This is a little out there, and not really a request for a feature to be actually implemented, more like something to think about.
What if a track was recorded with a certain Kemper profile, and you want to try another sound but for some reason you don't have the DI guitar tracks for re-amping? Theoretically, since the amped sound is the result of complex calculations performed on the original guitar input, it should be possible to reverse these calculations and reconstruct the original guitar signal from the amped recording -- if you know what rig was used. Right?
Unless the relationship between input and output is not a one-to-one mapping. And I suppose the discrete nature of digital sound data (in other words, the limitations of sample size and bit depth) would introduce inaccuracies in the two-way trip from guitar input to processed sound and back to reconstructed guitar input...
Just thinking out loud here really.
Adding distortion to a signal reduces its dynamics, i.e., constrains it to a smaller dynamic range. Since digital sound data are discrete in nature, i.e., a finite number of samples having a finite number of possible states between -∞ and 0 dB, constraining a signal to a smaller dynamic range should lead to a loss of information. For instance, if a single sample has (for simplicity) only 10 possible states over its full dynamic range, and if adding compression/distortion to a signal and amplifying it would substantially constrain it to, say, the loudest half of its range (meaning that the waveform spends more time in the high values between zero crossings), this would reduce the number of possible states for most of the waveform to 5. Meaning that information is lost, and that the original signal could never be reconstructed accurately since a one-to-one mapping between a system with 10 possible states (unprocessed sound) and a system with 5 possible states (compressed/distorted sound) is impossible. In other words: a loudness of 7 in the processed signal could correspond to a loudness of either 3 or 4 in the original signal... This is all hugely simplified, of course, but it suggests that my "de-amping" idea is not feasible, or at least could never be perfectly accurate. Still, I think it's an interesting theoretical aspect of using a digital amp.
