How To Create Multiple Songs With One Voice In Udio AI Music
Vložit
- čas přidán 22. 07. 2024
- In today's video I'm going to show you How To Create Multiple Songs With One Voice In Uido AI Music. Utilizing the upload audio tool we can recreate the same voice across multiple songs in a particular genre. It will take some effort to get the right results and you will likely burn through your credits but it is well worth it once you get it right!
Try out Udio here www.udio.com/
Separate tracks with bandlab www.bandlab.com/splitter
⏲Time Stamps
0:00 Preview of getting consistent vocals
1:07 Download an existing song
1:20 Isolate tracks with Bandlab
3:26 Why separate the vocals?
3:57 Uploading the audio track and settings
6:09 Example track 1, instrumental with vocals
7:40 Example track 2, vocals only
9:34 Cropping original track and extending
*Disclaimer Affiliate Links Below*
📸 Gear I use
Sony A7C CAN amzn.to/3spWX8C
Sony 35mm 1.8 CAN amzn.to/36Apekr
OBS (FREE) obsproject.com/
Editor Davinci Resolve www.blackmagicdesign.com/ca/p...
Audio: Rode Podmic amzn.to/35sSnxv
🎵 Udio AI Music
🔦 Find us on:
Discord: / discord
Instagram: / monzonmedia
Facebook: / monzonmedia
This method works pretty well, when Udio get's it right. I did burn through quite a few credits to get the right voices. Let's hope that consistent voices gets easier in the near future! Also I haven't been able to use this method and change the genre just yet. If any of you figure it out let me know!
Pro tip: if you want your track in a different key, change your intro track to that key. Otherwise all of your songs will be in very similar or the same key as your starting track. Great video!
Great tip! Appreciate it!
What is the key, and how do you change to said key?
@@MarcelousTV it would be nice if they said what key and tempo the songs were in to begin with
This video is a life saver!!! I ve spent my first 1 month credits trying to get some consistence. But this will be awesome. Thank you so much !!!
You're welcome! Glad it was helpful.
Man, you're a saviour, I was doing many kind of acrobatics for making the next hits of my AI star, and you came with this!!! Thank you from the bottom of my heart, you saved me thousand of hours 💪🏻😎
Glad I could help! Udio is working on a better way to do this but until then, this will have to do. 👍
Thanks for always making concise, detailed tutorials. You rock. I’ve been on Suno since the beginning and now want to explore Udio.
You're welcome! I enjoy both but tend to lean towards Udio more lately because of the better vocals. It's all fun for me 😁😊
The most iiddiiotiicc mindset of those who think udio better then suno infect suno if godfsther of udio thing eveytype if song generation suno is ahead even in quality and thing which this video meant to be@@MonzonMedia
It' works when you use the one with the music, but that's what I was trying to avoid to get some new music, It's cool though, I can learn from here and see what I can do, this was a great start to finding out new things in UDIO.
Excuse me sir, may I just say that you're absolutely a God amongst men for bringing this workaround to my workflow. It's not an exact science, but certainly better than getting a totally random voice altogether. Thank you so much for your insight. Well explained. Great job.
Hahaha! 😁 Well I wouldn't go that far but I appreciate the sentiments! Glad it was helpful.
There is no doubt....BRILLIANT!!!
Thank you! still Experimenting to see if it can be even easier. Or until Udio releases an update the remixing which I hear is coming soon.
Nice tips ! thank you :)
You’re welcome! I’ve got a couple more tips to share soon 👍🏼
Thank you, I needed this tip. I thought of a way to try to get the same voice but not this.
Glad I could help!
I use a similar method to turn my old instrumental tracks into new titles. I uploaded the last 5 songs today. Feel free to listen to them and let me know what you think. :-)
Works pretty well but I find it can go off the rails just as often 😂
Always great content mate, keep it up ! And a big thanks for all your hard work !
What I do is,I create a new song using the same prompt and set it on instrumental. (Or if u want a different style or genre just use another prompt.)
Udio will create a similar style sounding song with no voice, I download this, then use an external app to clip the voice and instrumental clip them together and reupload them as one. Now you have the voice,a gab between ,and the new melody that sounds nothing like the first.
You can Use inpaint to close the gab and glue the two pieces together if you want. Or leave the gap. If you leave the gap udio will just use the old voice in the new song with the new melody.
Works like a charm for consistent voice with new melodies or genres. Give it a try !!
Appreciate the support and kind words! I've been experimenting with a similar process but I also wanted to see how much I can do within Udio first. Thanks for sharing your process, always open to trying out new methods. 👊🙌
@@MonzonMedia Same here ! Cheers
Cool! This is probably is the first video I've seen on CZcams about Udio from a guy who ACTUALY use it! Of course I had the same idea by myself to force voices (and much other) into Udio, but I noticed It doesnt work with spoken words, for italian language at least... (I was trying to clone the voice of a famous announcer for my parody song) May be because Udio is strong on sing and not non speak...
I’m glad you found my channel and thanks for the kind words! I have not tried with spoken word but you could even hum the melody and it will mimic your voice.
If you grab a song, strip the vocals from it and then takes like 10 seconds of vocals, then generate a new instrumental and take 10seconds of that and make a 20 second audio clip with that, it will continue the new instrumental with the same voice from the vocals. Enjoy.
nice tip thanks!
My pleasure!
Love that song there pretty good
Appreciate it! The completed song sounds great!
Your videos are very helpful to me. Thank you. I noticed that after the changes, UDIO, when generating the next part of the song, tries to match the style of the previous fragments too much. Which is not always desirable. Until the moment comes when the UDIO algorithm decides that it's not what you want and "fires" something wonderful. Unfortunately, this costs your patience and credits.
Glad to hear they are helpful! Yeah sometimes it's just random but when it gets it right, it can be magical!
I have the impression that UDIO often generates sensations in which the weight of the instruments is too far to the left. This generally applies to major instruments. The background is usually distributed summarily into the left and right channels.
I’ve experienced this as well and just fix it in post. But fortunately for me it doesn’t happen too often.
How did you get a different groove when your guide track had drums?It seems when i create my tunes this way from my own created it keeps the same drum beat so i usually just use guitar only stems
You can change the initial prompt to influence the song more, play around with the advanced setting strength. I find using just vocals will give better variations.
why is it that when I try to imitate the vocals, the intro only shows the vocals without any instrumental backing, and it wasn't there before
Do you mean you are only getting vocals? If that's the case play around with the lyric strength, lower it to 30-40%. Even prompt strength you can try increasing it only 10-20%. All you need is some sort of instrumental to get things going.
I think I heard you say you did a video on Udio labels but I could not find. Can you link where it is?
There are a few. Here's the unnofficial Udio guide docs.google.com/document/d/1aQW2OBBnyM-MllWzmDzxdP6g9t6HJJxVNHyJDW8L8AY/edit#heading=h.7pptq4m0on6f And an extensive list of tags done by someone on Reddit although your milage may vary. docs.google.com/document/d/1_EKyUvY2RfeOIDSwc-U5hzW4z7Kb2LLDK7R5sTYGhH4/edit#heading=h.4442p49eacou
Hello, when I use the crop tool, the original audio is still in the new song
Did you see this part in the video? 9:38. You have to make sure you select the area you want to keep.
Hello, how are you, where can I listen to the songs created by users in wav quality? Where do you upload the songs created? Apart from the UDIO page itself, since you cannot appreciate the sound in WAV quality, I would like to download a fragment to compare the quality of the mp3 and wav, does it really sound like WAV?
To be honest I don't hear the difference but I haven't really done a thorough comparison. You just have to click on the song title and there are options to download in both formats.
@@MonzonMedia Could you share a song of yours in wav quality through Google Drive? I forgot to clarify that I am a free user. I'm curious to know what the wav version sounds like.
is there a way I can upload say a 20 second clip of my voice, and it can use my voice to sing the song? because my songs sound literally like Billie Joe from Green Day is singing them, which is def a no no if I want to use it commercially
Yes, I covered it in this video. czcams.com/video/cHRwcMLITO0/video.htmlsi=Rg8j3mDzTB3Dio_h. Mind you it's more of a workaround hack and it takes a lot of patience and generations. So you'll burn through quite a few credits. If you happen to play an instrument that helps get things going but if it's just your voice, it can be challenging to get the music going. I hear from the folks of Udio that this is something they are working on for a future update which would be amazing!
How would I get ahold of the audio file for a particular song?
Not sure what you are asking? You mean someone else’s generated song on Udio?
All I get is a capella with no music at all (although a different melody).
I have tried it many different ways, but I cannot get it to generate actual instrumentation. Any idea what I am doing wrong??
I tried adding instruments and ran it with different sections of the audio...but always nothing but the voice. No music at all. ??
It can be stubborn at times, you can try adjusting the lyric strength between 30-40% until you get some music that you like. Again this method is more of a hack so it may take some effort.
@@MonzonMedia
Thank you!!
Hi again. I've been trying to accomplish the following the past few days, with no luck. I thought the video above was the answer, but it turns out I was mistaken. I thought I posted this already but I can't seem to find it now (chalk it up to age). What you did with vocals, Ii'm trying to do with instruments. Part of me feels it can't be done, but your skill level may prove me wrong. Here it is what I'd like to do:
1. Two Titles Involved: Whiskey Eyes (has lyrics and instruments and is fully rendered) and Guitar Intro (no sung lyrics - 10 second guitar intro)
2. The song Whiskey Eyes is fully developed and rendered. But now I want to add my Guitar Intro, but not sure how to do that once the song had been fully rendered.
3. One approach I tried is to upload the 10-second guitar intro and add the custom lyrics of Whiskey Eyes. The only problem is when I do this, the New Whiskey Eyes doesn’t sound like the Original Whiskey Eyes. The tempo is not the same, the medley is off, and the singer’s voice is different. Probably the reason for the singer’s voice being different is because I uploaded a 10-second guitar intro without any vocals of the original singer.
4. So the question is, how can I add a 10-second guitar intro (no lyrics sung) to an original song that has already been rendered?
Thanks
p.s. After your reply to my first query, I was able to create a song using the voice I wanted. Thank you. I appreciate your effort to help others.
p.s.s. I did your transcript in PDF format, making it a bit easier to read. If you want to download and add here for others to download, please do. Here is the download link: drive.google.com/file/d/1iZQ3yr_EbvaayltGRZHJUzEwy9_AZcmq/view?usp=sharing
If you just want to add an intro, you can upload the song, select "intro", enter your prompt and tag it with [instrumental build up] or [guitar solo]. The problem is you will get a 32 second intro, so I suppose you can even crop that to be only 10 seconds, then inpaint the area to be seamless. Might be a bit difficult to explain here in text....I suppose I could do a related video. Maybe a tips and tricks video of sorts?
❤❤❤❤
🤘🙌
I followed everything you did, BUT... When I use an acapella sample I'm only getting acapella results back, I've done it over 30 times and every time its an acapella, and I want to create a new song in the voice with new music. It's captures the words just fine and it the voice is almost spot on, but there is never any variation of music. Great vid, well instructed I'm just not getting the results you did.
It can be very stubborn sometimes. Did you try to put some instruments in your prompt? One thing to try is extend it as an intro but instead of custom select instrumental. That way it will force it to create music.
@@MonzonMediaI followed your video I must have missed that part I will watch it again. Thanks
@@MonzonMedia could you explain this step easier? So I´ve choiced add intro, instrumental, used just short vocal, but didn't work - thanx a lot ... I spend days try to make production with my voice, that I use as mp.3 and upload. would be big help to check out this
Can I add my costume vocal?
Did you mean custom vocal? If so then yes, I go over it here using my own voice czcams.com/video/cHRwcMLITO0/video.htmlsi=TvnAYVjmsTJKlNYv
Hi
Great stuff. Music AI is pretty exciting. I hope it is ok I critique your video and make some suggestions. I think many of us are looking for the KING of Udio when doing videos. I find most of the videos confusing. You do a much better job than most. OK...Here GOES.
If I understood your video, you were basically working with a 10-second intro to create a new song by using custom lyrics with the singing/voice being similar to the 10-second intro.
The 10-second intro had both voice and instruments. This is the first confusing point for me. Let me break it down to be sure I understand.
1. 10 Second intro with voice and instruments
2. You used Custom Lyrics and wanted the singer of the custom lyrics to be similar to the 10-second intro singer.
Wouldn’t the lyrics sung in the 10-second intro also be sung when you played the custom lyrics' song? Once you have the custom lyrics sung by the same voice as the 10-second intro, is there a way to remove the lyrics sung by the intro singer? Basically, I want the same 10-second intro voice to be sung by the custom lyrics without the 10-second voice lyrics being sung.
I hope I’m not confusing you. Let me know if my summary so far is correct or if there is something I’m not understanding in your video. LET ME CONTINUE SUMMARIZING YOUR VIDEO.
Keywords Dolby Atmos and Original Master Recording are great pointers. I noticed you did not include in the prompt what the song was about, only used keywords in your prompt. I also use keywords “singing live” for a different vibe, “natural singer”, “raspy voice” (my favorite singer type), male or female singer. Haven’t figured out duet yet but would love that style also.
I think I grasped what you have done (unless the above is wrong) to this point. I started getting confused around the 6:09 part in the video. You were basically playing the songs you created. What I found confusing was the distinction between the 10-second part and the custom lyrics part. It may have helped if you explained what part was playing. But my confusion may be related to what I said above about lyrics sung from the 10-second intro, with the custom lyrics following. What I want to accomplish is to emulate the singer from the 10-second intro to be the same voice singing the custom lyrics without the lyrics being sung by the 10-second intro. It may require that the 10-second lyrics be removed once you have the singer you want to sing your custom lyrics.
May I suggest sticking to one subject when doing a video. I know using the vocal stem only is related but I think to avoid confusion it would be best to stick to one feature/subject only. Also, it may help to tell us the title of the songs being used. I basically titled the music/lyrics you used as “10 second intro” and “Custom Lyrics” so it was clear what was being used in your demo or when explaining a Udio feature. I anxiously wait for your feedback.
I have another request but this already is too long...sorry...if I confused things.
Hey there, first off thank you for the encouraging words! I put a lot of thought into giving value over flash in my videos so it means a lot to me. And yes of course feedback is always welcome as long as it is respectful as you demonstrated which I applaud you for. Sorry that it's confusing for you so I hope I can answer your questions. Keep in mind this video isn't necessarily a "beginner video" as I've done other videos that explain the process better. If I were to explain every little detail from the beginning the video would be too long. With that being said it sounds like you just want some clarification.
If you look at 9:34 you will see that I remove the original uploaded audio with using the crop tool. From there you can recreate the new song.
As for the key words yes, there is no need to put what the song is about if you have custom lyrics. You only have to do that if you are letting Udio do the lyrics for you.
The audio you hear at 6:09 is the original audio that was uploaded as a sample and at 6:27 you will see onscreen caption "original audio" then at 6:40 caption "generated audio" I even explain it as it happens. Did you miss that part?
As for sticking to the subject, everything I covered was necessary to understand the process and I can only suspect your confusion is due to the fact that you are new to Udio? You don't mention that but if you are I suggest you watch this video. czcams.com/video/h_Ry2VI01fA/video.htmlsi=J1KLpdL5bAHfF5HX
@@MonzonMedia Thanks for the feedback and quick reply. I think you'll be named the "CZcams King of Udio" . When I originally saw your video, I thought it may be the answer to what I wanted to accomplish. It wasn't because my effort is a bit different from what you did in the above video. Before I get to my request, let me say I did watch your beginning video (I think twice). I think now I'm trying to put into practice what I have watched. There's a good chance I may not quite grasp some things, but if practice makes perfect, I'll be following your lead in my efforts to practice what your video conveys. You are right I didn't catch some of the things you pointed out above but at my age there are a lot of things I don't catch these days :)😂
Here is what I'm trying to accomplish. If you can help, it would be great:
1. Two Titles Involved: Whiskey Eyes (has lyrics and instruments and is fully rendered) and Guitar Intro (no sung lyrics - 10 second guitar intro)
2. The song Whiskey Eyes is fully developed and rendered. But now I want to add my Guitar Intro but not sure how to do that once a song has already been fully rendered.
3. One approach I tried is to upload the 10 second guitar intro and add the custom lyrics of Whiskey Eyes. The only problem is when I do this, the New Whiskey Eyes doesn’t sound like the Original Whiskey Eyes. The tempo is not the same, the medley is off, and the singer’s voice is different. Probably the reason for the singer’s voice being different is because I uploaded a 10 second guitar intro without any vocals of the original singer.
4. So the question is, how can I add a 10-second guitar intro (no lyrics sung) to an original song that has already been rendered?
If it is possible, I'd appreciate your feedback.
Thanks again and keep up the good work.
I would like to hear the same with Suno
Definitely plan to!
Great video, misspelled Udio though :)
Oh FFS! LMAO! 😂😁 How did I not see that! Thank you sir, I mean it's only 4 letters!
@@MonzonMedia lol but quite easy to misspell :)
but you didnt tell us uploading Audio is ony for premium members
I mention it in a previous video. Guess you are new here? Welcome! czcams.com/video/cHRwcMLITO0/video.htmlsi=1twcBQ93mRUMIiTG&t=13 It's available for standard and premium plans.
did u use free udio or paid udio?
The upload feature is available on the paid plan.
@@MonzonMedia ok thanks,
Can we make karaoke only?
I suppose so but I haven’t tried. You could remove the vocals after with a stem separator app.
@@MonzonMedia yeah I tried but not good
I suggest everyone binge and stockpile as many songs as possible as the majors have slammed down the hammer and will shut these services down. Gigantic lawsuit just hit both Sumo and Udio and the Majors Alwaysssss get their way. Lets cross our fingers the Open source community can pick up where these two left off.
It will take years before anything is written in stone and besides, they just want $$ so likely there will be some sort of settlement. Generative AI is here to stay and not going anywhere. I do think it's important though for discussion to be had and laws to be made, most of the world is clueless about and not ready for it.
@@MonzonMedia The difference here is, that these platforms directly threaten their bottom line, so I don't see them just taking a payout and going quietly. And Lets be honest, these platforms are generating Peanuts compared to the Major Record Labels. They have insane armies of Lawyers that are ready to bring hell to whomever dares to try to take even a small piece of their Luxurious lifestyles.
Yeah I hear ya. Guess we’ll have to wait and see but I really don’t see things shutting down.
@@MonzonMedia Hey bro. I just thought about our chat when I came across this video. Up until I saw this I thought the record labels were going to have a tough time proving their case. Oh boy was I wrong, these platforms are absolutely done. I dont want to leave a link as it will likely go to spam, but look it up. The title of "Is Suno Sampling Copyrighted Producer Tags? " And you will hear literally to the exact frequency, blatant sampling of artist voices and tags like DJ Khaled. BTW My career over the last couple of decades has been DJ / Music producer. Im here on yt as @djnaydee , I know you're a musician as well.
modern song writing is more fun and faster now. i think corporate had these tools before. Pop star is AI manufactured.
I'd agree with you there. I've been having a lot of fun bringing my songs to life. Eventually I'd like to play some of the parts for real and do some live recordings of some of them.
Is this free or paid
paid
Paid, there is a free tier but with limitations.
Ty! 😊
Great, bla bla bla, can you do anything like that with only FREE stuff? no premium or paid account and only free software and websites. can you do it only from the prompt ?
Using Replay is far more efficient, and free.
There is no replay? Or do you mean remix? If that's what you meant sure, you can do it that way to an extent, playing around with the variance slider, change the prompt and tags can work to a point. But this method gives you more flexibility in terms of reusing existing songs and clipping portions to create more songs. And also if one would want to use a recorded clip from their own work, again this method works best. Either way, whatever works for the user and it's great to have options.
@@MonzonMedia “Replay” is the name of a free ai tool that runs offline where you can make models of actual recorded voices (such as your own) and make any vocal sound like it’s sung by someone else (the model) which is why it’s a shame you didn’t showcase that. My own mother can’t tell when I use Replay- the app :P
Oh the app! hahaha! It didn't click, thanks for clarifying. Funny you mention that...I recall seeing it on a few YT videos not long ago and meant to check it out and see how it works with Udio and Suno. Appreciate the tip!