Hey folks, I forgot to mention that you can't publish on Udio's site the music made with audio to audio at the moment but you can download it and save it locally and publish it anywhere else at your discretion. Also the confirmation part states to ensure that "you have the rights to use and distribute the clip" you are uploading. Be sure to read the TOS, basically by uploading the audio clip you are giving Udio permission to use the track for their own use.
Ok, I was going all in one this AI music thing (because: fun!), but just tried this feature, and not sure how I feel about this. Maybe works a little too well...
It's impressive that it can extend an existing segment of a song. As much as I appreciate these instant song generators, I still itch for some more control, like stem separation.
Agreed! I believe that's being worked on as well hopefully sooner or later. That will really be a game changer too. I hear Suno AI is working on that too.
@@RealityHertz I use bandlab myself for stem separation although it's not the greatest with AI music at the moment. Depends on the output though. Still a good option that's free.
I think it’s very unlikely that we will be able to export our generations as stems, If Ai audio generation is similar to Ai image generation then it’s very unlikely, as the way that works is organising a bunch of random noise into a format of things that make sense to the Ai, based on the input data it’s been fed, we’ve had good Ai image generation for a few years now, but none of them give the ability to separate an image into accurate “layers” (which would be the equivalent of stems). Sure they could probably run it through an Ai stem separator but there would still be a lot of unwanted noise and artifacts of the other stems leaking through into each other, unless they completely rebuilt the model from the ground up, where all the training data they feed into it also have the original stems fed into it aswell, and then Udio rebuilt its program around that (which would be a lot of work)
Thanks for this video and for using some rock oriented examples. I'm wanting to use the ai in almost exactly the same way as your examples. I have a lot of rock song idea tracks with drums/guitars/bass but would love to hear it with vocals that are better than me. I was surprised how well it handled the lyrics in your example. Also interested to see how it works as sounding board for transitions. Trying to decide between udio and suno for my efforts, and this udio demo was great.
I appreciate it! It’s quite like what I intended, the original was more subdued but always meant to do a more aggressive version. I wouldn’t change much to be honest. 🙌🏼
Amazing, this could turn your band into an online rock powerhouse! And later you could either find a singer that is able to sing as good or just perform with playback 😁🤘 great stuff!
I like the fact that there are no stems produced! Do some work...reproduce the parts yourself, I love that, actual musicians might still get a gig where that is concerned at least.
suno has a long way to get to this type of options! Can't believe they didn't have a UI ready with their launch of audio uploads recently! now i'm disappointed with them -_-
Suno does have an audio upload option now, it has some advantages but the overall quality with vocals is still quite lacking. I'll be covering that in Suno as well.
Great video again and a nice voice also! I also find that if you put the effort and time into it it will eventually spit out something good or in Suno’s case, something really good albeit with poorer audio quality. Let’s hope they manage to improve upon this soon. Many thanks
Great job on this video! UDIO's new Upload Feature is crazy good. I did a step by step tutorial video on my channel as well 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.
In almost 100% of my attempts the result was AT LEAST significantly different from the original and the stereo image was altered in a way that the transition from my music to Udio's music is extremely audible. I am not going to use it until it gets better. Though it is a great tool to bring back to life old ideas that stayed in the dark for years. 😊
@@MonzonMedia I tested it with a 1,5 minute unfinished song (with lyrics) and with an 1 minute orchestral song. Both didn't go well. The song with lyrics was almost humiliated (😂😂😂), the orchestral piece was extended in a very awkward way, like I was listening to bad recordings of the original and like the copy of the original was tortured by a sick mind. I am not even close to "pleased" with the result I got BUT I am sure it is going to get better and better in the future. What I am waiting for (and it will come) is the time when VST libraries will be AI driven (an ai mimicking the instruments you want to use). That will be a game changer.
Great video. For the last song, now you have to do an AI video of a warrior approaching a castle in the rain shot from behind. Oops, I can't stop myself from creating prompts. 😊
hahaha! 😁 Trust me, I've been wanting to do video for some of the songs I've made. It's very time consuming though but I hope to do this when time permits.
I've been using to redo some of my suno songs. And my big observation there is how much of an improvement it is over the original ai performances. Some of these songs were rendered a year ago, but when I process them this way, I can get more or less the original song out, just with a lot more energy and vocal range. It's given me a whole new perspective on my music and songwriting
I’ve just been trying that myself and have gotten similar results. But I hope Suno really focuses on the output quality for v4 otherwise Udio will be a step ahead.
Good video, man. Also, you're song is pretty good and your voice is not as bad as I think you think it is. LOL I was kinda hoping the AI was going to add some Alice In Chains harmonies. That would have cinched it!
I didnt know about the existnace of the Advanced bottom, and unsubscribed from audio because of a low quality sound.... LOL!!!!!!!!:_))))))) Thanks for sharing this INFO :-)
that's smart... leave it to the user to upload any voice they want to mimic so when they get sued they can just say "it's not us! it's them the users."
How can i improve the audio quality of the Udio downloaded Wav file? Is there anything that could be done to make the quality better other than the usual compression and EQ?
Firstly, it really depends on what genre you are prompting for and the capabilities of the model. Some genres just sound better than others. However you can try to prompt for "dolby atmos, original master recordings, remastered," to get better sounding generations, it doesn't always work but it helps. Also you can use a program like bandlab or ripx to separate the tracks to edit them with other effects besides EQ and compression. There is only so much you can do with this quality though.
Thanks for the excellent video explaining how to use the upload on Udio. Much appreciated. It's an amazing improvement, a game changer for AI and producers. Please check your statement that is need to be checked when uploading a file. I think you are attesting that you have permission to upload it. I don't think that you're signing for Udio to use it. I hope not!
Yes you are correct I should've been more clear on that, I'll add it to my pinned comment. This new feature is going to be so very useful for me personally, definitely a game changer! Appreciate you checking out the video and giving feedback!
Not at the moment but they say it's something they are working on. You can use bandlab's splitter to do that. I cover it here using Suno. ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-Gt1d8iKOGO4.htmlsi=CCdKwcO9zuh2Q--2
Hey, thanks for your quick reply ✌️ Ok i see. Splitter still have some artifacts but! thats just a question of time. Logic Pro X‘s 11 splitter is also working amazingly good. Very excited about this upcoming Udio Feature. Amazing times for Musicians 😋 Cheers & thanks again
@@jvlbme Yes, I'm experimenting with that...it can work if you get the right generation. Tried inpainting over clips as well to add instrumentals to vocals but no luck.
@MonzonMedia good singing , just not a fan of the lyrics, u could of been big ,never to late ! Good music is timeless, and you playing instruments was next level
All good again I wrote that when I was really young. But lyrics are meant to be interpreted by the listener how they wish. I’m not the best song writer 😊
I've been quite happy with it so far. One issue I am having is whenever I try to download the wave (or video) from inside the song page McAfee says it's a virus and isolates it, but I can download the mp3 from the ellipses on the main page with no problem. Anyone else?
@@MonzonMedia I pay for an annual subscription for it. Seems to do a good job. I'm assuming there is no virus but right now I'm content with the mp3 format.
With all due respect to the comments, you can’t call yourself a composer just by prompting 3 words…..it takes a little bit more to compose music guys….so, out of respect for real musicians, do not call yourselves composers please……
