I think the original round of songs was generated with Suno. What does your pipeline look like? Did you try the new models in the space that have emerged since, or is Suno still the best?
I’m not a composer by any means, but, every so often, I come up with a tune that I’d like to larger melodies. The companies developing music models seem to be focused on the `type in or generate lyrics → specify style in natural language → song is completely AI generated` pipeline.
I have a musician friend who is also frustrated with the existing AI tools. He wants to be able to write some lyrics, sing a tune, and have the AI generate all the accompanying instruments with that tune. But none of the tools can do that.
I think the original round of songs was generated with Suno. What does your pipeline look like? Did you try the new models in the space that have emerged since, or is Suno still the best?
I’m not a composer by any means, but, every so often, I come up with a tune that I’d like to larger melodies. The companies developing music models seem to be focused on the `type in or generate lyrics → specify style in natural language → song is completely AI generated` pipeline.
I have a musician friend who is also frustrated with the existing AI tools. He wants to be able to write some lyrics, sing a tune, and have the AI generate all the accompanying instruments with that tune. But none of the tools can do that.
Suno is surprisingly good at that! It has only gotten good enough very recently, but I had some decent success the other day.