Acoustic Sim parameters.
Is the flat mode a result of correction in the Kemper software?
I think it is rather : How to correct KemperCab frequency response to have XXX Speaker Response using special filter(Imprint) response . Something like Tone Match for measured Speakers.
Anyone remember the Slate thing where they said "we have profiling technology"? Turns out they didn't.
They did. They just can't sell it to the people(KPA patent).
And that's partly the reason - I don't like patents. They stop competition and progress.
MIDI clock is a part of a group of messages called "System Realtime" messages and it has no specific channel.
Set the KPA channel to e.g. 2(different from the channel of the kb) and thats it. (System->MIDI Settings -> Midi Global Channel)
If you have Omni then KPA recives every keyboard controllers MIDI messages.
The neural guy talked about this on the gear page, including some of what makes their approach different, I think.
Yea - I just searched for this paper .
And it makes sense. They had to analyze how KPA works to bypass the patent. And this document largely reflects the KPA structure - that is, two EQ blocks and one nonlinear block in between. Quiet sounds examine the sum of AmpEQ and CAB EQ. High signal level examines CAB EQ... and so on. It is worth reading into it if someone likes to know how something can be constructed and how it probably works.
Of course, we will never know 100%, but one can probably assume that this model could have been used.
I'm curious in an academic sense how the Quad Cortex will do profiling with an approach different enough to not violate the Kemper patents.
If it’s not only ancient EQ matching and if the results are great sounding and feeling this will be an interesting machine.
CK patent is here :https://worldwide.espacenet.co…pacenet.com&locale=en_EP#
but I think the better explanation lies in this paper: https://www.hsu-hh.de/ant/wp-c…r_amps_with_WH_models.pdf
Udo Solzer and Felix Eichas have some very interesting papers on approaches slightly closer (and potentially better) thank Kemper's.
But our approach relies on a much more fluid structure.
Kemper seems to follow a more rigid Weiner-Hammerstein topology, for which the non-linearity transfer function is static (hyperbolic tangent or something of the sorts) and the profiling process adjusting the FIRs and saturation block gain.
Our deep learning approach allows us to use more sophisticated loss functions (ours is based on human perception, which results in a much more "accurate" black-box model). Of course, you need a lot more computational power, which is another reason we had to make such a powerful unit: when using Neural Capture the neural network uses 4 SHARC cores in tandem for training).
This approach generalizes better (it can cope with a wider range of sounds), and can even learn parallel signal paths (such as multiband distortion or pedals with a clean blend, dual microphones).
Honestly, this algorithm is the most intriguing . Will they be able to detect and simulate distortions from pream and poweramp simultaneously? Will it be able to properly profile amplifiers with built-in gates or parallel distortion paths?
"keep the tail of the delay or reverb"
It looks like the device will also have the option for delay or reverb spill over.
I just wanted to add that if the truth is that they've used AI or machine learning to profile nonlinear distortions, this will be another big step in nonlinear audio simulation.
Even if it doesn't match the competition.
Unless it turns out to be just marketing and hype.
That is why it is worth waiting and seeing the results. In the end it counts.
For me the "biomimetic AI technology" or "neural network algorithm" souns very interesting from technical stand point.
I believe that you can use similar algorithms as in Ray Tracing to reproduce the image based on machine learning.
No to super Fajnie że reprezentacja PL użytkowników KPA się powiększa .
Just set SPDIF Output to Git/Stack and record it with XXX (not comercial)profile and post this here with XXX profile .
I think something might be terrible wrong.
That clip you posted does indeed sound awful.
This might actually be good as I don't think I can sound that bad even If I wanted, which most probably means that there's something that can be fixed.
I'm also using mostly TopJimi profiles (the AFD pack.)
And it does sound exactly like his videos (guitar differences apart).
It translates well from Gibsons to Strats or Teles (which I do own).
That clip just doesn't sound right.
My first thought was.. dead batteries on active pickups.
Clip attached from a random Diezel profile I found With a p90 Les Paul.
Way to much gain and a bit bass heavy and apologize for the random noodling.
Just to give you an idea...
Where are you based?
I believe if you could sit down with a fellow kemper user with it's own Guitar and Kemper you would be able troubleshoot whatever issue you have there pretty quickly.
I think you have some issue there for sure.
Also I believe a lot of people here should make an effort to be more polite.
If they can't don't reply.
I think he's coming here humbly asking for help.
I agree, that's why I asked to post :
1. Clean DI recording
2. Profile ( free profile )
3. KPA OUT recording.
@Terence T we can reamp it just to compare . This will tell you and us where the problem is. Please do that or this conversation dosn't make sense anymore.