Clone ANY Voice for FREE Using AI & Your PC

[Music] okay sleepy Joe's get ready to make some memes oh please you can't even make your own breakfast without going into a national crisis oh great it's the bidenator here to ruin my tutorials I just can't catch a break man vote man vote kick vote band please vote ban this guy out of the video I'm gonna call you Biden because all I want to say is bye bye Mr Biden vote ban doesn't work this isn't an OverWatch Lobby I can't believe you don't know this okay so I was wondering how to make these memes for a long time so I did a little bit of research and I found a slow but free way to do it and a paid but fast method this works best with American voices because it did have some troubles deciphering my European accent here's a sample of that I was only nine years old I love the B Movie so much I had all the merchandise in movies I'd pray to Barry b every night before I go to bed but anyways let's make some memes if you have an Nvidia graphics card you can actually generate these voices from text using only your computer completely free using the tortoise text-to-speech software first you'll need to install Anaconda I'll leave download link in the description foreign visit this link and select your configuration for me it's stable Windows conda Python and Cuda 11.7 making these selections generates this command line you'll need to install Pi torch so to do it open Anaconda prompt it should be installed together with anaconda and just paste this command line in to paste it just right click it it'll take some time to install and just type in yes once asked next we'll need to install git the link can be found in the description together with all of the other links I'll be honest I'm not a big git user I'm just changing the default Editor to notepad plus plus and leaving the rest at default I'm not quite sure if this is the best configuration but it worked for me after you've installed git you can open up Anaconda prompt and run these git commands to clone the tortoise text to speech software I'll leave the link to the original author of the software in the description as well [Music] and finally one of the last things that we need is actually the python sound file that you can install using this command that will also leave in the description or the pinned comment down below [Music] and at this point we have an instance version of The Tortoise text to speech software that's going to be able to run on our computer as you can see here it's already pre-loaded with some voices and some voice samples and I'll show you how to load in your own voices and voice samples later in this video okay so let's get to the fun part of actually generating some voices here's how you generate speech using one of the provided voices open up Anaconda prompt and type in this command you can type in the text you want for it to say and I'm gonna use one of the voices provided at random for this and the fast generation preset it will now start the degeneration process after it's done you'll see this results folder where you can actually listen to Three variants that were generated to see which one you like the best hey guys I'm just testing out this voice hey guys I'm just testing out this voice hey guys I'm just testing out this voice now you can listen through all of the voice samples provided and you can actually choose which one you want to use for example there's even a Morgan Freeman preset however you can also upload your own voice samples and add those as a new voice to use the easiest way to do so would be to install a free program called audacity and to get a good sample we need voices without any background noise and around three to five ten second samples once you're inside audacity make sure to export them in the 22 050 Hertz sample rate and add no metadata whatsoever now save these samples anywhere on your computer go into the text to speech program and create a folder with your desired name upload all of your samples there now you'll be able to use them when generating voices [Music] foreign so I'll be honest the quality won't be great at first but we can actually improve it by the way if you're getting this error once opening Anaconda prompt that means you're not in the right folder you can type in CD tortoise TTS to fix it let's improve the quality of my voice by adding emotion into the brackets and setting the preset from Fast to standard hey guys just letting you know that you should like the video so I think this model doesn't really pick up accents and it was actually trained on American voices so it americanizes everything however there are paid options that provide a faster and a better way to do it the one that I used for the intro of this video is called 11 laps American emit here just letting you know that these artificial voices are actually pretty good once you get the hang of it anyways you should have liked the video by now instead of waiting hours it only takes seconds to render out the voices but it still makes me American I've tried out the same service with samples of trump and Biden voices from YouTube and it was what you heard at the start of the video vote pan doesn't work this isn't an OverWatch Lobby I think they turned out like really good anyways I hope you've learned something new because I definitely did and it was quite a fun tool to mess around with my friends I was even able to clone one of my friends voices just by using samples from his YouTube videos with his consent of course if you also want to turn your computer into a personal storage server not just an AI voice generator watch this video next

Channel: Emit.Reviews

Views: 103,140

Keywords: ai voice generation, how to generate ai voices for free, ai text to speech, ai meme voice generation, ai clone voice

Id: GBZlkh4qS2o

Length: 6min 34sec (394 seconds)

Published: Mon Mar 20 2023