flipnaut wrote:the question is also: Nice to have core 6 and soon core 7 in nebula server... but which library-developer already uses the new feature for nebula 3?
Fortunately, no one. 3rd library developers are library developers they don't beta test.
But therefore 3rd party developers produce libraries of excellent quality and stability hardly any bugs or other malfunctions like most of the acqua plugins which are even beta tested, but still are not as functional as they are supposed be. So don't sit too high on the horse - it's sill a long way to go until AA acquas become something professional!
for the exception of compressors, we are not introducing new "variables"
The new engine simply does better things which are already in production: loading time, latency, multiband instances, less mistakes during deconvolution - ie level jumps, generic errors.
There is not a "new feature" there
We have tons of features on the gui side - the skinning language
On compressors/dynamic side there were a lot of improvements on the speedness and cpu load (how aria compressors work, they are amazingly light on cpu if you take in account the prog rate used)
There is a new big developement about envelope followers and nat detection and this is really new. Anyway it happened also in nebula3: we created compressors way before 3rd party developers could sample them.
Compressor templates were finally available 2 years later, in 2009, when we were confident about the tech - and everything started to be fully automated.
Yes, our compressors were a lot worse than what developers released, and they used the same tools. So we can expect big improvements even without creating anything new starting from now. It is possible we have already the tech for the best compressor possible in the software realm, already now. Time will tell you.
It should be stated that many current acqua plugins were sampled by experienced developers. Stedal is very good, for example.
On compressor side we had the first "clean automatic deconvolution" (without my bugfix) in ultramarine, and later on in prime comp mk II and ivory comp.
So we are close to a sort of automatic new procedure.
In the future it is possible developers will move it to state-of-art, exactly like today some of them are using old tags for getting complex things not planned to work that way.
Today they do not have variables linked to time and they are forced to hack the system in several ways. We have now attack time, release time and charge time sources, for example. Nat can decompose an attack or release shape to sub-shapes, or implement automatically a program dependant compressor in a very clean and simple way. You can sample a threshold. And so on.