Wow. That buffering pause blew me away. That is showing a real sense of comic timing. That seems quite sophisticated! And here I am using Udio to make formulaic pop. Thank you for these ideas!
It loves just pausing on random lexemes, it actually looks like it is just a very fortunate coincidence - but I don't think there is "intent" behind it, as it were. Still neat, and at some point, we'll definitely see intentionally and novel mannerisms bleed into vocal performances.
Hi!, if you want to get consistent verses and choruses just add to your next verse or chorus, the metatag [verse repeats] above each new verse you add or [chorus repeats] above each chorus, and forget about [verse 1] and [verse 2]
Thanks for doing these videos. Please keep experimenting and exploring Udio to give tips for how to get more control. There’s a lot you can do that you’d never know if you didn’t find someone show you, or just happened to figure something out. I bet it can do even more we just haven’t figured out the magic words, even the devs probably don’t know.
Yea they actually said themselves in the discord that theyre blown away by the things people have made. They didnt realize some of the possibilities it can do.
I love Udio from day one and create a lot of Songs with it on my german channel! Subscribed to your channel and hope for more tips and tricks like these! THANK YOU!
Thanks, for the details, I stumbled on to this by accident & just by messing around I got udio to spit out some decent jokes & a 3 min. routine on the nature of reality 😂 crazy time to be alive…
@@AtomicGains-ql8hgit is scarily realistic. I'm actually in the process of planning a live show which uses ChatGPT in real time to deliver jokes based on audience suggestions and comments provided. This Udio capability has given me plenty of ideas on how to take the concept to a new level. I'd need to make sure the comedy voices and applause tracks can be turned around quickly as the live audience will require minimum lag. Thanks though for the ideas.
It's so damn impressive, that it flew over my head until it was pointed out. I thought it was just a pause for dramatic effect. But if there was video footage of the comedian doing a freezeframe moment where he stops dead in his tracks on stage holding a pose - that would drive it home. I wonder whether it was intentional additional meta joke though or a lucky coincidence that this particular render chose that moment for a long pause? I guess if it's trained on comedy routines, then it's intentional! Very cool!
@@robertjones9598 I think it's showcasing the level of context awareness here. I've been generating some stand-up with this thing and have been getting some wild results that sync perfectly with the context of the entire segment. It's hilarious 🤣
@@KeithPhillips I wonder whether somewhere in the vast amount of sets it's been trained on, some lone forgotten comedian made some pauses whilst referring to buffering. I'd be truly gobsmacked if it just pulled it out from thin air! (Given the repeated assertion that 'no one knows' what truly happens in the black box of neural nets, even their creators, due to too many 'atomic' level firing of 'neurons' to rival our own mysterious and for all intents and purposes 'unfollowable' vast amounts of chemical moving parts in our own brains, a Maxwell's demon type scenario; combined with our own lack of definitions and understanding of whatever constitutes consciousness; I'm secretly entertaining notions of there being a 'ghost in the machine'. That's coming more from the hard materialist view that our own consciousness is a kind of misunderstood illusion though, just a pattern recognition mechanism spawned through a regenerating chemical chain reaction run amok that attributes too much worth to it's own sustainability. That's my intellect talking rather than my core ontological beliefs.)
That's insane, thanks. Makes me want to experiment with the 'Lyrics' function a lot more :) Your speach experiments are a bit scary for me as I imagine a KI talking to me like that :))
As someone who has enjoyed radio airchecks of DJs from the 1960s, I think it would be cool if Udio could do that. I tried it once but I did it wrong and Udio made me a song about a radio DJ.
If the instance is mostly right, but it misses something at the end, it's best to break it at the first syllable of the last word it gets right, and extend starting with that word.
I´m currently feeding Udio with alot of Lyrics that I wrote over the years... Works very very good even in German. Only problem with Udio is that the songs tend to have alot of compression on the "seperate tracks", which can make it tough to make some postmastering. Especially drums are often drowning at certain parts which you can hardly change afterwards. Overall quality is much better than say Suno, but the latter is better with genre mixes and most of all better for post-mastering due to less compression. Needless to say the voices in Suno sound too artificial most of the time. Long story short: Using AI with own lyrics is great and I love Udio :)
Maybe you can tell me how to bring udio to pronounce a single word with another language. I had German text with single English words and it did not recognice that this needs to be pronounced different. Maybe there is some sort of markup like in html to make clear that something is a different language.
can you create skits ex. background music or theme, then a scene with dialogue, and sound effects? Lastly, what about two different genres in one track. ex. a pop vocal going into trap rap verse?
I haven't tried doing background music and then a scene with dialogue...I'll have to give it a try though. And yes you can change genres in a track. Thanks for watching.
Yes you can. It’s easier to start with music and spoken word prompts and put stuff like [spoken word] or [speaking] etc etc meta tags in the custom lyrics which does also have an effect. You’ll need to experiment but if you do that it sort of puts the music in its idea and it’s easier to have both. You might be able to set the look back shorter in advanced tools, as I didn’t have this when I tried it, but when I tried to make it do music after a long spoken word thing it was practically impossible, not impossible, but it got very confused. But when I started with some musical theatre thing I had music at the start and then it had spoken word, it did have some background music over some of it, like a musical theatre piece where you’d also get some comic musical sound effects accenting certain words and phrases (which is crazy because it did it so meaningfully). Meta tags and experimenting with it is essential for making spoke word performances.
@@AtomicGains-ql8hg Im curious why laughter exist without any lyric prompt. Seems like adding the standup comedy prompt adds laughing track automatically. I dont have inpaint so its rough to get a good generation.
You make AI generated laughter under your "own jokes"? That sounds a bit sad, plus making a tutorial about how to use AI for creating self-affirmation mental-masturbation loops sounds like a weird future we are heading towards. Sorry for the provocative comment, I had to. Art is free, my udio-music also contains applause sometimes but I have a self-ironical perspective on it. Seriously, when I created a jazzy song with a live audience, the audience gave me good advice when I generated the saxophone solo. The virtual crowd was agreeing with me about which variants of the solo were the best - I think I had 16 variants of the first half of the solo, and it was a hard choice until that one came where the audience was cheering. The song is named "Die Totale Erleuchtung" in case of interest.
@@AtomicGains-ql8hg Yes I know, it´s just also an interesting metaphor for what can happen when people get lost in their AI loops of self-entertainment instead of actual socializing. That "AI-girlfriend" thing, basically^^
@@AtomicGains-ql8hg Just to be clear, your video was informative. I just don't think the music and comedy these tools produce currently has any value, but that's not a criticism of your work at all.
@@jeffreygordon7194 what dou yo mean by value? in artistic way it depends of the "hand" but is tru it is not yet very "customizable". In money way, there are people earning lot of money with this just now. It is the beginning. Better to be in the path.
Wow. That buffering pause blew me away. That is showing a real sense of comic timing. That seems quite sophisticated! And here I am using Udio to make formulaic pop. Thank you for these ideas!
Glad you gained something new from watching this :)
It loves just pausing on random lexemes, it actually looks like it is just a very fortunate coincidence - but I don't think there is "intent" behind it, as it were. Still neat, and at some point, we'll definitely see intentionally and novel mannerisms bleed into vocal performances.
Hi!, if you want to get consistent verses and choruses just add to your next verse or chorus, the metatag [verse repeats] above each new verse you add or [chorus repeats] above each chorus, and forget about [verse 1] and [verse 2]
Thanks for the tip, I actually did that in my other video about the audio inpainting in UDIO. I didn't add into this video, I should have though.
Sorry for my English is weak so is there any tutorial out there to understand more about meta tags
You're simply the best! Thank you a bunch!
Thank you so much!
Thanks for doing these videos. Please keep experimenting and exploring Udio to give tips for how to get more control. There’s a lot you can do that you’d never know if you didn’t find someone show you, or just happened to figure something out. I bet it can do even more we just haven’t figured out the magic words, even the devs probably don’t know.
Thanks! It's definitely fun to see what we can do with these tools, i'm sure there's a lot more to uncover.
Yea they actually said themselves in the discord that theyre blown away by the things people have made. They didnt realize some of the possibilities it can do.
Oh, I'm gonna have some fun with this.
It's fun to play around with, hope you get some good results!
I love Udio from day one and create a lot of Songs with it on my german channel! Subscribed to your channel and hope for more tips and tricks like these! THANK YOU!
Wow, thank you! It means a lot to us that people enjoy our videos. There will definitely be more videos like this in the future.
I just checked your channel. Why u didn't choose the location to Germany
Bot
@@Eddierath BS
@@PreetamMizar why should i ?
Thanks, for the details, I stumbled on to this by accident & just by messing around I got udio to spit out some decent jokes & a 3 min. routine on the nature of reality 😂 crazy time to be alive…
Yes it definitely is a crazy time to be alive haha. I love how creative we can be with these tools. Thanks for watching.
That 5 second "buffering" in the joke is actually ridiculously impressive.
I know right! it surprised me.
@@AtomicGains-ql8hgit is scarily realistic. I'm actually in the process of planning a live show which uses ChatGPT in real time to deliver jokes based on audience suggestions and comments provided. This Udio capability has given me plenty of ideas on how to take the concept to a new level. I'd need to make sure the comedy voices and applause tracks can be turned around quickly as the live audience will require minimum lag. Thanks though for the ideas.
It's so damn impressive, that it flew over my head until it was pointed out. I thought it was just a pause for dramatic effect. But if there was video footage of the comedian doing a freezeframe moment where he stops dead in his tracks on stage holding a pose - that would drive it home. I wonder whether it was intentional additional meta joke though or a lucky coincidence that this particular render chose that moment for a long pause? I guess if it's trained on comedy routines, then it's intentional! Very cool!
@@robertjones9598 I think it's showcasing the level of context awareness here. I've been generating some stand-up with this thing and have been getting some wild results that sync perfectly with the context of the entire segment. It's hilarious 🤣
@@KeithPhillips I wonder whether somewhere in the vast amount of sets it's been trained on, some lone forgotten comedian made some pauses whilst referring to buffering. I'd be truly gobsmacked if it just pulled it out from thin air! (Given the repeated assertion that 'no one knows' what truly happens in the black box of neural nets, even their creators, due to too many 'atomic' level firing of 'neurons' to rival our own mysterious and for all intents and purposes 'unfollowable' vast amounts of chemical moving parts in our own brains, a Maxwell's demon type scenario; combined with our own lack of definitions and understanding of whatever constitutes consciousness; I'm secretly entertaining notions of there being a 'ghost in the machine'. That's coming more from the hard materialist view that our own consciousness is a kind of misunderstood illusion though, just a pattern recognition mechanism spawned through a regenerating chemical chain reaction run amok that attributes too much worth to it's own sustainability. That's my intellect talking rather than my core ontological beliefs.)
Great video and content mate, subbed ! More of those, please !
Thanks for the sub! There will definitely be more. Thanks for watching.
Fabuloso
Thanks for watching!
Another great video thanx
Glad you enjoyed it
👏🏻👏🏻
Thanks! :-)
Im hooked on Udio! 20 country songs in 3 weeks!!
Wow! It's a very fun tool to use.
I use Udio for story telling with the Prompt: spoken, story telling, documentary, noir, dark, alfred hitchcock.
It does a real good job.
I will definitely have to try that. Thanks for the tip!
That's insane, thanks. Makes me want to experiment with the 'Lyrics' function a lot more :) Your speach experiments are a bit scary for me as I imagine a KI talking to me like that :))
Thanks for watching! Have fun experimenting.
Great comedy tool!!
Glad you enjoyed it!
What a cool video! Amazing what is now possible with A.I. Thanks!
Glad you liked it!
Pretty cool. Thank you!
Glad you liked it!
Words in parentheses (curved brackets) denote backing vocals so could sing or say anything, not just an echo effect.
Great to know, thanks for the tip!
The comedy skit seems almost real. Good work
Thanks! It is getting hard to distinguish what's real and what's not these days.
That first broadway sounded cool I want to hear more of that
Thanks, I'll make sure to do some more experiments.
great _ I was not aware of this function in udio:)
Glad you found it out now, thanks for watching!
As someone who has enjoyed radio airchecks of DJs from the 1960s, I think it would be cool if Udio could do that. I tried it once but I did it wrong and Udio made me a song about a radio DJ.
Interesting, I'll have to try create one! Thanks for watching.
You can literally do that. You have to get the right prompt.
@@Edbrad Cool! I'll to give it another shot.
If the instance is mostly right, but it misses something at the end, it's best to break it at the first syllable of the last word it gets right, and extend starting with that word.
Thanks for sharing.
I´m currently feeding Udio with alot of Lyrics that I wrote over the years... Works very very good even in German.
Only problem with Udio is that the songs tend to have alot of compression on the "seperate tracks", which can make it
tough to make some postmastering. Especially drums are often drowning at certain parts which you can hardly change
afterwards. Overall quality is much better than say Suno, but the latter is better with genre mixes and most of all better
for post-mastering due to less compression. Needless to say the voices in Suno sound too artificial most of the time.
Long story short: Using AI with own lyrics is great and I love Udio :)
Thanks for watching. We are still in the early stages of AI music generation, and I believe it will grow quite fast.
Helpful
Glad you think so!
You could extend any of the instances with add intro, and just put in:
[spoken: male vocalist]
Hello London!
[crowd yells]
[instrumental intro]
Great tip, thanks!
Thanks heaps for this, I never would’ve known 😊
You're welcome, thanks for watching!
Lmfao the second Broadway tune nailed the effeminate vocals, understanding Broadway tunes 😂
I know right! It's getting scarily realistic.
Pretty good considering Udio is still in beta, I like it _ Thank you
Thanks! Yes, I believe we are still in the early stages of this software, will be very interesting to see how it evolves.
The comedy bit about Netflix buffering reminds me of an Amy Schumer comedy special
It was interesting how it added the pause...I haven't seen Amy Schumers special, i'll have to check it out. Thanks for watching.
what AI tool did you use to generate the graphic with the how to spell apple? AI tools that I've used mangle words
I actually used ChatGPT for that image, it did misspell it but i prompted it to correct the spelling...and it did!
Maybe you can tell me how to bring udio to pronounce a single word with another language. I had German text with single English words and it did not recognice that this needs to be pronounced different. Maybe there is some sort of markup like in html to make clear that something is a different language.
Great question, I haven't tried that...but i'll give it a try and hopefully make a video about it soon. Thanks for watching.
Try to put the language in [. ] before the word. This should work. Often enough udio just makes up new languages and words though
can you create skits ex. background music or theme, then a scene with dialogue, and sound effects? Lastly, what about two different genres in one track. ex. a pop vocal going into trap rap verse?
I haven't tried doing background music and then a scene with dialogue...I'll have to give it a try though. And yes you can change genres in a track. Thanks for watching.
@@AtomicGains-ql8hg Sure! Thanks for your dedication on the research because I took a whack at the comedy set and it was hilarious.
Yes you can. It’s easier to start with music and spoken word prompts and put stuff like [spoken word] or [speaking] etc etc meta tags in the custom lyrics which does also have an effect. You’ll need to experiment but if you do that it sort of puts the music in its idea and it’s easier to have both.
You might be able to set the look back shorter in advanced tools, as I didn’t have this when I tried it, but when I tried to make it do music after a long spoken word thing it was practically impossible, not impossible, but it got very confused. But when I started with some musical theatre thing I had music at the start and then it had spoken word, it did have some background music over some of it, like a musical theatre piece where you’d also get some comic musical sound effects accenting certain words and phrases (which is crazy because it did it so meaningfully).
Meta tags and experimenting with it is essential for making spoke word performances.
WHERE ARE YOU GETTING THESE JOKES MAN. I WOULD ACTUALLY LISTENT TO IT !!
Just made with chat GPT.
@@AtomicGains-ql8hg WOW INSANE
wow
Thanks!
I tried narrating a TH-cam video script! This is crazy!
Awesome, thats a great idea.
Thats a pretty generous crowd lol
Lol, yes they are pretty generous for the content they are getting.
That intentional buffer? Skynet timeline confirmed
I know...scary!
Have you figured out a way to get it to do an alvin and the chipmunks song? I've been trying, and i cannot come up with a prompt that works.
I have not tried that, i'll have to give it a go though!
i tried the first tip but it still generates music.
Did you make sure to have the 'manual' setting on? Otherwise UDIO will add elements like music to the track.
@@AtomicGains-ql8hg yes
put the Hello London in parentheses (Hello London!)
Ahh yes, i'll try that. Thanks!
Thought echo was done by putting () on the phrase or word needed to echo and adding at end with () was for adlib. Im sure its prob both.
Yeah you're probably right.
i misread your channel name as atomic glans 💀
Oh no lol! Maybe I should have made a stand-up comedy routine about our channel name hehe.
Some other guy did this a few days ago, I tried it in Chinese, no good. Suno can handle Chinese and Cantonese fairly good.
Yeah I think both SUNO and UDIO have their strengths and weaknesses, I believe UDIO will soon add those features though. Thanks for watching.
What is the deal with airline food....(laughter)
😆
The second one sounds like T.J. Miller.
Oh yeah, I didn't notice that!
I tried a few stand up routines but they came out all gibberish, not even proper english.
lately, without inpaint, lots of moderation errors and gibberish are generated.
Try copying famous jokes, they’ll possibly be in the training data
It's very trial and error right now, can take a few tries to get good results. Try get chatGPT to generate a comedy script and see if that helps.
@@AtomicGains-ql8hg Im curious why laughter exist without any lyric prompt. Seems like adding the standup comedy prompt adds laughing track automatically. I dont have inpaint so its rough to get a good generation.
I did a tape recorded exorcism in Udio xD and 1/4 of the Shrek movie
Wow, never heard of anyone doing that before.
LOL, the first one is literally Tom Segura
Oh yeah, I didn't notice. Thanks!
Are you an AI generated host? (serious question)
Hi, i'm 100% real :-). I guess it's hard to know what's real and what's not these days.
Use punctuation. Especially a period.
Thanks for the tip!
You make AI generated laughter under your "own jokes"? That sounds a bit sad, plus making a tutorial about how to use AI for creating self-affirmation mental-masturbation loops sounds like a weird future we are heading towards. Sorry for the provocative comment, I had to. Art is free, my udio-music also contains applause sometimes but I have a self-ironical perspective on it.
Seriously, when I created a jazzy song with a live audience, the audience gave me good advice when I generated the saxophone solo. The virtual crowd was agreeing with me about which variants of the solo were the best - I think I had 16 variants of the first half of the solo, and it was a hard choice until that one came where the audience was cheering. The song is named "Die Totale Erleuchtung" in case of interest.
I appreciate the comment, I just test AI software to see what is possible to create with it. Thanks for watching.
@@AtomicGains-ql8hg Yes I know, it´s just also an interesting metaphor for what can happen when people get lost in their AI loops of self-entertainment instead of actual socializing. That "AI-girlfriend" thing, basically^^
I'm not diasgreeing that it's interesting. But it's all garbage. Terrible music and terrible comedy.
I didn't set out to create a masterpiece or a comedy gem, it was just there to showcase what can be done.
@@AtomicGains-ql8hg Just to be clear, your video was informative. I just don't think the music and comedy these tools produce currently has any value, but that's not a criticism of your work at all.
@@jeffreygordon7194 what dou yo mean by value? in artistic way it depends of the "hand" but is tru it is not yet very "customizable". In money way, there are people earning lot of money with this just now. It is the beginning. Better to be in the path.
Thanks bro. Don't forget to read the holy Quran once in your lifetime. You will thank me on the day of judgement.
xd
Thanks for watching!
prompt :
Migos.
custom lyrics :
bippity boppity, bippy boppy?
bippity boppity, bippy boppy.
bippity boppity, bippy boppy.
bippity boppity!
bippy boppy!
bippity boppity
bippy boppy.
you're welcome.
Thanks! I'm sure it sounds great.
@@AtomicGains-ql8hg its just for a good laugh. just good fun to hear dont waste credits haha.