The BEST Text to Speech Software | This is WILD

hey guys I'm super excited for today's video we're  going to be checking out the script overdub you   may be familiar with the script it's like an  all-in-one podcasting and audio editing tool it   also does voice transcriptions which is really  great but they have this other feature called   overdub which is a text to speech feature so they  have stock voices but you can also record your own   voice give a sample to descript and it uses AI to  generate your voice back to you so you have text   to speech of your own voice and I'm super excited  to try this this descript overdub review is part   of a bigger review I'm doing where I'm actually  getting a custom avatar made by Synthesia so   you're going to be able to see me as a deep fake  custom avatar and then descript provides the voice   but that's not the only reason I'm excited for  descript a lot of times when I'm recording here   for my YouTube videos sometimes like maybe I'll  misspeak and so what I have to do during the edit   editing process is re-record the voiceover that I  messed up on and then dub it back into my videos   and it would be so nice just to have a text to  speech option so I don't have to do that again   it might be really time saving so we're gonna see  if my AI voice sounds at all like my real voice   how natural it is we're actually going to try  to generate a text to speech voiceover and drop   it in to one of my YouTube videos and see if it  sounds natural and passable all right let's just   dive right into it so right now I'm on the free  trial of the Pro Plan but here is the pricing   structure for the script you can see here that  the overdub has a thousand word vocabulary for   the free trial for the Creator tier it's also a  thousand words at twelve dollars a month but then   the pro version is 24 a month and an unlimited  vocabulary now if you want to try the script for   yourself I'll drop a link in the description now  once you sign up for it you have to download the   app and this is what the app looks like I'm going  to hit this blue button here create new voice I'm   just going to name the voice gen and this is where  you create your own voice into script overdub now   you can record your own voice right here into  the descript app and they actually provide these   training scripts for you to read from and they  even go up to 30 minutes in length the script   does recommend at least 10 minute reads for your  AI voice so you could use the record feature but   luckily for me I have lots of samples of my own  voice because I pretty much talk all the time   here on YouTube so what I did was I opened up  a 11 minute project in Final Cut Pro and just   exported that recording as an audio file I made  it a wave file and I'm just going to drag that   WAV file right here into the app and I can set  my transcription language I'm going to leave   it as English and it says that my transcription  will be ready in a few minutes that actually only   took about 30 seconds and I'm just going to do a  quick scan through my transcription to make sure   that it looks pretty accurate now I'm going to  hit this blue button here called submit training   data and now I have to record myself saying  this consent form we are definitely living in   the future here you can't just sign it you have  to record your own voice I hereby verify that and I get a message that my voice is now training  and that they'll email me when it's done through   the tutorials I watched earlier I know that  it could take between 2 and 24 hours to get my   voice back from descript so I'll see you guys in a  little bit either later this afternoon or tomorrow   okay it's the next day it took about 15 hours to  get the email back from descript saying that my   overdub voice was ready I have not listened to  it yet but I'm super excited a little nervous   let's check it out alright so I think what I need  to do is start a new project which I guess I can   do here at the bottom of the screen or over here  with this blue button kind of confusing I think I   want to start writing I'm going to click speaker  okay this is where I select myself as the speaker   and then I guess I hit the play button  I'm confused I don't hear anything   oh wait look waveforms have appeared so I guess it  takes a second to generate the audio once you've   typed it in that's totally fine let's just  let's just listen to it I'm a little nervous   let's hear what this voice sounds like I mean  that sounds like me doesn't that sound like me   let's hear what this voice sounds like I think  the inflection is off but that is 100 my voice   that's a little scary I even added an exclamation  point at the end to try to like you know tweak   the inflection a little bit and I don't feel that  that's reflected in this text-to-speech voiceover   let's listen one more time let's hear what this  voice sounds like I think that's pretty amazing   and a little scary I'm pretty impressed with that  all right but let's do the real test which is   that real life scenario where I fumble something  while I'm recording one of my YouTube videos and   I want to go back in and fix it without having to  re-record everything through my camera mic again   let's see if it matches the sound of the camera  mic I've transcribed a quick sentence from an   existing real life video and I'm just gonna paste  that line in here scroll down in the inspector   window to find the 3D text menu okay this is  wild okay so now I'm gonna hit the I guess the   publish button is that the next step all right  it's going to give me a WAV file which is nice   and let's bring this into Final Cut and really  see if it holds up with my original audio all   right so here's the original audio scroll down in  the inspector window to find the 3D text menu and   let's bring in that overdub voice okay what I'm  first noticing is that the length of the generated   voice from the script is significantly shorter  than the original audio which is fine because   you know when I'm talking and I'm thinking like  sometimes I leave gaps between my words so it's   nice that it kind of tightens it up so let me  mute the original audio you can see it's great   out there if you're not familiar with Final Cut  and then this clip right here is our wave file   from descript I want to pick something not too  chunky and you'll see why later all right let's   scroll down in the inspector window to find  the 3D next menu now if you're not seeing all   of these details here knowing myself and knowing  the way I speak I do think that the inflection in   my like very astute hearing isn't quite right the  way I would say it but I don't think anybody else   would notice the difference like ever this is such  a helpful tool this could be a game changer for   me with my editing I'm like really surprised  by this let's just do one more just to see in total I got seven proposals in a day and of  those seven I selected three I wanted to work with   I think I have too many commas I'm gonna remove  a couple of these commas and see how it sounds in   total I got seven proposals in a day and of those  seven I selected three I wanted to work with all   right let's export it and drop it in and see how  it sounds first let's listen to the original audio   in total I got seven proposals in a day and of  those seven I selected three I wanted to work with   okay that's what I actually sounded like when I  recorded this let's listen to the descript overdub   um about their pricing maybe a little bit  about their Creative Vision in total I got   seven proposals in a date and of those seven I  selected three I wanted to work with I'm really   excited about the three designers I chose I  mean it's it's uncanny like I said I can tell   the difference but I don't think anybody else  would know the difference so what are my final   thoughts on descript overdub I mean obviously  I'm really impressed I don't think that's a   surprise I think this voice 1000 sounds like me  the one critique I would have for the script is   that the UI the user interface is not good I find  it super hard to navigate it's everything's just   very white and I find that when like I'm using an  app that if there's like brighter colors where it   kind of draws my eye about where I'm supposed  to click um I think that that would be really   helpful and to make the script a little  bit more user friendly I don't think the   UI is good at all and I see a lot of room for  improvement but that's okay because the basis   of the technology is a runaway hit as far as I'm  concerned and I'm really excited for next week's   video where I pair this voiceover with my custom  made avatar from Synthesia this is going to get   real weird so if you're not subscribed make sure  that you subscribe so you don't miss that video   if you guys want to check out the script overdub  I will link to it down below thanks for hanging   out with me today I picked out some other videos  I know you're gonna love and I'll see you again

Channel: Jenn Jager

Views: 277,256

Keywords: best text to speech software, best text to speech software for youtube videos, text to speech software, best text to speech software 2022, text to speech for youtube videos, best text to speech, best text to speech for youtube videos, text to speech app, text to voice for youtube videos, best tts for youtube videos, descript review, descript overdub review, descript vs speechelo, descript vs talkia, best text to speech generator for youtube videos, tts software, descript

Id: t0zXc7SGr9I

Length: 9min 9sec (549 seconds)

Published: Mon Dec 05 2022