As a full time musician…. This is truly unbelievable. You nailed it that anyone who says “bro, I know this is totally AI generated” is lying. Pretty mind-blowing stuff

In my experience in using both Suno AI and Udio, I found Suno both easier to use for longer songs and better overall in terms of music and vocal composition and continuity within each song (remembering how the verse/chorus/ sounded earlier) whereas Udio's overall sound quality is better. Udio sounds less compressed/aliased in general as compared to Suno. This is especially apparent with orchestral music which on Suno can sound like a low-bit-rate mp3 while Udio sounds nearly "CD quality."

First AI came for artists, and i said nothing because i don’t draw… then they came for photographers, and i said nothings because I don’t take pictures. Then they came for songwriters and there was no one left to speak for me.

I really appreciate your videos because you not only provide a "how to", but also discuss the extremely important philosophical considerations. Great stuff, man!

If suno was higher fidelity it would win hands down. If Udio solved the consistency issues it would win hands down.

Wow. I was feeling depressed because of my circumstances. Nothing was shifting my mood... then your video came up. Now I can't wait to play with this. Thank you.

Multitrack AI music will be the game changer for this industry and whoever does that first will win the game

I just came here from a 7 day trip, "creating" a complete metal-album. love it! some minor issues i came along: 1. the engine ignores some prompts completely, e.g. add a "storm" or "rain" effect to an intro with nothing else. so, the later you add the intro to the whole song the less will it be composed. 2. female / male vocals are complete randomly used or even ignored. 3. as it is mentioned in this video: chorusses are rarely the same. could give the song a more natural touch, though. (if the differences are not too extreme) 4. add [piano solo] and udio will give you a guitar solo instead...oO..or a vocalist 5. instrumental intro tend to have singing bots in a fantasy language^^ after 15 seconds. 6. The creation of the CD Coverartwork crashes if the song has too much text. a. i would like to have (besides the cutting tool) an option to shorten or stretch the single parts. b. and an option to modify the whole song structure after finishing it. like text, prompts, [x], and so on And futurepedia, thx for the video :) a question: is there a list which things i can put in brackets?

Don’t get me wrong, this is crazy tech. However as of right now there are a few things that I find AI music does consistently that can set it apart: 1 - it has very little idea of song structure, often putting emphasis in weird places and structuring lines in unconventional ways. (It can have a cool effect, though) 2 - AI tends to layer the vocals a lot more than normal songs (I assume to cover up slight inconsistencies in the voices) and the mixing is always a little instrumental-heavy (likely for the same reason) Again, this is still crazy tech, but as of right now those are a couple things that set human made music apart.

The human voice sounds rich when recorded/mixed well - but the key with AI is to not make voices sound 'tinny' when shifting registers and keys. BUT it's getting there. Love this video.

I am a song writer and this has really helped to make my demos to present my songs I write to the artist, if I choose to use an artist.

Amazing!! Wow and yes blown away by Suno and Udio and big thanks for Noisee, am creating some vids now 🎉

In UDIO it is much more difficult to make a good long composition than in SUNO, but if you try hard, you can achieve simply amazing results. Using my own lyrics, I made several tracks in different styles. And some of them, I dare say, are really good. Alternative rock in the spirit of Skunk Anansie: kzfaq.info/get/bejne/itWPjdOY07vFfXk.html Progressive rock slightly reminiscent of Porcupine Tree: kzfaq.info/get/bejne/edWVa61otp3Ro4U.html Progressive metal similar to a mixture of Mastodon and Haken: kzfaq.info/get/bejne/rdOmppV8mNPbm3k.html Alternative metal vaguely reminiscent of Tool and Deftones: kzfaq.info/get/bejne/f5d2o7ukqLa9hKs.html Progressive rock in the spirit of early Genesis: kzfaq.info/get/bejne/hNtznsx1xrvdm3k.html Alternative rock in the style of Alanis Morissette: kzfaq.info/get/bejne/lc6Khpmdl9-Xaas.html Eric progressive rock with female vocals: kzfaq.info/get/bejne/jq6VmbCEl-CrnYE.html

I’m blown away how well mixed and mastered these sound.

The part where they had Udio just doing comedy and sports commentary and such just blew my mind. The ASMR made me chuckle. But yeah, singing is so hard that when you can sing, you can do anything. Artificial human voices have been solved.

Happy birthday, Irvin!

Wow, a acompetitive alternative to Suno already?!? Scary for the whole music industry, from Berklee and Juilliard to Avid and Waves and of course all instrumentalists, media composers, songwriters and singers. One ring to rule them all.

The number one problem has been noted by a lot of us who write songs. It's a deal breaker for me. I don't care how realistic it sounds if I can't write a full song with consistent choruses. I still prefer Suno.

The idea of "musical brands" using AI to pump out songs, then hiring artists to perform them on tour is both genius and dystopian at the same time. I could totally see that model taking off though, as sketch as it sounds. Hopefully transparency about the AI involvement at least.

thank you, very informative

That's mind-blowing. Thank you for presenting those tools👍

As a creator of tracks with a jazzy vibe Udio works great!

Imagine an endlessly generated playlist, where you can press "like" so it learns and generates more songs like this in the future and if you press "dislike" it generates less of this in the future. Getting better and better, running on millions of humans to train the AI, ...wait are we the tool-user or the tool?

This is the best video about AI Music on KZfaq right now.

is it okay to use for your youtube? Who owns the final project and music?

Suno and Udio should join forces!

Not gonna lie, as a musician myself, the thought of AI cranking out radio-ready hits does make me a little nervous for the future of human artists. But you make a fair point that true artistry will always have a space. Maybe this leads to dope human-AI collaborations?

Thanks for the Info. But question is will the youtube channel be monetized if i generate song with Suno?

AI implementation in the plugin/Daw will come soon

13:04 "Extrarrestial" craft lading :D

I have been making punk industrial techno, heavy metal on suno and it takes me a few tries but I got great results (the singing not the best at times), But with udio so far I'm having a hard time making something that "new", or making longer songs, in suno I'm making at times some 9 minutes electronic songs that are great beginning to end. Hopefully Udio is gonna get there too.

I've already impressed clients with this - THANK YOU!

i couldnt't sleep because this was way too amazing to use haha xD

Too cool! I jumped right on it, and gave it a try. Wow!

It's CRAZYYYYY how much of a Nas this sounds like 3:48 :D

Also, I could not figure out how to start new section of a song in Udio at certain time like

What does the future hold if AI can be used to make music and videos ?

Impressive! I did two songs with my own words

15:40 "the nr 1 problem... chorus" SOLUTION: use=> [chorus repeat]

like your prompts they are good

Udio does sound the best, but it suffers continuity and indolence problems. It is also offering a restricted service at the moment (4./29/24) due to over-subscription.

hi. If we use the songs we produce here on KZfaq, will we have problems with copyright?

Omg I can finally have the asmr I want and not the horrible ones (or the same ones I have watched hundred times) yipi

A lot of thanx 👍

Hi do you know if i can use the songtext from Udio Users and make new Songs with it?cause in udio u cant Write to a Creative and ask

Great help, thx a lot !

Anything like STABLE DIFFUSION which runs locally for THIS MUSIC GENERATION?

Anyone can tell me how they did the non-music like sport analysis ?

Can we use the music in our content though? Or will we get copyright like most of the other ai music generators

Site appears to be down. Can't seem to login, keeps coming up with errors. Pretty amazing overall though.

How do you get udio to do non singing stuff?

I like udio, but Suno is better when it comes to doing full songs. I've been using it to re imagine songs I've written. Genius stuff

Right now I'm using suno but I'm willing to try other ai tools

I Like Udio's Interface but notice a lot of the songs sound like they are stealing someone else voice. One song I heard on Udio sounded Like David Bowie.

You're wrong, I can definitely tell it's AI, but not because of the lyrics, phrasing or instrumentation. There's a tonne of phasing issues with current AI generated music, as well as mixing (and lack of mastering). Having said that, this is early days and I don't think it will be long before these things are resolved. As a composer I'm horrified but intrigued. It's an absolute marvel, with horrifying prospects.

One thing I don't like about Udio is that it's hard to go from say a female singing a hook to a male rapping without it becoming the same female voice sing rapping. Suno handles this much better the majority of the time. The audio quality though is a huge factor in why I'm hooked on Udio right now though.

Thanks great vid ai is just amazing

Has anyone tried making reggaeton? How good was the result?

I'm waiting for traditional tracks to add, chop and splice from other songs that were generated, then let the AI do its magic from there. That would be "true" AI music generating. Can't believe that is not a thing yet.

I wonder if u can use the same singer across the Generations

Wow, wow, wow! for decades I have dreamt of this kind of software and now it's here. It's beyond my expectations.

AI music is not even boring anymore.

We are getting out the hood with this one $

Have your tried BFF AI? Has all the tools you want

Can you pay to have the copyright on these songs? Can you download stems and midi?

Help!! Ai jobloss is the only thing I worry about anymore.

It is difficult to get the right mood for classical music on Udio.

I was today years old when I found out The Wiggles genre is pub rock

tried 5 different prompts and crashed every time ???

KZfaqr Monzon Media shows you how to get consistent choruses.

Will A.I replace me and the rest of type beat producers?

I don't know. Even in your examples, I felt every Suno song sounded better than every Udio song. Suno follows lyric inputs pretty well whereas my (limited) experience with Udio has it almost completely ignoring my prompts. I think that once one of these companies makes a generator that will follow lyrics to the tee AND will allow input of musical notes, they will have won. I've made hundreds of gens of a couple songs I'm working on and neither service will generate what's in my mind so these are just toys really where we can roll and get something fun but can't use them as tools.

Humans can’t compete with this. New regulation is required. They’re feeding all of humanity’s expression into a blender to spit out made-to-order songs built on the backs of everyone, but without compensation. That’s what I see at least (and I know I’m not alone).

Poor Seth

09:37 - GEMIDÂO DO ZAP [LOUD MOANING GIRL MEME]

Chapters 0:00 Intro 0:41 Prompting overview 1:27 Blues examples and comparison 3:26 Rap examples and comparison 5:15 Mind-blowing example 7:00 Noisee for music videos 7:56 Rising Sun Gospel 8:55 Noisee reference image 9:28 Alien Ocean 10:06 Disco Funk 10:25 Pop Punk 10:52 Classical 11:08 Broadway 11:43 Classic Pop 12:27 Non-music examples 13:33 Remix 15:00 More tips 15:40 Limitations 16:43 Thoughts on the future of music 20:50 Futurepedia

I found udio to be very inconsistent when extending songs. Hard to create a longer song. Suno is way better at that.

Not working . Constantly giving errors.

UDIO IS NOT WORKING FOR MANY OF US Does someone from here have this issue ?

Suno makes better, catchier, _music_ , Udio makes better quality and _voices_ .

Going to just copy paste my dilemma. I felt a little nicer when I wrote this, so I'll keep my rage to copy and paste. Been playing professional Bass and Drums for 20 years now. This is the beginning of a very dehumanizing future, fun stuff now but long term is something people don't take into consideration Of course when new toys come out that is currently mind-blowing everyone is going to be so positive and defend it to the death. But this isn't my idea of something that will be a companion to help you expand your music. It is replacing your creativity and will make alot of people who already struggle with confidence they will get to that place they dream about as a musican, completely give up. Why spend a year of ups and downs as you are start from nothing but ideas in your mind and see them turn into complete songs. This will become the norm like everything else, our collective attention span is awful. Nothing slows down so anyone can catch their breath before something bigger and better arrives, its pretty unreal I don't see more people being realistic of the path this will most likey take for the majority of humans who lived a life already feeling left behind by society. Where does that lead us? There is beauty in creation over time to see your final product and look back on the hard work. Those are moments that makes life have meaning. Family and those accomplishments that lets you show how hard you've worked as you publish who you were at that moment in your life, captured through the language of music (something I beleive is such a beautiful way to say who you are and around the world its the language you don't need to learn too understand the meaning of why it was written). I really hope I'm wrong, but seeing people already pushing this so hard it doesn't matter about what I'm saying. Its becoming more impossible everyday to have anyone step back and realize this could push us completely out of the way in place of a prompt that takes less effort then making coffee in the morning. And its not just the music apps, its a frenzy to see if they can make AI for everything eventually that gives you the proud moments we hang on too. I know theirs amazing possibilities in this. But THINK. I'm more concerned about where we sit with these amazing creations. It's pretty F*cking low already. And it gives me a sense of sadness cause I know the feeling of not understanding where you are supposed to fit. Be excited, just sharing a different perspective on what its making me feel.

Only reason i still use suno is that udio does not remember its own chorus, so repeating chorus is impossible unless you do it 30 seconds aftet the first one, nonsense

It's perfect for video game tracks and advertisement jingles!

My prediction is at some point the big music studios will start buying up all these AI music generators so they can claim rights to anything anyone makes.

I hate Ai in terms of full production music and video. I also hate those who promote what will ultimately be the END of humanity. Regarding music generation it breaks every copyright law that exists. It cannot work without sampling all music that is released over the years.

My limited experience with udio seems that suno makes much better music while udio does superior voices.

First, you have zero rights to any outputs, under any circumstances. So you cannot use even 1 second of an output anywhere commercially. Non-commercially you have to give credits. This is understandable, it may change when they move into a paid plan, but... There is a big BUT there, I tried it with my own lyrics, non-english language, Turkish actually. The prompt was "alternative rock, male voice". It gave 2 outputs. I started to play and I was like WTF!? They were like; the first one is exactly the same vocal from called "mor ve ötesi" and almost exactly one of their song with my lyrics. Second was the same vocal of a singer named "Teoman" and almost one of his songs with my lyrics on top of it. The outputs were copy-cats, nothing creative. So that makes me question the capabilities of their ai model. Also while they have such legally binding and strict "terms of service", the way of they are violating artists' rights felt a bit hypocritical. I am not against using such training data on model training, all data should be usable on model training "in my opinion". But the model you created must be used as a piece of knowledge, not a source to copy and paste. If the model does this, that means you failed in model creation. PS: I didn't have that experience with Suno. All the outputs i got, felt unique (I barely used both of them so take the Suno part as a grain of salt)

Seems no one ever uses harder forms of "rock", why is that? Could it be that all A.I.s are bad at it. and what about "intros"? it ALWAYS places what sounds like vocals

I for one, welcome our new AI beat master overlords.

What I don’t like about you is that they want you to shout them out or - give them arbitration. So I think right now Suno is better because you can do whatever you want with your generated music and you don’t have to shout them out. Does anyone know of a good AI stem splitter that will also help us create notation? Because I do want to take these songs and make them more of my own and bright parts to it.

Suno is still way better.

Cool! Hope you guys like Noisee!

No you can tell it is AI, how? ITS SAYING TOO DDAMN MUCH , NO MUSICIAN WOULD SAY THAT MUCH SHIT , listen to Elton John, his lyrics are simple enough he wont get confused while he is playing. these ladies would get confused in their playing, or while singing they would lose track of their playing slightly.. I also noticed a few plays where they waited a little bit too long to end the syllable.. But the thing that bothers me, is there is only AI sound generators, there is nothing like a AI based waveform generator, one that you can dial in a prompt and get out a mod on an input wave, so that you can start deep faking sound or even create fucked up inbetweens.. Most AI sound generation is about creating finished songs. AI in general stops at just being art collection, you don’t really control AI, you collect art from AI and pick your best outputs, but where do you come into being relevant to the AI art generated? There is no collaborative expression, cause there is no realtime AI mirroring.. It will become me when I can sing with the AI and it can respond to what I’m saying or backup my singing or play a saxophone that matches the mood of my singing, or where I can play something and it manages to make lyrics to match my playing.. There is something like that for AI art generation, that is the image guide, or the combined input of an image guide and the prompt.. Now that SDXL Turbo is out, you can talk about AI image generation feedback, I haven’t seen someone do it yet cause their brain hasn’t gone there and offered it to the public.. I maybe could do it if I knew how to leverage the models in python to do that.. But AI sound generation needs its on parallel to the SDXL Turbo, where you can change the output in realtime.. But really, I think once people start making a lot of it, peoples brains will be able to identify the artifacts , and it will lose its effectiveness. The thing to do is not to try to replicate people. but to include them by making it easy for them to interact with the process. But AI is not something you can control like anything else, its like with coding, no programmer programs in assembly, cause the language for every processor is different, usually programmers code in a high level language like Python. A LLM like chatgpt is a step above that.. But each LLM is different and has different features.. Same is true of Image generators, the output you will get is different, you can’t control it as precisely as a lower level language like python or C, but when you change the models, you are not going to get the precision you get with coding it directly.. This will be the same with AI image generation or sound generation. And its cause the AI cannot know exactly what you mean unless somehow you can express things in its language and use its choice of words to get the precision. It will probably get easier over time, but I find myself using Gemini to generate my prompts for other AI gens. Cause when I ask Gemini for something , I know it has everything anyone has ever said echoing in its memory, and when I ask it to make something its pulling the precise words that can get to what I mean, and if not, I can pinpoint it more definitely with more interaction. Now I may use this program to see if I can create specific genres of music or if I can create something really bizarre.. I’m no interested in making pop music which is the examples you give, and not all musicians are going to want to make the same crap on the radio causethat crap is for people who are not into making music, they are more into wanting to shag the singer.. And this offers some comedy cause with this we could screw over a world of really stupid music lovers. And in that there is a lot of fun to be had..

Those who have successfully fooled themselves into believing that they are " creating " something by using this type of AI should be mocked and shamed , and it looks like it is starting to happen .

Bye Bye Ms American Pie....

Udio is great

Yeah let’s see it do something that doesn’t sound like it was done by a real person but better eg not a glorified cover. Get back to me when it writes a hit

Seems like everyone got paid a lot to promote Udio which is far inferior to Suno. Unless someone is "reviewing" no one prefers Udio

This will end human music. The listeners won't care

No lyric moderating. Enough for me to leave Sudo in the trash can where it belongs.