I know
What I'm kinda hoping for someone (maybe the OP) to test, is a completely different thing, though.
I guess I can see where you're coming from. If the Kemper and the QC were traditional samplers - which they most certainly aren't - then looping audio samples back and forth, over and over again between two samplers, would inevitable result in generational degradation of a type that the Rate Reducer partially can simulate. Even that isn't the same thing, though, but yeah - I get what you're saying.
The process I propose has very little to do with trying to manipulate sampling frequencies, bit depths or aliasing errors per se. Sure, there may be some of that creeping in, too, as I already mentioned in my post, but the main point lies elsewhere. I'm curious about how the fundamental (?) differences between the algorithms and analysis of the two systems - as well as the inevitable imperfections or randomness of the respective processes - when "crashed" into each other repeatedly, would accumulate and warp the original amp tone and characteristics generationally, over time, and perhaps lead to useful results beyond what an amp normally can do?
Remember that the raw material here are amps - not sound sources or audio samples. We would not be resampling audio over and over again. It's more about trying to find a new way to influence a multitude of (hidden) parameters by which any type of input audio - for example that of an electric guitar - gets shaped when fed through a device that after some serious mangling maybe can be pushed intp producing a behaviour that still resembles and feels a bit like a regular guitar amp, but now also has, perhaps, completely new characteristics not found in any real world amp. It would not be a basic, static effect that is slapped on top of the amp sound. Instead it could, perhaps, be another way of changing the amp behaviour itself - by using these devices in a particular, but technically "wrong" way. It's exploration - that''s all. It's probably complete nonsense, I know, but still...
As I already mentioned in my previous post, the probability that it'll sound like s** is very high, of course. Kemper even warns us that profiling digital modelling amps sometimes can lead to unexpected artefacts. I fully expect artefacts from this process too, but as the QC seems to be a little bit like the Kemper - I'm thinking more complexity, dynamics and user induced process randomness here - , I started to wonder if maybe those artefacts sometimes could lead to something musical or musically useful, too? Of course, the starting point would likely influence the end result to an extreme degree in an iterative process like this - ie. if the original amp sound is distorted or not, for example, would matter a lot.
I have no idea what sourcery is afoot inside the Kemper when it profiles. I do, however, have reason to believe that the makers of QC probably make use of deep neural network technologies for black box modelling of audio circuits in their device. Force-feeding real sourcery into a neural network training model, over and over again, sounds like a bit of weekend fun to me. I'm already married, of course...
In conclusion, this is not a very serious proposal. I get that this is not what these devices are designed for, nor what they're bought for. I'm simply a bit curious, and I decided to put this out there - in a tongue in cheek spirit - in case someone with both devices would have the time and energy to maybe try it out?
Okay - I'm done hijacking this thread now. My apologies to the OP. My excuse is that you kind of already started this process, in a way Also, you're the first person I've encountered that owns both boxes, and isn't a YouTuber Thank you for your contribution!
Cheers,
Bag