

Performance with AI music
Here's one way to present your ai music and video to a live audience. Play it through a monitor stuck on your head lol This is performance art meets ai generative content. You can't deny the human ingenuity that's gone into thinking this through. Apologies for the poor quality film but you get the point.
How will the watermark work?
Sorry if this has already been asked. Does anyone know exactly how an audio watermark might work? Will it be an audio signal outside of normal human hearing or will it be some kind of embedded metadata? Will it be at the start of a track or randomly placed in a track? For those using a daw, will it be identifiable in individual stems or only present in one stem? A piece of music is very different to an image (where watermarks are widely used and understood). I'm interested in how this concept will work. Is the technology already in place or does it need developing.
AI Software Instruments
This is an interesting development by Magenta. A prompted instrument generator as plugin software. What's your views on this? https://magenta.tensorflow.org/mrt2#apps-plugins
Does £4 for a coffee = game over!?
Coffee lovers? Does £4 or more for a miniscule flat white or other mean game over for your coffee habit? Or are you going to pay regardless? I'm interested to know as it could be the beginning of the end for the coffee shop boom.
Sit down concerts or lively bar?
What would you rather play? A sit down concert where everyone is quiet and listening to your music. Or a rowdy bar with a listening audience but free to chat, drink and wander about or dance. My preference is the second option probably because it suits the band and is what I'm more used to. A concert is a completely different vibe. But I'm interested in other musicians views on this?
Taking a break from Suno
I'm taking a break from Suno and have unsubscribed. It was becoming way too addictive. And I came up with lots of unfinished projects whilst continuing to get new ideas for new music projects. That was frustrating for me as I like to finish stuff. My workflow is using stems in a DAW so once these are initiated the process slows down. I take a lot of care with mixing and producing so it does take time. It's all good though and I've enjoyed using Suno and very likely will renew my subscription when the time is right. Has anyone else been through this?
Anyone else getting a few bum notes in their outputs? Notes that aren't in key with piece and completely out of tune? I can remove/replace them by downloading stems to a DAW but I'd rather the notes were right in the first place. Any other work rounds?
I mostly write my own lyrics but sometimes AI lyrics can also be quite funny. Like I don't worry too much about rhyming words because it limits what you want to say so much. But AI lyrics seem rhyme obsessed. One of the early AI tracks I prompted was a joke thing asking a friend if he fancied going out for a few drinks. So the phrase I put in was:
"Oli do you fancy a beer some time soon?"
I was messing about and in a hurry so I let AI add the next line. Which was:
"It's been too long, we're orbiting the same moon."
It's so bad it's good and I laugh every time I hear it. I'd never have tried to rhyme soon with moon.
Has any one else got examples of terrible lyrics that are also randomly funny or stupid!?
(Names have been changed to protect the innocent)
Just for fun I made an AI song about AI slop. Oh the irony! But my reasoning is to advocate careful and considerate use of AI to create quality content, rather than churning out song/music spam. The lyrics are all my fault.
[Electro Punk] Stop the Slop! by Songwriter Mart