Greysound - A pro recording studio that engineers itself

Hello Product Hunt!

My story

My name is Anshu and I'm the creator of Greysound. I've been playing guitar since I was 11, singing (not counting pre-lessons shower vocals) and playing piano since I was 17, drums since 21, and faking it on bass guitar since 19 (guitarists know what I mean). When I was in college, I was in a punk rock band and all I wanted to do was sing and play all instruments (Dave Grohl style) on a studio EP that I wrote. I couldn't afford studio time, but at 17, I had endless time on my hands (between CS classes) to teach myself music production and audio engineering. I spent about a year getting good at editing, mixing and mastering audio on Pro Tools (an industry-standard DAW by Avid) and maybe another few years before I felt like my ear and production tastes were radio-ready.

Others with the same dream as me, but with adult responsibilities, don't have the luxury of spending years on a side quest before they can record their music.

That's why I built Greysound: a DAW with an AI CoProducer. Now, humans that want to make music don't have to also learn how to engineer audio, and even pro audio engineers can automate tedious parts of their craft to focus on the artistic parts of sound design.

My mission: Build the anti-Suno

Some seem to believe that making music is a chore that needs to be done away with. When AGI hits and takes away all our jobs, will we all just sit on a beach doing nothing? Will we just fry our dopamine circuits with AI slop while our hands and minds atrophy? There are some things that are chores for our species and AI should absolutely automate them. There are other things that humans enjoy doing, and AI should lower the barrier of entry for all humans to be able to do those things. Music, and art in general, is one of them.

Ethical AI

Our core AI (CoProducer and music creation in Programming Mode) is not trained on any human-made music or IP. It's trained on music theory, composition, production techniques, DSP algorithms and when to use them and other concepts that a human producer would be trained on. Sampling Mode uses AI generated samples provided by a third party provider who trained their models on paid-for, licensed, and ethically sourced music.

What's Greysound built on top of?

I built the web version of Greysound (this launch) from scratch leveraging the web audio API (that most modern browsers have in-built support for) for the core audio engine.

Our desktop-native version that supports third party plugins and AI-controlled hardware automations (coming out soon; an alpha build is out on our discord!) is written in C++ on top of an open source library called JUCE.

All AI features are built in-house except for Sampling Mode models and off-the-shelf LLMs that call our in-house models.

Ask

I'd love your feedback to make Greysound the best DAW on the planet, AI or not. We have a lovely Discord community of ~150 people that have been incredibly supportive and have been battle-testing Greysound. Please join us if you'd like to share bug reports and feature requests with me, or just follow along :)

I've been playing with this every day for WEEKS now, and I've been having a blast trying out wild ideas and stretching the creative limits of what Greysound can do. Anshu has been fantastic, practically superhuman with how dedicated he is to the project, and is constantly working with the community to make it better. This is a fantastic application.