of the moment and not ideal. Here are 3 better examples imo. That first one had some "rice crispies" that I can't do anything about and probably annoy some people.

Lossless
Lateral Reflection Test
zSHARE - THE ROOTS - Y_All Know Who.flac
Invisible Speaker Test
zSHARE - The Pharcyde - Officer.flac
Just because test. (a little bit of an artifact threshold test lol)
zSHARE - Gnarls Barkley - Crazy.flac

Lossy
zSHARE - THE ROOTS - Y_All Know Whoa.ogg
zSHARE - The Pharcyde - Officer.ogg
zSHARE - Gnarls Barkley - Crazy.ogg