Vital audio created an AI to create presets from prompts and I tried it and only one preset from like 30 sounded good right away, but maybe had some interesting sound that are crude but can be tweaked and refined. However I think just follow YouTube tutorials and importing Serum wavetables is a better process and yields better results.
I abandoned it because I changed, DJing and making music became everything to me and I lost all interest in all the nerdy Linux/BSD stuff. I haven't even used BSD since 2022. I do love mixxx in Linux and that is my workhorse for DJing, I may do some production on Linux with Renoise but for now music production is on Windows.
@charlie_root@3T@dcc@kirby I have started making explainer short videos on Youtube I think its been around 2 months... I am improving and learning along the way... This by the way is made from 100% FOSS software and public domain data... the voice is generated from Coqui-TTS meanwhile it is animated using Blender... The topics are chosen from AI.. Basically I tell ChatGPT to generate 100 titles for a certain topic then check each title in the keyword searching tool of vidIQ... Only those which score high 60+ are chosen... This narrows it down from 100 to like 8 or 9.... then I tell ChatGPT to generate the script based on certain template... and then I animate it using QGIS for creating maps and then the geotiff and shape files are imported into Blender using FOSS BlenderGIS plugin and then I animate it using Blender.. Its not the best thing out there but I am constantly improving.
The description, tags and hashtags these are generated from the AI tool of vidIQ.
@charlie_root@3T@dcc@kirby you were good at it.. i liked your style.... i don't know why you abandoned it but i can get it.. youtube is like a full time job and its very difficult to churn video after video unless there is some type of automation or you have like more than 1 person at least an editor.