alt Hacker NewsShow HN: LatentScore – Type a mood, get procedural/ambient music (open source)
Hey HN,
I've used Generative.fm for years and love it, but I always wanted to just describe what I was in the mood for instead of scrolling through presets. So I built this.
You type a text description of anything - from "mountain sunrise" to "neon city" - and it generates a procedural/ambient stream matching that mood. It runs locally, no account, no tracking, no ads.
Under the hood it's a custom synthesizer driven by sentence embeddings, not a generative AI model (although you can choose to use one!) — so there's no GPU, no API calls, and it starts playing almost instantly. The whole thing is open source: https://github.com/prabal-rje/latentscore
If you're a developer and want to use it programmatically it's also a Python library - pip install latentscore — one line to render audio. But honestly I just use the web player myself when I'm working.
Fair warning: it's still alpha and the synth has limits, so please don't expect full songs or vocals. It's ambient/procedural only. But for focus music or background atmosphere, I think it's pretty good.
Would love to know what vibes you try and whether they land!
- Prabal
Comments
I get clicks and pops every few seconds, using Librewolf.
But otherwise very cool!
I kinda liked how well you can fine-tune parameters of the music. Could be useful as dynamic soundtracks for games in low resource settings
Really elegant approach - mapping sentence embeddings to a deterministic synth feels more like building an instrument than generating content, and the instant playback makes it great for flow.
Would love to know if the same prompt always yields the same sound (reproducibility could be powerful), and whether you’ve considered semantic morphing between two moods over time.
The Fast demo model is already very impressive. It was way better than expected, but still required being a bit verbose since it didn't seem to understand rarer words ("sauna" didn't get me anything pleasant, "hot sauna" did).
The generated palette seem to be a great indicator of whether the model understood the prompt or not.
I Haven't checked out the Python SDK yet, but it seems very interesting!
I'm curious to know if there is any reason for why you picked Gemma 1B for the Expressive model. Did it generate more cohesive parameters than other 1B models? Or was it just the first one you picked?