i'd be happy about a little explaination about the 10000 thing, what is it exactly 10000 samples?
the axe-fx can handle "only" 1024 sample IR files IIRC so is there just a differnce in lenght or
will there be some secret extra sauce going on?
cheers
S.
I guess length would not be useful, plus, it's not a technical problem. It's just a cab impulse is certainly (stop me if i'm wrong) not that long. It's only the image of the response of 1 dirac input, it's not likely to be much longer than that.
As Ermz say, I guessed from what I understood it could be more a "definition", a "resolution" increase. Since most ppl complain about the unrealistic fizz, high end, "scrthhschh" sound that comes with the digital domain, maybe increasing the frequency/bandwith of the impulse will make high end closer to reality because these are precisely the frequency on the spectrum that get close to the impulse freq/bandwith (I don't know how much time does 1024 samples represent). If there is a way for the impulse loader to read those 10000 samples as fast as our current 1024 samples impulses are loaded, the increase of precision in the high frequencies would be augmented, closer to a virtual "perfection" since the human ear couldn't, maybe, make a difference. This would be a little like the retina display on iphone 4 : a double of resolution that gets close to the human eye maximum ability to distinguish "pixels". I guess a "retina 2" would be the last step since it's already hard to distinguish iphone 4 pixels.
If then you could have layers of impulses chosen depending on the amplitude of the input, the dynamic response would be closer to reality as well since cabs are not likely to be linear. I remember some tried this on Nebula but failed cause it induced a frequency or phase problem in the end that wasn't solved. But that would be awesome.
The only last step would be implementing secondary effects that I don't know (resonance ?) but this would be only the last % of increase of quality and would need much more CPU power. I haven't done signal treatment theory long enough to know other stuff than all that.
Dunno if that's what DAR is based on but that would be awesome. Maybe they found something better.