Replay: The EASIEST way to create AI Cover Songs!

[@flosstradamus]

Hey!! Im part of the team that made replay! thanks for making this video!

[@BlairBryngelson]

Cool!

[@BlairBryngelson]

Can you share some BTS? How big is the team? What do you do? Is there anything you can share that isn't announced?

[@BobDoyleMedia]

I’d love to connect and chat if you’re willing!

[@thomashannoveraner356]

Can you please tell me, why i get a malware trojan horse error on the replay. exe? So its too dangerous for me to install on Windows 10.

[@nolow_life]

@flosstradamus I noticed weights has many of the replay features built in. What is the benefit of using replay rather than weights directly?

[@MisterDivineAdVenture]

9:15 I wasn't all that impressed until you put that female voice into play - OMG I can't believe it. This is the new Band in a Box. Vocalist in a Box I guess. Holy cow.

[@delcoy71]

Right and the even better news, this isn't the only program that can do this. I saw another video with a guy doing very similar things and getting amazing results. Honestly as a vocal arranger this is gold for having a vocal demo of your song for those wanting to purchase and use it. That usually requires a lot of extra work but now you send a flow sheet showing the lyrics and/or pattern (intro verse chorus etc..) and the demo with parts split for easy learning. This is truly the best era music techwise.

[@MisterDivineAdVenture]

@@delcoy71 Right! To get close to this in a sample library (and no actual lyrics!) was going to cost me me $400+ that I had to put off and off and off. This is just for my fav house-sounds.

[@fret-less]

Unfortunately not. You still need a good vocal track to begin with. This won't add any expression like dynamics, timing, breathing etc like a real singer would. (You know all the hard stuff that singers do). It simply replaces the sound of the voice.

[@briansimmons3349]

​@delcoy71 you should see what we are doing in 50 years

[@RimeBeatz]

This is some next level stuff man ! As a beatmaker, I can only imagine how many different ways there is to utilize this and come up with crazy ideas..! Thanks for this video !!

[@hancockluxellelin9144]

Bob, I have no idea why, at first I thought this video was scammy but then you ended up helping me make music I got extremely excited about. Thanks much for the concise info

[@moilimyacoub4580]

Just discovered about Replay today and I'm really blown away. Kudos to the Team behind. Honestly its sure that down the road Replay will have Pay options like any other app that comes FREE first but serisouly, we must encourage this developer to share an almost "perfect" tool for free with all these options... its simply mind blowing and YES this must be used as a TOOL. And as a Beta tester for major Vst Audio plugins companies, REPLAY is a MUST for post-production / Podcast creations and Social media contents. Bravo> Thanks for also making this video Mr Doyle ;-)

[@BobDoyleMedia]

Thanks for watching!

[@daz71]

And I'd intended on being real productive this weekend! Now I'm producing Miley Cirus singing Sound Of Silence! This is truly magic stuff!

[@BobDoyleMedia]

Yeah it really gets the dopamine going.

[@wormjuice7772]

Finally me and my buddies can have sweet vocals on our drum & bass tracks. Thanks for showing this!

[@daxtonbrown]

This is life changing for me. I've been writing an AI themed novel for 32 years which has now merged with reality. There is a female singer who is the femme fatal. I have the lyrics and the music, but now I can do the female voice and orchestration. I will be able to make an entire movie with this technology on my own when combined with video generating AI like Sora. Mind bending.

[@Octoberisms]

You definitely can make the movie! I have been using Haiper AI for video production until Sora is released to the public .

[@daxtonbrown]

@@Octoberisms Thanks!! I already hav started making some small clips. Can't wait for Sora.

[@ad_burner_sc]

Mind blown is an understatement!!! Thanks for the heads up dude !!!

[@WyvernParadox]

Thanks for using Replay!

[@Fastlan3]

Why does Windows Defender want to stop me from installing?

[@WyvernParadox]

@@Fastlan3 Its a false postive!

[@Fastlan3]

@@WyvernParadox understood.

[@jimmysrandomness]

Thanks for this Awesome detailed tutorial 👍

[@OdinsHarem]

That "I took off to New York in a summer breeze…" sounded _so_ good with Billie Eilish!

[@frankieAllan-vr6qn]

I also thought it was brilliant but Ive tried a few voices even the Dean Martin and it just doesnt compare to Bobs in fact it comes out poor..I see quite a few here are pleased with it so can only assume Im doing something wrong

[@pico12-xq2uh]

I'm blown away by this easy AI song magic!

[@charlesnardi5750]

Yep- mind blown. 🤯 Liked and subscribed!

[@BjutifulMinds]

Your voice is just amazing! And thanks for the tutorial,you just got a subscriber

[@eafortson]

The UI for this tech has come a long way in a short time. Really impressed and cautiously curious and interested in where this is all headed.

[@andrewmartin8565]

Thank you very mutch for your recommendation of "Replay". The program runs like a charm on my 13 years old Mac mini with 16 GB. The combination of "Suno" and "Replay" creates a universe of amazing new possibilities.

[@superwassou]

You gotta that US radio voice, really. 😮 Your video is perfect to discover this tool and the way to use it. Welle done.

[@majahanson311]

When you're transposing the instrumentals, it might be worth splitting the drums from the rest of the instruments. Transposition is a lot more audible in drums, so if you leave them where they are and only transpose the other instruments, you can often get a cleaner result. Also, if you try adjusting the voice by 7 semitones instead of 5, that will be a perfect 5th interval. The vocal might just work with the instrumentals without transposition. It's worth a listen at least

[@BobDoyleMedia]

Great tips!

[@TruthSurge]

? not if more than about 5 notes are used in the melody. Some keys have note overlap but I would be surprised if a song only had 5 distinct notes.

[@skycladsquirrel]

Awesome! Awesome! Awesome! I can’t wait to start playing with this!

[@MikkelGrumBovin]

nontalent

[@obitrice5790]

Bob, you're a great man. Your work is greatly appreciated.

[@SteveSherron]

Thanks very much for this video. This will be very helpful to me. I could have never done it without your help.

[@cagnazzo82]

I'm writing a story using Claude Opus. And I was thinking about putting elements together to probably turn it into an animated youtube channel. I did not search for your channel and this randomly popped up, and it's exactly what I was thinking about (but didn't know existed). Wow. Thanks for the tutorial. Definitely gonna have to try this out this weekend. Subscribed!

[@BobDoyleMedia]

Thanks for watching! So glad it's helpful!

[@NonsensicalGaming]

I'm writing a sci-fi, dystopian cyberpunk novel with Claude Opus, 130 pages down in two weeks... I dreamed of us having artificial intelligence when I was a kid, and I'm definitely not disappointed. Good luck on your story project!

[@olusegunmaximus2241]

Gosh!!! I just got to the merge models and it's soooooo insane!!!!!

[@samkang]

Replay is so convenient to use, thanks for your sharing

[@BobDoyleMedia]

My pleasure!

[@LithiumProductions]

Just found your channel...instant subscribe!!! Is there a way, or do you think they will implement a way to upload your own music and typewritten lyrics to SUNO and it creates just the vocal track? Or just upload your lyrics and create a vocal only track?

[@joelake7986]

Many thanks for this. I had a song about the hula hoop that I thought would have worked better with a Little Eva vocal, instead of the one I had. Unfortunately, it also had background vocals which were pretty integral to the song. Little Eva came out great, but the background vocals came out sounding like a croak. For some songs I think will be very useful to me.

[@mikesrandomchannel]

Just incredible, thx so much for this. Please make a model of YOUR voice available too, thx 😆 Subscribed so fast I pulled a finger muscle… 😅

[@ClassicMusicianChannel]

I don't think you have to worry about using other voices. There are plenty of cover artists on KZbin. BTW, turn up the volume on the music so that it matches your narration volume.

[@wahyung9669]

Oh yes make a duet.Cant wait to try this out, love replay ability to transpose in semi tones on both tracks. What im look forward to is using this with Melodyne/ Autotune to add another singer a male/female duet and having them harmonized together. I can imagine the slurs, counter melodies (would a challange)mthat adds more fun to an already known tune to impress a few privately. Love Ai, get what you wish you never had in your wild imagination. Why not even add a SATB section using stems to seperate the rythme/bass. Well easily said, than done. I guess some seperated instrumental track varies so much in volume is due to side chaining to duck the intruments in the studio (so the vocals dont get burried and cut tgrough the mix - depending on frquency clashes). I see that as a positive and the voice you replace you dont need to redo what already been done by the engineer.

[@BobDoyleMedia]

Yes, adding harmonies is something that I'm playing with for sure, but I don't have Melodyne. A clunkier way to do it would be to use the Autoharmony features in Band in A box, which I will probably try at some point. I'm actually posting a "duet" video next, but it's not really harmonies. Just different octaves. Still sounds nice though when editied together...at least I think so.

[@vjaycanada]

thanks again bob for another great ai tool, YOU ROCK !

[@HubertEtAlice]

Really cool stuff. I am discovering your channel. Thank you so much.

[@ElaraArale]

LETS GO THE MAGE IS BACK WITH WISDOM.

[@dmitryy2199]

Bob, you are my goto channel for creative AI audio uses!, Thank you!

[@MikkelGrumBovin]

F... !!! AI - its lazy , dumb, cheapness - fake , and WILL ruin all humanmade art ! - You should be ashamed of yourself - pushing this evil BS ! - We will - from now on - never know if an artpiece , a song , a piece of music - is Human crafted or machine made - a 5 minute routine , input a few words , suggestions - and VOILA ! Youre suddenly an ARTIST !? - Its criminally lazy , and utterly shortsighted ! We (us musicians) are out of work in 5 years from now - and - human thoughtful handcrafted art, will forever be only a Myth ! - and you dont givafuck ......

[@MainelyElectrons]

This is awesome, I’ve been having a blast with this!

[@Jonecky3115]

Thanks a lot for this, was looking for good solution and this is just perfect!!

[@AdamIverson]

Excellent tutorial, but please be careful with it. Whatever you do, DO NOT use the song owned by UMG (Univeral Music Group), they are known to copyright strike anyone with AI cover of the song. They WILL copyright you, even if it falls under fair use, doesn't matter if it's direct song or AI cover of someone else's cover. My AI cover KZbin channel was completely terminated because of copyright strike with no way of fighting it back.

[@BobDoyleMedia]

Agreed, this is a dicey area to be sure. I was nervous enough to post as it is! But I do believe that this kind of technology can be used in legitimate ways, so I share it.

[@Cre8tvMG]

I hate UMG. Government needs to fight back for consumer rights to derivative, satire and educational works.

[@thedevilsadvocate5210]

I thought you got three strikes

[@AdamIverson]

@@Cre8tvMG Fair use law was supposed to fight that, but apparently, that doesn't stop UMG from abusing the DMCA. Agreed, this need fixing.

[@AdamIverson]

@@thedevilsadvocate5210 It did, on my other KZbin channel. I don't use the account I'm currently using to post comment to upload video and also to separate my identity.

[@CarlosGonzalez-lw7cu]

Hey Bob! Awesome video! Can’t wait to play around with this. Question… do you or anyone reading these comments know of any software to improve the quality of the music portion (not the vocals) after you’ve separated it? To get rid of some of the noise and sometimes tinny sounds you hear in suno music.

[@Moritz.87]

Cant help you for your specific question but you can upload the instrumental to vocalremover to find out the BPM. Then you can download instrumentals (or buy the licenses, so you have the distribution rights) and put all together in Adobe Auditon or what audio editor you like. Not the answer you wanted to hear but maybe a solution...

[@Dreamovie.creators]

Hi great video thank you!! If I want to create my own model, how do I train it? do I need to upload a long audio file? specific words maybe? cause every file I upload, it still doesn't sound like me

[@MrBatooon]

Thanks, this SW is pure gold

[@irlsolidspace]

brilliant, so much fun. Quick question: I make internal corporate presentation videos. I am looking for a tool to upload my video to and I would mark at what times the video has important transitions and changes. The AI tool I am looking for, creates a music from prompts but implement audio changes using the time markers I set. Any advice?

[@martindevin5914]

@10:03 That Billie Eillish take is incredible. 👏👏

[@JimCPettigrew]

Looking forward to this - thks for posting 👍 Don't know if this is a bug but - just started to use Replay 3.91 - when I click on the microphone symbol (Record your own) nothing happens - your video at 14:42 shows a box open saying start recording any ideas please?

[@Fastlan3]

Did Windows Defender pop up and want to stop you from installing? THis happened for me.

[@EmileChill]

amazing, got an entire broadway musical soundtrack made in 10 minutes lol CRAZY!! Thanks!! This is great for my full AI movie.

[@BobDoyleMedia]

Have fun!

[@EmileChill]

@@BobDoyleMedia i'm blown away, crazy stuff thx again

[@2BsWraith]

It can have a bit of trouble with thinking instruments are vocals for a split second and also has trouble with reverb, but otherwise this is the best ive seen and experienced yet.

[@autofashi0n]

Dude this is just what I need! Thanks

[@nicomuziek]

Finally someone giving us clear AI information. Thank you.

[@knivesout33]

You don't need these GIF inserts. Trust me. Your content is good and relevant enough. ;)

[@TheTMax]

Very interesting! Would it be possible to use this workflow to create a parody of a song, changing the lyrics, but having the original artist sing it? What about other languages, or is English the only supported language?

[@pisandonalingua]

Hi Bob! Excellent video! Do you know what AI tools I can use to change the vocals generated by Suno for my own voice?

[@BobDoyleMedia]

You can use this or any other RVC model player, so if you have (or learn) RVC (not hard) then you could do that.

[@deandresnago2796]

This really is amazing you should have said it's the second most easiest since you showed the weights Gigi website which probably is easier to do since you just plop in the KZbin link and pick a voice

[@dinfluence30]

Great stuff to experiment..very fun. Thanks

[@TruthSurge]

video card to process audio? I've never heard of that before.

[@PraiseChristAlmighty]

Sounds like processing bitcoin mining 🤔

[@Onticxp]

You don't understand what a GPU does then

[@darksoulnj]

Great tutorial! Can you do a quick addendum on how to do duets? Thanks for the great info!

[@eranfeit]

Very nice. How to create my own model for my voice using this software ? Thanks

[@OneroomBeatz]

What I try to do with Udio is for instance create a Michael Jackson track 80's etc. then let ai seperate the vocals and istrumental and change the vocal to MJ original voice or simply use it as an Acapella. ai can be great for creating Acapella's, too bad you don't have fully control of what you're getting with Udio.

[@saurabhaacharya1749]

Hey mate, what is the best model to upload your own song, then get sample voice of any one and add that sample voice to the song basically it is a need to add any voice to any song. Appreciate some guidance here please.

[@p_p]

woah last seconds really made me push the sub button lol im having so much fun with replay too! how you did that last seconds video?

[@RastaRecords]

Hey . Im from Poland . I got questions. I saw on facebook that someone upload song of popular polish raper with changed lyrics . The song sound like original but some words are changed . Im wondering how they did it ? Do you know any program that can do that?

[@unyro]

Thank you! How can you change lyrics with this technology within specific track? Is there a solution?

[@niemandschuldet]

That was funny, informative, logical, entertaining, and pretty cool! Thanks!

[@RideShareRocks]

Wow. This thing is amazing! Great vid dude. Subbed!

[@BobDoyleMedia]

Thank you!

[@Fastlan3]

@@BobDoyleMedia Could you please address why the Microsoft Windows Defender pops up to stop installation. Did you get this warning?

[@BobDoyleMedia]

@@Fastlan3 Yes, I got this warning. Over the years, I have seen this MANY times. I've never been harmed by any software I've had this warning with, but I've also only ever done it with software I know has been around a while and been tested by OTHERS first. Once I know its been commonly used without problems, I go for it. It's a fairly common message when you install software from some of these indie developers.

[@Fastlan3]

@@BobDoyleMedia thank you for the reply. It being free and getting this warning, was concerning. I believe the creator can contact MSFT to have them review the software, and they will remove this warning, which I think is important. We need these warnings to be legitimate given all the scams and viruses out there.

[@NSCCYT]

Can you have your custom voices too?

[@shamrockisland]

Love the video 😂😂. You got a new subscriber

[@sidtube12]

Great job guy! I like it very much.

[@Sunshinecoastphotography]

Great video! I just subscribed! I have a Suno song that I'd really like to try another voice iin, but there's a few harmonies in the song and Relay doesn't seem to do well with them.

[@Pheakdey2000]

Hello Dear i got an error request So any problem with it?

[@AntonioRagusa70]

How can we put our own voice on a song created with Suno or Udio? If possible, could you make a video about it? Or is it already covered somewhere in your channel? Thanks in advance, and congratulations!

[@UnchartedWorlds]

Bob We can also use model that was trained on our own voice with Replay? For example model based on our own voice we created using Applio?

[@BobDoyleMedia]

Yes, it’s the exact same technology.

[@UnchartedWorlds]

@@BobDoyleMedia Perfect, I'll become a singer now! :)

[@hailchellwallace1183]

Can you make a tutorial for remaking/remixing instrumentals with AI. I don’t see videos like that anywhere

[@jorgeauden1698]

Outstanding! Do you work with spanish voice-overs?

[@MarcelousTV]

Listen, i cant tell you how much i loved this video and cant wait to get to my computer to download it. Im an avid Suno user and have created some amazing stuff with Suno and Audition and am looking forward to adding Replay to the mix. Humbly and truly, thank you. I can’t wait to check out your other videos. Thanks again!

[@pizzaiq]

This is fantastic for narration.

[@louiserichard1729]

Good video Bob. I got a question please : what is the cheapest Nvda graphic card I need to create models ?? I saw one is 1000$ + but is there a decent card less than 1000$ ?? Thanks a lot Bob and that’s everyone here that could help me on this? Thanks

[@cboscari]

There goes all my free time. Thanks a lot, Bob. ;)

[@SuperAkiliz]

Nice! I have my own voice modeled as a pth file. I hope this works better than RVC

[@BobDoyleMedia]

Well, I’m pretty sure this is just an elaborate player for RVC but would love to know how it works with your model.

[@Digimattik]

Great video Bob - thank you. Question - do you know if you are allowed to use these voices in commercial projects such as in your own music or videos - are they essentially royalty free and allowed for commercial use?

[@DreamingWhileAwakeMusic]

Great information and video, thanks for posting ✨🙌

[@coquibunny8797]

What if the song has backup singers? Can they be separated individually? Keeping one but discarding the other? What are the system requirements?Any phone apps that can do this?

[@tomhite3510]

Really amazing new tech.

[@phantomreacts]

Somehow i can't drag and drop the module after I've extracted it. Pls help

[@nalisten]

Powerful New tool for the arsenal. Thanks for Being Alive🙏🏽

[@BobDoyleMedia]

Thanks. Beats the alternative I guess. :)

[@tsis-k-koj]

🤯 wow thanks for sharing this.

[@willscorner8423]

Why separating within the Programm? Most songs are available as instrumental versions on the internet anyway. As well as isolated vocal versions which can be found easily as well.

[@scotter]

This rocks! Apologies if I missed something: Can I enter my own voice as a "speaking model" and have it made to sing as a vocal track for a song?

[@mdoerkse]

I had to come find a KZbin video because the Replay website is useless. Thanks, nice demo.

[@viewpoint9100]

i think it's great for animation, you don't need to pay for voices, it's just you, your scripts, and your art

[@mikekellyco]

This is AWESOME!!! Thanks Do you know if there is any app that will let you change some or all of the lyrics to an existing song? EX: Car Wash by Rose Royce. I'd like to change the lyrics from working at the car wash to working at the golf course. Thanks

[@notme5744]

Find anything?

[@tubeman1030]

Getting error: Microsoft Defender SmartScreen prevented an unrecognized app from starting. Running this app might put your PC at risk

[@danielbanic3738]

This is great. Is there a way i can make my own voice models with cloned voices i created ?

[@BobDoyleMedia]

Yes, and I’ll be posting a video on that either later today or tomorrow.

[@danielbanic3738]

@@BobDoyleMedia great thanks

[@3dvfxprofessor]

Have you tried Metal songs? I gave it a test with "4 Non Blondes - What´s Up (metal cover by Leo Moracchioli)" and it failed badly to cover the voice. ;-)

[@Salty_Knuckles]

Excellent video. Thanks. You got a comment, thumb and sub! Edit/update: Windows defender is throwing up a scary warning about this installation file. I’m doing a manual full scan.

[@Wyl7]

any update on the scan results? So much “free” software is packed with tech to collect and offshore your personal data.

[@briccs3830]

any update?

[@NiclasLinden]

I get this message while trying to log in to weights "Server error There is a problem with the server configuration. Check the server logs for more information."

[@YakovJames-y6w]

same here

[@backyq4926]

thanks for making this incredible tutorial!! but when i was trying to create a new song it said "An error happened while trying to locate the file on the Hub and we cannot find the requested files in the local cache", do you know how to address this issue by any chance? 🥺🥺🥺

[@garaiss]

Is this also possible for other languages too?

[@ARogers]

I'm not sure this software helps me with what I'm trying to accomplish. Maybe you can help. I re-wrote the lyrics to "What It's Like" by Everlast to be entirely about my cat. I have the instrumental to the song. How can I record myself singing my new lyrics but sound like Everlast/Whitey Ford on the instrumental?

[@anonnymouse2402]

I would love to convert an audio book, using a different voice model for each character.

[@vegastrina]

I don't know if I missed this, but I did go back to your download file link just to see where your downloads were going. Is there an option to import an audio file already on your system? I don't see why not, but let's pretend I have an mp3 of a voice I'd like to use, would there be an issue?...or are there restrictions? Thanks.

[@MegaDoyle666]

I always get the error message "numpy.ndarray object has no attribute device" when I start processing a song.

[@amkire65]

Great video once again. I can see this is going to be another lost weekend (worse with it being Easter) due to AI! I currently have Bing Crosby singing a Suno-generated Sea Shanty in the background. So many musicians I'd never have listened to before getting hold of software like this. The only problem, though, is there are bits of the song where his voice breaks like a teenage boy going through puberty! lol Some is probably down to the separation of the vocals from the music, but seeing we are still at such an early stage in the life of this technology, it's doing a great job on the whole. Where the voice breaks, I wonder if changing the settings would help in those parts of the song, then I could edit those into the version where the majority of the song sounds good?