Smart Black-Box Neural Networks Recreate Classic Guitar Amp Sounds in Real-Time

Researchers at Aalto University and Neural DSP Technologies claim to have created a neural network capable of emulating any guitar amplifier with enough accuracy to be indistinguishable from the real deal in blind listening tests.

"Deep neural networks for guitar distortion modeling has been tested before," explains Professor Vesa Välimäki of the work, "but this is the first time where blind-test listeners couldn't tell the difference between a recording and a fake distorted guitar sound! This is akin to when the computer first learned to play chess."

"The tests were conducted to validate the performance of models emulating either the Blackstar HT5 Metal or Mesa Boogie Express 5:50+ tube amplifiers," adds doctoral student Alec Wright. "The models were created with a focus on real-time performance, and all of them can be run in real-time on a desktop computer."

It's an impressive achievement: Recreating the effect of an analog amplifier in the digital realm isn't the most straightforward of tasks, but one which is key to preserving the feel of vintage hardware that is rapidly reaching the end of its usable lifespan.

Previous best efforts in virtual analog modellng have relied upon traditional circuit modeling techniques, a labor-intensive process that must be repeated for each target amplifier and which produces a model too computationally complex to be applied in real time. Instead, Wright and colleagues relied on black-box modeling: applying an input signal and observing the output, then creating a model to replicate it without requiring an understanding of what is going on in the middle.

More on the project is available in the team's paper, published in a special issue of the journal Applied Sciences under open-access terms.