Hi everyone,
Just released my new single, “Cosmic Energy”! It’s a dreamy, atmospheric deep house track perfect for late-night drives.
Watch on YouTube: https://youtu.be/BI8TExTGlJE?si=1_WyomfOO8OBOlRV
Stream on Spotify: Spotify
Hope you enjoy the vibe. Any and all feedback is welcome!
Cheers,
Alex
Very nice melodies
Pretty girl and nice cosmic energy song.
Well done!
Sounds very nice.
Just one question: singer is AI?
Nice song and recording. I was also curious about if the singer might be AI given no name mentioned (and also the sound, though that could just be processing techniques for the genre).
The video notes also mentioned you conceived it and created in collaboration with Google’s AI. Definitely wouldn’t mind learning more about what that involves.
I don’t usually comment on music which is posted here but I have to say this is impressive. The mix is excellent with that sense of space instrumentally. I like you have chosen to have the singer up front with minimal reverb as a contrast. It’s really quite hypnotic. I would love to know how the vid was constructed. Looks very Unreal Engine-ish in a good way but knowing how long it takes to put UE stuff together I’m curious to know how it’s been produced.
Hey everyone, thank you so much for all the kind words and for taking the time to listen! I’m really glad you’re enjoying “Cosmic Energy”.
I saw a couple of great questions about the production, so I wanted to share some details on how it was made:
Regarding the video: You’re right, it was made using generative AI! I used Google’s Flow to generate the key scenes. Then, I took all those clips and did the final editing, color grading, and storytelling in DaVinci Resolve. It was a fascinating process!
As for the vocals: That’s a great question! The vocal was created with Synthesizer V by Dreamtonics. It’s a vocal synthesizer, so while its sound engine is powered by AI to be realistic, the entire musical performance is human. I manually programmed the melody, timing, and fine-tuned all the expression parameters note by note. It’s more like playing a virtual instrument than using an “AI singer”.
The rest of the track, including all the synths, drums, and the final mix, was produced in Cubase 14 Pro.
Thanks again for the great questions and for your support. It really means a lot!
Cheers,
Alex (AxelB)
Thanks for the info. I’d never even heard of Flow before. I use DaVinci Resolve as my video editor these days, but, thus far, I’ve really only been using it to edit live performance videos and create lyric videos with a constant background (typically something I’ve created in Photoshop, with or without generative AI). Flow may be something to experiment with in the future, especially for trying to create more typical music videos.
Vocal synthesizers have come a LONG way since my early experiments with Vocaloid. (I reviewed MIRIAM for CakewalkNet back in 2004 and also tried LOLA a bit – I actually did use LOLA for some background vocals on one of my recordings, albeit to layer with my own BGVs.)
Hey Axeleon, thanks for explaining. Like rickpaul I’d never heard of Flow or the Dreamtonics stuff. I’ve been using live vocals (a very convenient wife except with age she is finding singing accurately more difficult) and Sonarworks AI Voice which is good to a point. I shall investigate.
The video especially….when I think of the hours we’ve put in to our videos for an okay-ish result and what you have achieved with Flow I think it’s pretty mind-blowing and potentially a great resource for creatives.
I’ve added the track to my reference tracks (I’m quite into the deep chill genre currently.)