AI Replaces Thousand Dollar Guitar Tube Amps
Written by Harry Fairhead   
Sunday, 11 October 2020

The faithful still love their old fashioned technology in the form of analog tube amps specifically, but not always, for guitars. The only problem is that they can cost thousands of dollars. Why not let AI lose on the problem? That's what SmartGuitarAmp does and it's free and open source.

Is it worth it? It all depends on your point of view. I battled with tube amps back in the day. They were subject to hum, unless you had the extra cash for DC heaters, they acted as microphones if you dared to knock the amp, and they were very easy to drive into distortion -  without care distortion was the basic operating mode. Transistors made things so much better - a lot less noise, no hum and crystal-clear audio. The problem is that a lot of people just don't like the sound. Too clinical. Too lacking in warmth and quality. Oh hum....

If you want a tube amplifier you can get one, but as they are specialist engineering put together with a touch of brand name marketing, a lot of mumbo jumbo and a high price tag. In this digital age, it seems backward to construct tube amps even for an audience that is ready and willing to pay. This is where AI enters the picture. Take some recordings of guitar music being played via a tube amp and train the network to reproduce the effect. This is what Keith Bloemer has done. Building on the work of Eero-Pekka Damskägg (WaveNetVA) he has created SmartGuitarAmp, which has a GitHub repo and an Apache 2.0 licence.

It has been implemented as a VST3 plug-in and it even looks the part:

Only problem is that it wouldn't keep you warm on a cold day.

All of this highlights a problem in the audiophile world. Before digital you could spend a lot of money on excellent speakers, a solid turntable, a good amp and you would play vinyl at the best possible quality. When CD's came along you could still get reasonable quality and you could pay more money for DACs that promised to make the sound better, but, and this is the problem with digital, the sound quality is always limited by the digitization parameters. Streaming has made things much worse. I now have access to more recordings than ever, but all in very low quality. If you move up to high quality digital streaming then the range of choice is much lower and many of the things I want to listen to are jsimply unavailable.

With the storage and bandwidth we have available today, it is time for mp3 and other lossy formats to die. Could AI help with this problem?smartamp

More Information

Related Articles

Learning to Sound like a Fender


To be informed about new articles on I Programmer, sign up for our weekly newsletter, subscribe to the RSS feed and follow us on Twitter, Facebook or Linkedin.



Natives In Tech Accuse Apache Of Cultural Appropriation

The Apache Software Foundation is being urged to change its name and the names of the projects it hosts by members of Natives In Tech, a collective of Native technologists crafting free and open  [ ... ]

Future Learn Unlimited Offer

Programmers love learning. In particular they have embraced online learning with many developers revealing in surveys that they enrol in multiple courses - sometimes concurrently! So here's news of an [ ... ]

More News





or email your comment to:


Last Updated ( Sunday, 11 October 2020 )