@Dietz said: 2. Algorithmic reverbs are _of course_ trying to recreate real rooms, too (... ever heard the "Vienna Hall" from a System 6000?), and they _do_ need a lot of processing power - why is it just today that we see the real Lexicon algorithms for native machines? They relied on dedicated DSPs, that's what made it seem so "easy" on the machine.
Just for the record you surely know much more about reverb than I do. It's true that lots of names of patches on Algorithmic reverbs come from real spaces, usually (I've heard) because of the early reflection character of that patch (and to give us an idea). But of course on a Bricasti, Lexi and VSS4 (and others) you can vary the the amount of the ER's, vary the length of the tail, control the modulation, etc. coming up with sounds that no room will give you. I think of them as supra-natural. And I have a feeling that the "lush" sound that many talk about, is a little supra-natural (something no room could give you).
We're all come accustom to certain sounds: plate reverb vocals in the 70, certain verbs (EMT) on snare drums, etc. And human ears have for a few hundred years been hearing orchestras in great halls. So that "natural" hall sound is what we deem as correct. I can't help but think that after decades of getting used to "bigger than life" orchestral sounds that come from (for example) a real orchestra in a great room with a little Bricasti added in, we might start considering that sound as the standard.
For example, the last Batman score (The Black Knight) was recorded in an L.A. scoring stage, and then up to 8 Bricasti M7's were added on in the mix. That's an impressive sound, but not one that works for a more classical situation. We are all used to different sounding pianos, snares, guitars, etc, and yes even need orchestras that can be molded to sound different for different situations. So we all need (IMO) dry samples, MIR, and maybe also a Bricasti for more larger than life sounds.
As for the power, I was surprised to learn just how little Lexicon and TC verbs actually take. On a 8 core MacPro someone got I think over 50 instances of a stereo Lexi verb, and TC system 6000's use 10 year old dsp chips (like ants rowing inside to generate the power!) yet the verbs sound great (!?!). Bricasti on the other built very thirsty machines which are I think about 10 times more powerful that a System 6000. And I remember reading here about MIR, and all the processing power needed to achieve that quality. [Y]