Thanks, this is cool, but I find it a bit challenging to get the sound I'm looking for if I can't provide artist names to sample. Not that I want to use that artist in the track, but how do I describe wanting a track that mixes the jazz sounds of Frank Sinatra with the ambient beats of Thievery Corporation.
Is this a copyright issue? Or just that the system doesn't know what it was trained on?
Thank you for thinking of this and implementing it!
But, I'm wondering is there any major difference between just clicking the "..." next to the song in the default interface and selecting "Download->Audio"?
You need to create an account to access download links (and I only see this option for songs I've generated). However, it looks like each link simply points to an mp3 file based on the song id. eg:
1. Open the website https://app.suno.ai/.
2. Click on the share button above any song to obtain the share URL.
3. Paste the URL onto our website https://sunoaidownload.com/.
4. Click 'Download'.
God, I remember a month or two ago when Google released its first music synthesizer that was making distorted songs with barely recognizable lyrics. yet it was a song nonetheless. I was so excited and told myself this thing would probably be much better in a year or two.
A month or two later and Suno is all that I was imagining and even more. The AI tech is advancing like crazy. This right here is an undisputable proof of the value and usage of genAI that nobody can deny. Suno wrote and performed in 3 minutes a song better in almost every categories than those "created" by the likes of Travis Scott or Justin Bieber, at least in my opinion.
Even if the tech is frozen tomorrow and nothing but minor improvements is made, this alone can already shake up the entertainment industry.
My mum teared up to a personalised ChatGPT poem when I was demo'ing it to her. She would love to listen to some blues/country tailored lyrics of her own devise.
I have found I have developed a strong distaste for most things AI-generated. When I try to articulate it, even if the end result looks superficially well made, I think it is because it is missing some human connection, some authenticity. It has a shallowness and repetitiveness similar to scripted reality vs actual documentaries. Similarly I can acknowledge people drawing photo realistic paintings takes skill, but most of the time I find them dull and uninspired.
I had just created a meme for a throwaway laugh using copilot, I hope it doesn't make me a hypocrite.
29 comments
[ 4.3 ms ] story [ 73.2 ms ] threadIs this a copyright issue? Or just that the system doesn't know what it was trained on?
But, I'm wondering is there any major difference between just clicking the "..." next to the song in the default interface and selecting "Download->Audio"?
https://cdn1.suno.ai/$SONG_ID.mp3
I imagine that Cafe's etc would love this.. they could tweak the music to suit their clientele and not have to pay any royalties.
A month or two later and Suno is all that I was imagining and even more. The AI tech is advancing like crazy. This right here is an undisputable proof of the value and usage of genAI that nobody can deny. Suno wrote and performed in 3 minutes a song better in almost every categories than those "created" by the likes of Travis Scott or Justin Bieber, at least in my opinion.
Even if the tech is frozen tomorrow and nothing but minor improvements is made, this alone can already shake up the entertainment industry.
I had just created a meme for a throwaway laugh using copilot, I hope it doesn't make me a hypocrite.