Author Topic: AI will DESTROY Recording, the DAW, and Plugins  (Read 1644 times)

Offline Chaser

AI will DESTROY Recording, the DAW, and Plugins
« on: January 20, 2024, 10:06:00 AM »
I try and stay as current as possible in regards to AI and a "Heads Up" when possible.
I have posted AI Splitters etc etc and as it advances so have the Models and training.
The Tech is advancing rapidly.Microsoft,Google and Amazon are in a Race manufacturing AI Chips for Generative AI.
New Copyright laws etc etc or existing will need to be adjusted and maybe some form of embedding into the song and AI will be able to determine/search Copyright Violations throughout the Internet..Licensing/Rights etc is different using FREE and Paid Subscription See FAQs.
The Video gives examples of AI Models in Training including complete  Songwriting..Lyrics,Instruments Full Mix etc from just a Text prompt in under 2 mins.

The Youtuber (Benoni Studios) is someone I have known for quite a few years and was a heavy user of BFD before all of their problems.He was blown away by how advanced the Tech has become.This is a great example of how far/fast AI is advancing..Metal is one of the Harder Models for AI as generating the drums, distorted guitars,Vocals etc is a bit more difficult to accurately reproduce than mainstream music such as Hip Hop,Pop and a number of all the new genres today that implement lot of Autotune ,sampling etc.

The most advanced Model out of all in this Video is Suno (Formerly known as *Bark) and will give you an idea of how fast the future is approaching using AI in Recording or DAIW as Benoni calls them
Good..Bad..Hate it or Love it..just like years ago when DAW's were introduced and the move from hardware to software..Sampling wars in the late 80's etc etc if you are into DAW and recording , especially as a Pro you'll want to pay attention as your competition will no longer be required to have years of experience or High End Studio and gear and the time to start is at the infancy.
I see those with little to no experience writing hit songs in the next few years or so with nothing but an iPad/iPhone and a prompt..and more Virtual concerts with onstage Holographs..

The Video is about an hour..the middle of the Video demos other Models that generate from different sources such as Humming etc..if you want to skip those for now and see just Suno results , watch the first 20 mins..and 45 mins to end.
There are 2 examples generated LIVE and currently a verse and a chorus around 1 min in length..(there is a continuation feature that adds segments of 60 seconds based on your subscription plan see FAQs) great for a sketch pad/ideas or try generating drum less tracks of your own etc... as time/training goes on the Model and songwriting capabilities/options will advance quickly.

Since the release of this Video , I have noticed a few changes and at least one major change.
During the Video Benoni used Artist names in the text prompt.."In the style of"...Slipknot..Taylor Swift etc..That is no longer allowed...you will be prompted that Suno is for generating Original Music.



Suno AI is Subscription base  Maximum of 2,000 songs per month.

The Basic version allows 10 songs a day which if you know what you are doing and how to manipulate text... create an AI generated Album or AI generated Soundtrack everyday for FREE.

There are currently over 250,000 members on the Discord Server and average 30-40,000 online.
There are currently under a dozen genres and few Showcases/Videos as everyone is in the learning stage

At the heart of Suno AI are two major models: Bark, which specializes in vocals and lyrical performances, and Chirp, designed for instrumental backing tracks....They work in tandem.




AI will DESTROY Recording, the DAW, and Plugins

Suno AI

FAQs

*Bark
Bark is a transformer-based text-to-audio model created by Suno. Bark can generate highly realistic, multilingual speech as well as other audio - including music, background noise and simple sound effects. The model can also produce nonverbal communications like laughing, sighing and crying. To support the research community, we are providing access to pretrained model checkpoints, which are ready for inference and available for commercial use.

EDIT:
Added Images/Links
« Last Edit: January 21, 2024, 09:24:13 AM by Chaser »