PG Music Home Forums General Discussion Off-Topic Suno - AI Music Generator

Suddenly I feel so..... what's the word?????

Useless?

Yeah, it appears to be the "next big thing" in AI music. And unlike Suno, it appears that Udio is actually licensing their material.

However, AI generated material still can't be copyrighted.

But while I find AI generated music interesting, I'm not really interested in using AI generated music because it does all the parts of the music that I consider to be what makes it my own.

Although not within Band-in-a-Box, in this video Bob Doyle Media shows how Suno can be used as a great starting point, especially when you pair it with Band-in-a-Box!


-Callie

Suno, or any other Ai, and BiaB are just tools and you can choose to use them or not.

It is easy to become fearful and irritated by new ways of doing things. I think the older we get the more change becomes uncomfortable.

There is good reason to fear some techonology, atomic weapons come to mind. AI has some issues. Can AI kill off all the humans? Perhaps. Is that a big deal or any great loss? Likely not. We are most likely much less important than we would like to believe.

We will be able to play and create music fora long time into the future. We will also find it harder to monatize and collect on those efforts as time goes by. AI will play and create music and at some point will be most likely be indistinguishable from human made music.

Many of us have concerns about ethical issues but I fear we are in the minority.

Good, bad, right, wrong and ethical issues tend to disapear as distance from shore increases...lol

Riders On The Storm


Cheers

Billy

Revisiting this topic. After doing a load of research and speaking with our AI guys at work I have a better understanding. Suno and others like it are not regurgitating others music, or sampling it or copying it or any of that. It was trained, supposedly on copyrighted music and based on all that training when given a prompt it actually creates a unique output that meets what it knows about that genre and any other qualifiers you may have used. In some cases it is quite spectacular. It still has hallucinations, gives you nonsense lyrics when you have checked the instrumental toggle for instance. But over all very impressive. I like it to many of us who learned how to play our instruments by learning from the people we studied or looked up to etc. I learned to play guitar copying Chet Atkins records, note for note. Now I can create something completely mine but I did it by learning what he did.
So in summary Suno and others are not just combining pieces of other music and calling it new. It generates from scratch a song, using programming of course, that mimics what it has learned that song should sound like, what instruments it should have, how that should sound, and what they should play given a chord structure it created as well. You ask for a blues instrumental in slide guitar and it pretty much nails it.

In some cases, it is "regurgitating other's music" - or at least portions of other people's copyright material.

For example, it's been shown to generate producer's tags - basically audio signatures placed at the start of tracks.

Those "qualifiers" are how it categorizes elements of the songs that it's learned. Using the right qualifiers, it's possible to make Suno create output that is quite similar to copyright materials.

Given that Suno is able to generate copyright-infringing output, it's clear that no "supposedly" is needed. Sunu is being trained on copyrighted material for which they have no license to use.

AI companies attempt to stop this by preventing users from using certain descriptors, but with some clever prompt engineering people have been able to get around that restriction.

Yes and no.

It's not "combining" in the sense that it takes the first few seconds from one song, and then a couple more seconds from another song.

But it is combining in the sense that it's learning songs, and then creating output by morphing elements of those learned songs.

It's certainly more sophisticated that cutting and pasting between songs. But without having learned those songs - and stored them in its neural network - it would be unable to create new songs.

I'm not sure what "using programming" means, but it doesn't really generate a song "from scratch". It generates a song by morphing together learned elements of songs from the training set. It just happens to have a huge amount of songs to draw from, so it's often difficult to see where it's come from.

But choose enough descriptors that match the training data, and it'll spit out elements of those songs, showing that it's just copying after all.

My wife, daughter and I were listening to a choral festival this afternoon. My kid was in the concert hall (one of her works was on the program) while we were streaming it at home in California.

A piece came on and it was clear that the “composer” had asked an AI bot for something in the style of Sibelius. The words made no sense at all but the music was way, way too close to Finlandia for comfort. As we were discussing this at home, the signature chord change occurred and my daughter texted, “Turn left at Helsinki”. That text was easily the highlight of the piece.

Sibelius is still under copyright in many countries for another three years. Finlandia is in the PD in the US, however, so there’s nothing that anyone would do on that score. The blatant sound-alike was cringeworthy per my daughter’s post concert report and my wife and I thought the same thing. A new arrangement using the tune with decent lyrics would have been preferable—what we heard was embarrassing.

Whatever about the legalities involved, we can let the lawyers sort that one out, Ai music making is sure as hell entertaining and will only get better.

You can probably tell I am a big fan!

Billy,
well said!

This all reminds me of the Sony's pitch of their Aibo robo doggie in 90s. "Just like a real dog..."
I prefer music (for the most part) that is made by humans, not a parody - even a very realistic one. A taste thing.
Curiosity - sure it's a fun technology. A tool - likely useful, if you are not substituting a skill that you already have.

If you have a $100-200 printer and can print 100 variations of Mona Lisa a day with crisp detail and vivid colors, this doesn't make it an "art", or you - an artist. Same as watching sports - doesn't make you an athlete. A new generation of Promtomusic (c).

P.S. What is great, that centuries of music is available to purchase or rent either digital or CDs, LPs, etc. No shortage of good entertainment for my lifetime

I lost the thrust of my point, I think we all create music, using what we learned from other music, in a similar way that our AI engineers here at my work describe to me what is going on. It just has the capacity to ingest 1000's of those songs and use it to generate its output. Do I have to attribute Chuck Berry when I use a double stop lick in my playing? Nope, but everyone that hears it and knows Chuck Berry can pretty much know it was lifted from him.
I know in my experimenting, I can't get it to create anything that sounds like something else I already know except in style and genre etc which it nails almost every time. does it use chord progressions that are prominent in those styles? Sure it does just like every songwriter on the planet. ii V I anybody?

It will certainly be decided by the lawyers and the courts. The big "record" companies are mad because they think they should be the only ones exploiting artists. how dare the computer guys get in on their game. If anybody is going to rip off an artist they want it to be them and nobody else!! LOL.

Lots of good points being made here by some level headed people though.

In the meantime, some of the AI stuff I have played with has given me some really good ideas!

In this explosive fire storm of technical innovation we are all witnessing (aka AI development), my hope is that BiaB doesn't become obsolete.

More than ever, companies large and small must adapt, incorporate, pivot and invent or face decline.

Yes, indeed. The sky is falling — again.

When you create music, you don't draw from the audio signatures of millions of songs to generate not only replicate chord progressions, but instruments, vocals, and studio effects.

Not "1000's", but millions of songs.

Songwriters remember songs.

AI doesn't remember anything - it stores the information in a neural network and uses that to create songs that are based on that information.

AI is designed as a tool that learns the elements of millions of song - including the voices and the instruments - and then creates new songs by mixing elements of those songs. Every song created is a remix of a prior song.

That's entirely unlike how people learn and replicate music. As a small example, AI isn't going to include ideas inspired by a T.S. Eliot poem, the death of a loved one, or a seeing a rose in a garden.

Every word that AI writes is taken from lyrics from someone else's song, using voices copied from other singers.

But AI doesn't know whether something it replicates is copyright infringing or not, because it's all potentially copyright infringing. Everything it does is created by a process of copying. The difference is that the process of replicating the sounds can usually hide the source material because there's so much of it.

That's not because it can't do that, but because the designers have intentionally hidden those controls from you. But AI can very much create songs with Elvis' voice, or riffs created by Chuck Berry.

Because it mixes together inputs, AI songs tend to be an average of the training data.

So it will tend to make generic choices, except when it drops signature sounds, like guitar licks and vocals.

Equating the process that AI uses with the process songwriters uses is a false equivalence.

AI doesn't have its own voice, the ability to play guitar, or the ability to consider the emotional impact of its lyrics.

Unfortunately, this is likely to be 100% accurate.

It seems that companies that own the copyright aren't interested in preventing AI from competing with artists, but are more interested in making sure they get a piece of that pie.

I have been using Suno for a couple of weeks now and I’m very impressed. I’m am trying to create a bunch of blues songs for a “concept” album about my journey with Parkinson’s.
I’ve created all of the backing tracks in my usual workflow. Start in BIAB, then to Studio to improve parts that I am capable with guitar and limited keyboard skillls. But I hate singing voice.
I became aware of different Ai apps that can either improve your voice or replace with one I might like better. In the process I stumbled upon Suno.
I put in a prompt for “a Chicago Blues Shuffle with Horns 120 BPM” I entered my own lyrics.
What I got was astonishing! I got a great blues tune with appropriate breaks and song structure. It created a chorus portion that I never would’ve imagined and a great ending. There is virtually no support for the product. So I watched a lot of YouTube videos and learned more about writing prompts and part markers I.e., verse chorus intro outro pre-chorus etc.
I created a song that, after 3 regens that was what I was looking for and much better than I would’ve done on my own. I had a blues song with an excellent, very expressive male singer and a great song structure with lots of dynamics and breaks that I never would’ve thought of. In my headphones it sounded perfect. The guitar solos were really excellent (although the tone was a little too distorted) the horns were enough in the background to keep their artificial sound obvious. The rhythm guitar was spot on as were the drums and the bass. The cymbals had a hiss that I hope I can resolve.
So I split all the tracks on Moises Ai (then only stem splitter I’m aware of that can split rhythm and lead guitar). The quality of the individual stems were less than professional particularly the drums.
I don’t have BIAB 2025 because I’m on Mac, but it looks as though some of the new features might allow me to recreate parts that need recreating.
Ai is still in the infant stage. Maybe the stage where the infant starts crawling. But in a few years it will be very good.
It will still require a skilled person to create the music and eventually the public will differentiate between canned music and music that required talent and skills.
People who resist it will be left behind.
I hope Peter and the rest of PG are staying on top of Ai (I think they are) because it can be the death of BIAB or it can enhance BIAB and become part of the music creation workflow of the future.

It's only plastic if you use it and are inexperienced with music structure. Just inputing a few prompts and pressing a button is certainly going to produce something of lesser quality at least to some of our ears.But for others it can be a means of expressing something locked away in their minds that they otherwise couldn't release. I think we've lost touch about what creativity and expression are. We, also, often confuse creativity and expression with performance. Both sometimes require talent, especially the performance part. However a performer doesn't necessarily mean they are creative... they have maybe never written a song/score in their lives.

Expression on the other hand isn't about talent, it's about getting out what we are feeling and using whatever tools we have available to us to accomplish that. As a musician I understand this fundamental aspect.

I'm a musician (guitarist)/songwriter/producer and have played around with SUNO. It can be an impressive tool for musicians/songwriters if there were finer controls. However, it continues to evolve and every couple of weeks or so, they are adding tighter controls for users. What I like to do is record some riffs/melody, on guitar, upload it to SUNO and have it re-interpret my audio sample, then begin fleshing out a score around it. I then save as a WAV file and use OpenVINO (in Audacity) to seperate the stems. I can then bring the drum stem into Reaper and use Groove Agent to rebuild the drums (manually), then the bass lines and add any other instruments I want. Then I can record any guitars myself. I see AI, in the future, simply becoming like a word processor (as it is to writers), but for songwriters. Take a look at Ace Studio, it is impressive for AI vocals, giving very minute control to the producer and it is trained ethicly. Once those fine tuned controls are in place, songwriters (in the near future) will laugh at those of us who worked 2yrs on a song, never to get it the way we envisioned it, but they can produce the ideas locked away in their heads with absolute precision in a few hours or days!

I would suggest playing around with some of the other tools in SUNO... (Extend, Replace Section and Crop). They will give you finer (albeit limited) control over the final piece. These tools didn't work all too well back in December and January but are working much better now.

Also, if you are familiar with Audacity, you can add the Intel OpenVINO AI plugin to it and after importing your WAV file from SUNO, you can split the stems. It does pretty well on drums, bass and vocals. It will not seperate rythym guitar/lead etc though. If you use OpenVINO I recommend using '8 Shifts (the max)', found under the Advanced Options to seperate the stems. It takes longer but seems to produce better results on drums, bass, vocals.

Honestly, it's all subjective. AI music creation is really only a little more than a year old. It's a baby by any standard. Once finer controls are put in place, there won't be too many artists, in the near future, NOT using it in their creative processes. I don't know of any writers still using a typewriter (and some don't write at all, instead, hiring someone to write their ideas/story for them). Ethical training of AI, however, is certainly a concern, but another matter altogether.

Also, expressing oneself, doesn't require 'skill'. Performing on an instrument does, of course, but that doesn't make that person someone who has 'creative', or even 'expressive', skills. Songwriting requires some skill, but that doesn't necessarily mean that that person can perform on an instrument very well. Creative processes are about getting expressive thoughts and ideas out, in tangible form, what one has bottled up inside them. If the tool that allows one to do that just so happens to be AI, then all the more power to them.

Final thoughts. Used as a tool, skillfully, it can be a great aid to musicians/songwriters/producers. For those who have expressive ideas locked up inside their heads, but no skills on an instrument (or knowledge of music structure) or access to those who do, it can be a fun and creative way to let it out. However, I see it as garbage in/garbage out (but again, that's my opinion and is completely subjective), but I have heard some extremely creative stuff generated by a few people that has impressed me as a traditional musician/songwriter/producer, with over 40yrs experience, and I have no doubt that there was a lot of thought and effort behind the input/lyrics, etc.

LOL, we old dogs just have to adapt! Been learning to use various AI tools in my creative process since December and have been having a lot of fun, but it requires me to learn a lot of new toolsets, which aren't so easy to learn... and that is NOT so fun!

If AI in music takes after AI in Teaching, our goose is cooked. I can create just about anything I want as a teacher in AI in seconds. It just strips out everything mechanical or mundane. All I have to do is read it and check it. Then iteratively improve it. What's more there are dozens of programs to choose from. I reckon teachers are starting to use AI far more than students. I suspect music will go the same way. As DJJimBob commented, I suspect it's how you use it to make it yours that will redefine our creativity. We're on the cusp of a brave new (possibly dehumanized) paradigm in many areas of life. Anyway tried the programs to make a website using AI? Quite an eye opener as well. A few funnies, like Ms Chan and Mr Wong with pictures of Caucasians of the wrong gender. Just create fake people in a graphic AI program to replace them. A brave new world.