guitarists: integrating neural amp modeling python code?

protozoa's icon

Neural amp modeler (https://www.neuralampmodeler.com/) is a python package that uses audio files to accurately model amplifiers and speaker cabinets. There is a VST that works in Max and a standalone application. Both use profiles and impulse responses that users can upload here: https://www.tone3000.com/

The code is open source and I'm wondering if integrating it (not just the VST) in Max would be doable or advantageous for cpu utilization.

Thoughts?

Steve Rummage's icon

Proto, did you find anything on this?

protozoa's icon

A native Max external is doable, though I still don’t know if it would confer any advantage in processing efficiency. I think I lack the skill to find out definitively for myself. There are DSP routines in the Neuro amp modeler core that can be read by Max to process audio and the impulse response convolution would need to be done separately I think.