It does have inspiration based generations! You need to decrease the context length quite a bit and next you need to adjust the prompt strength to create a whole new song using the same style.
@@Zerobytexai Yeah, but it always goes from the nearest point to xx seconds, so if you want to repeat a Chorus from 1 minute ago, it will be confused by most recent part, at first...
Fabulous video Bob, Thank you. I particularly enjoyed the snippets of you playing into the AI. It really made the whole presentation rich and meaningful to me on many levels. Best wishes
I properly burst out laughing when it abandoned English right at the end - as if it couldn't understand your accent. 🤣❤️ Time to experiment, and to see how it compares with Suno's new update. ❤
hey Bob, you make the best videos in this area and explain everything sensationally!!! Except for the videos you show at the top right, they are just really bad, not lip synced! It's enough if you're all alone there. keep it up 👍 you've given me an incredible amount of input. Thank you for your outstanding work. Greetings from Germany
That’s the best example of how to use this feature that I’ve seen so far. I had been trying some very old original songs just to see what it would do and didn’t like the result. But I was uploading the whole thing at once. Using snippets of new stuff seems to be the way to go. What I’d really like to do is be able to upload a quick full song sketch and let it recreate that with embellishments and vocals. But that doesn’t seem possible yet. Once that’s possible, that’s a fast way to get a fully copyrightable song fleshed out.
DOYLE RULES!!! Been testing the AI's ability to sing in two distinct voices, created a rap song about being Puerto Rican in New England but in Spanglish, Udio nailed it, then I tried a rap song about the world united and prompted multiple languages, again Udio didn't disappoint. Finally I prompted for a duet sung by a Spanish Female and Italian Male, I was able to get Suno to do it with custom setting. Udio is generating now.
@@marcosazambuja not a whole lot tbh, technically I don't own the rights pero copyright laws, however, I can use them as inspiration and write my own tracks, I'm a lyricist myself so I can drum up My own material and create similar sounding beats to create something totally original. But as far as the tracks I generated it's more for experimentation and fun. As far as monetizing them, that's a no go but they're cool to listen to and it's interesting to see the AI interpret my prompts
That’s incredible! The possibilities to customize your own output are now infinite. Finally. This is like a few years ago when AI Art programs started letting you input your own photos and drawings.
I want a rock voice and can't sing but made the music on my DAW what site do you recommend to drag and drop my vocals and turn them into some good thank you
I tryed this new feature and is very interesting. It can, for example, if you prepare a file, join two different music, filling the gap. As seen in your video, it can also build a song around an instrumental riff you gave it. The only bad point it's weak in trasforming human voice in a real song. A tryed to sing a musical idea with my voice, hoping to get some instrumental base around, but 99% of time it will simple extend your voice singing "a capella" with random variations. Very rarely it tries to add a base, and if it does is pretty random. Definitely it can't be used as "from singing your idea to get your song"...
Thank you for your comment. I had hoped that it would be possible to underlay or continue vocals with instruments. But apparently the development is not that far yet.
SUNO has just presented this same possibility of uploading an audio exclusively to certain TH-camrs, it is in Closed Beta, I would like to see it in Open Beta.
I’m intrigued to know how you got it to sing the correct words near the end of this video! I enjoyed watching and Learning from your video. I take it you recorded into a DAW via external microphone then saved it as mp3 then, uploaded that to Udio every time? Thanks for sharing
"The bayou is so wet, the trees are so low that I touch the roof, paddelling slowly. Throughout the swamps, where shadow is concealing. I hope that the years will bring bread but never chilli peppers. I'm dancing to the sound of the old accordion, the moon at my feet" Well that's Cajun AI ;-)
how many people will be uploading sections of songs be existing artists - extending them - then using inpainting to remove the origonal section? if they are not already running uploads throgh a recognician system like shazam already i think they will probably be forced to prety soon.
How about complete songs? Can I upload a complete song and let Udio handle the production? That would be a life changer fir Songwriters like me who wouldn't sleep at night knowing part of my catalogue was not generated in my own mind.
Great job on this video! UDIO's new Upload Feature is crazy good. I did a step by step tutorial video on my channel showing how you can take an 8 year old demo of an actual song written by Nashville songwriters and create a newer modern day version of the same song using Udio's upload feature. Technology is just getting better and better quicker than expected which is scary.
I wonder if Udio users are now unwittingly training the AI by uploading our music to the service to be extended feeding it more data to analyze and copy. For example I wonder if your voice that it cloned will remain stored in the system and end up being added to others generations. So far since I have been using Udio I have gotten perfect sounding cloned voices of Reba McEntire, FKA Twigs, Method Man, Nicki Minaj, Kali Uchis, and Cardi B.
Does anybody know where he's getting that video of the song? As far as I can tell, that's not something udio offers but if it is, I'd love to be able to use it.
hi thank's for your vidéos...is it possible to download one of my song and ask Udio to remix it in other style ?...( if not ..is there any ai which can make this ?) thank's
Very insightful: underscores in my mind that Suno is a better music generator than udio. The music just seems to be more dissonant and incoherent. Looking forward to Suno’s version of this feature when it launches
As cool as it might sound, I strongly recommend professionals to stay away from this new feature because you won’t be able to publish nor get copyright on the final product based on the terms and conditions of Udio……but for home users and to have fun is fantastic! Great video man!
@@kevinsturges6957 it is. Ask ASCAP or BMI. To get ownership you have to contribute with more than 70%…..35 years as a professional here…..ASCAP member since 1990, WGA and PGA member…..
Here’s a crazy idea why don’t we get the humans to create part of the song. Yea good idea! Here’s even a crazier idea why don’t we get the humans to create the whole song. That is crazy! 😮 Great video Bob as always, just not a fan of this use of AI!
IT will not permit more than 25mb sound file, the AI can only augment things it can see so 2:10 minutes.. It will only be able to do something with audio it recognizes, it may not even use your samples.. It can do better with copyrighted music than it can do with the non-copyrighted music, this is because the model was trained on copyrighted music.
The first two' ai generated parts sounded bad. I suspect you might do better by having it make an instrumental song first and make it add lyrics later if it can do that.
A lot of these Udio songs have monotonous melodies, the verses and choruses don’t contrast, the lyrics are meaningless. This isn’t music, this is an imitation of music.
Oh my god. I haven't had a laugh like that in a very long time. Best few minutes of the Internet this month! #asliceoflive #pizzablues #happyaccidents As far as the AI breakthroughs are concerned, however... Now things are getting VERY interesting! Already thinking of some creative experiments to try with this in my flows. And I'm looking forward to seeing how Suno claps back at this with v4 and their own upload model.
That’s a piece I wrote in 1990 or so. I think I have it on Soundclick somewhere with other stuff from this video actually…www.soundclick.com/share.cfm?id=4793106
You left out the part that Udio now owns the song. If you were to upload to Distro Kid for streaming and don't put in the credits giving Udio you break the TOS of them. If your song becomes popular making you money they could sue you and win. The only way around this is the song must be 50% or more different than Udio's output. I wish creators would identify this in their videos. They are not like Suno with their TOS.
Again Awful imho , udio. Has no consistency bad controls inly 30 secs... , i like suno better even if sometimes the sound quality us not As good.. Even...
I wish instead of extend/intro it had an “inspired by” option. Essentially remix something you upload.
won't late
Same. Udio was teasing some new updates and I hope that’s one of them.
It does have inspiration based generations! You need to decrease the context length quite a bit and next you need to adjust the prompt strength to create a whole new song using the same style.
@@Zerobytexai Yeah, but it always goes from the nearest point to xx seconds, so if you want to repeat a Chorus from 1 minute ago, it will be confused by most recent part, at first...
What happens when you put a full song that you’ve created into it? Do you guys have to worry about all the same issues?
Just randomly found this page...I love this guys real time reaction to Udio!!
Fabulous video Bob, Thank you. I particularly enjoyed the snippets of you playing into the AI. It really made the whole presentation rich and meaningful to me on many levels. Best wishes
So glad you enjoyed it!
I'm italian and confirm the italian voice is perfect...
Love hearing that!
I properly burst out laughing when it abandoned English right at the end - as if it couldn't understand your accent. 🤣❤️
Time to experiment, and to see how it compares with Suno's new update. ❤
hey Bob, you make the best videos in this area and explain everything sensationally!!! Except for the videos you show at the top right, they are just really bad, not lip synced! It's enough if you're all alone there. keep it up 👍 you've given me an incredible amount of input. Thank you for your outstanding work. Greetings from Germany
That’s the best example of how to use this feature that I’ve seen so far.
I had been trying some very old original songs just to see what it would do and didn’t like the result. But I was uploading the whole thing at once. Using snippets of new stuff seems to be the way to go.
What I’d really like to do is be able to upload a quick full song sketch and let it recreate that with embellishments and vocals. But that doesn’t seem possible yet. Once that’s possible, that’s a fast way to get a fully copyrightable song fleshed out.
Yeah, I woud like too. But probably this new function is not intended for this purpose...
I love Radio, as a guy doing it for decades, and i love your channel! Please keep making this stuff!!❤❤❤❤
DOYLE RULES!!! Been testing the AI's ability to sing in two distinct voices, created a rap song about being Puerto Rican in New England but in Spanglish, Udio nailed it, then I tried a rap song about the world united and prompted multiple languages, again Udio didn't disappoint. Finally I prompted for a duet sung by a Spanish Female and Italian Male, I was able to get Suno to do it with custom setting. Udio is generating now.
And what will you do with these tracks?
@@marcosazambuja not a whole lot tbh, technically I don't own the rights pero copyright laws, however, I can use them as inspiration and write my own tracks, I'm a lyricist myself so I can drum up My own material and create similar sounding beats to create something totally original. But as far as the tracks I generated it's more for experimentation and fun. As far as monetizing them, that's a no go but they're cool to listen to and it's interesting to see the AI interpret my prompts
First two ones are amazing too, I'd like to listen to the full versions
That’s incredible! The possibilities to customize your own output are now infinite. Finally.
This is like a few years ago when AI Art programs started letting you input your own photos and drawings.
'inspired by' the upload would be better and allow picking a segment we like from songs we don't own the rights of. Thanks ❤
That was cool Bob, I have it setup just needed a bit of direction and I got it. Thanks
I want a rock voice and can't sing but made the music on my DAW what site do you recommend to drag and drop my vocals and turn them into some good thank you
I absolutely love what it did with the klezmer music at 13:00 (the first sample) !!
I tryed this new feature and is very interesting. It can, for example, if you prepare a file, join two different music, filling the gap. As seen in your video, it can also build a song around an instrumental riff you gave it. The only bad point it's weak in trasforming human voice in a real song. A tryed to sing a musical idea with my voice, hoping to get some instrumental base around, but 99% of time it will simple extend your voice singing "a capella" with random variations. Very rarely it tries to add a base, and if it does is pretty random. Definitely it can't be used as "from singing your idea to get your song"...
Thank you for your comment. I had hoped that it would be possible to underlay or continue vocals with instruments. But apparently the development is not that far yet.
SUNO has just presented this same possibility of uploading an audio exclusively to certain TH-camrs, it is in Closed Beta, I would like to see it in Open Beta.
I’m intrigued to know how you got it to sing the correct words near the end of this video! I enjoyed watching and Learning from your video. I take it you recorded into a DAW via external microphone then saved it as mp3 then, uploaded that to Udio every time? Thanks for sharing
"The bayou is so wet, the trees are so low that I touch the roof, paddelling slowly. Throughout the swamps, where shadow is concealing. I hope that the years will bring bread but never chilli peppers. I'm dancing to the sound of the old accordion, the moon at my feet" Well that's Cajun AI ;-)
super Bob :) Curious what you think of the new audio input from Suno Ai - I had a big laugh again :) never have this kind of pizza in audio haha
how many people will be uploading sections of songs be existing artists - extending them - then using inpainting to remove the origonal section?
if they are not already running uploads throgh a recognician system like shazam already i think they will probably be forced to prety soon.
be nice if they implement exporting as stems
Thanks, been waiting for this one.
17:56 This music gives me hard Y’s VIII vibes.
What about the copyrights?
You do realize they're just training the "A.I." on your music. You're giving them free data.
satoru gojo on you profile picture, cool
You r so funny dude.. i like it 😂
If I have a complete song but it is only instrumental, can Udio add vocals to it?
Lovely! Would be cool to hear what it does with the bluegrass into synth pop 😅
Love the Sound of the ukulele ❤ whats the make ?! Great playing
Sonoauto was 1st one for this AI tech but i couldn't make it to work. But now udio will do it. 🎉
How about complete songs? Can I upload a complete song and let Udio handle the production? That would be a life changer fir Songwriters like me who wouldn't sleep at night knowing part of my catalogue was not generated in my own mind.
How does the music video that appears every time music plays in the top right play?
Great job on this video! UDIO's new Upload Feature is crazy good. I did a step by step tutorial video on my channel showing how you can take an 8 year old demo of an actual song written by Nashville songwriters and create a newer modern day version of the same song using Udio's upload feature. Technology is just getting better and better quicker than expected which is scary.
I wonder if Udio users are now unwittingly training the AI by uploading our music to the service to be extended feeding it more data to analyze and copy. For example I wonder if your voice that it cloned will remain stored in the system and end up being added to others generations. So far since I have been using Udio I have gotten perfect sounding cloned voices of Reba McEntire, FKA Twigs, Method Man, Nicki Minaj, Kali Uchis, and Cardi B.
I believe this to definitely be on the horizon, if it's not already happening.
someone please send me a link to the pony dreams song , it's lovely and I can't find it .
I really like your content, thanks a lot !
woooooow!
Does anybody know where he's getting that video of the song? As far as I can tell, that's not something udio offers but if it is, I'd love to be able to use it.
Which video are you talking about specifically?
Now they are training the AI with your music for free. I will pass on that
Still in my AI punk phase, but gotta check this out soon! (Is it available for everyone? Didn't see it on the site yet).
hi thank's for your vidéos...is it possible to download one of my song and ask Udio to remix it in other style ?...( if not ..is there any ai which can make this ?) thank's
10:40 and 15:26 are great
This my.fave tool udio wins plated all weekend
Pizza pleassse chessiest one please. New York style. 😂
Thanks
Very insightful: underscores in my mind that Suno is a better music generator than udio. The music just seems to be more dissonant and incoherent. Looking forward to Suno’s version of this feature when it launches
As cool as it might sound, I strongly recommend professionals to stay away from this new feature because you won’t be able to publish nor get copyright on the final product based on the terms and conditions of Udio……but for home users and to have fun is fantastic! Great video man!
As long as you write your own lyrics, I’m not sure if your comment is true?
@@kevinsturges6957 it is. Ask ASCAP or BMI. To get ownership you have to contribute with more than 70%…..35 years as a professional here…..ASCAP member since 1990, WGA and PGA member…..
What YOU actually create is yours.
Here’s a crazy idea why don’t we get the humans to create part of the song. Yea good idea! Here’s even a crazier idea why don’t we get the humans to create the whole song. That is crazy! 😮 Great video Bob as always, just not a fan of this use of AI!
I inserted the voice...but I don't hear any instruments, only the voice. How should I go about putting a backing track with AI? thanks from Italy
.. but you still don‘t have stems and it‘s mp3 quality…. Isn‘t it / haven‘t you?
IT will not permit more than 25mb sound file, the AI can only augment things it can see so 2:10 minutes.. It will only be able to do something with audio it recognizes, it may not even use your samples.. It can do better with copyrighted music than it can do with the non-copyrighted music, this is because the model was trained on copyrighted music.
Suno has this in Beta and I have to say I think Suno has overtaken Udio. Not sure how one goes about out early access to the Suno one.
That video was super cool.
Thanks so much!
You’re welcome.@@BobDoyleMedia
The first two' ai generated parts sounded bad.
I suspect you might do better by having it make an instrumental song first and make it add lyrics later if it can do that.
12:25 you can not tell me that udio used audio of jacob collier for their ai...
Please show how u get the videos
Everytime I click video I don’t know where it go es I can’t find it. I see u have videos pizza singers?
Can you add your own lyrics like Sora?
Absolutely
Use a little more genres like trance, house, etc.
😂 im gonna get out all my old guitar strumming recordings and see what it can do 🤙
...and can it do anything serious? With a goal in mind, rather than the cartoon stuff?
A lot of these Udio songs have monotonous melodies, the verses and choruses don’t contrast, the lyrics are meaningless. This isn’t music, this is an imitation of music.
Most of it is, but I’ve been able to meet my needs by digging.
As we always say, it's as bad as it's gonna be. Great overview! Thanks. Suno does this as well but it's in limited testing atm.
omg! Mimd Blown!!!!!
Oh my god. I haven't had a laugh like that in a very long time. Best few minutes of the Internet this month! #asliceoflive #pizzablues #happyaccidents
As far as the AI breakthroughs are concerned, however... Now things are getting VERY interesting! Already thinking of some creative experiments to try with this in my flows. And I'm looking forward to seeing how Suno claps back at this with v4 and their own upload model.
anyone know the song at 17:00?
That’s a piece I wrote in 1990 or so. I think I have it on Soundclick somewhere with other stuff from this video actually…www.soundclick.com/share.cfm?id=4793106
I wonder if Jacob Collier loves online pizza 😂
I wonder if you mentioned you wanted the roaches to sing it, if it would do it
I wish it was free
When I click process it’s like a slot machine.
🍕🍕🍕
14:06 NEXT!!! lol AI rarely gets it right on the first try. At 14:06, I would have been deleting that take and going to the next one.
Go suno test upload bro)
That video is actually being edited as we speak.
O! - M! - G! How stupid do we have to be to give them our songs as training data for FREE?
You left out the part that Udio now owns the song. If you were to upload to Distro Kid for streaming and don't put in the credits giving Udio you break the TOS of them. If your song becomes popular making you money they could sue you and win. The only way around this is the song must be 50% or more different than Udio's output. I wish creators would identify this in their videos. They are not like Suno with their TOS.
Not true lol 😂 I can't give you the gems for free lol
why is your audio suck? it has small volume. i have to maximum my volume in laptop just to hear you
Those Italian songs that Udio made are terrible and sound quality is terrible.
Still too much coincidence at Udio. But not bad.
sounds like it's pasting random phrase samples together to be fair, ha ha ha ha ha.
Again Awful imho , udio. Has no consistency bad controls inly 30 secs... , i like suno better even if sometimes the sound quality us not As good.. Even...
CAN ANYBODY HELP ME ???? Thank you with my post below
TH-cam removal of the dislike count was one of the worst decisions I've ever seen in the history of the internet.