#How do i make my voice models text to speach (normal convos and singing)
1 messages · Page 1 of 1 (latest)
You can add talking samples in the dataset of the voice model you are making
I have the voice changer and I use it to make my voice different. But idk how to make thoes voices into text to speach or singing.I never done that idk anything about it no idea how to do it or where to start
So you want a tts that talks instead of u in call right?
Well in call ig but I was interested in just testing it out since I saw the viral presidents talking ai's that are trendy and it was funny so that got me into it
So you just want to use a tts voice right? For that check inside #1212368971127590922
@scarlet apex idk how to explain
Obama Eats A Donut🍩
DISCLAIMER: The following video contains AI generated voices of several real people. The A.I. voices here do not reflect their real counterparts. All things said by all characters in this video were not said by these people in real life. I am in no way trying to create false information.
For The best videos, be sure to subs...
Somthing like this
Oh you want to use a rvc voice model too with tts, yea you can do that, for example online if you dont got a good pc you can use ilaria rvc https://docs.aihub.wtf/rvc/cloud/ilaria-rvc/ ,if you got a good pc you can use applio in the local folder of the docs
Last update: Mar 8, 2024
Thank you man rn am busy but tommrow i Will test it out and if I have some problems i Will ping you here or you want me to DM you?
Thanks for the help
Really appriciate it
You can write here dw
@scarlet apex hello curently am watching ilaria's video on "ilaria RVC". Yesterday i also downloaded the program but today i also needed do i need to download it every time? i will ask more questions if i dont get somthing lol
download? you dont really download anything on your pc, i guess you are talking about the google colab cell where it says install ilaria rvc, and yes you have to everytime you re open the colab as its cloud pc, not your pc
but it dosent take up any space on my pc its just basicly like a lunch? btw thanks for quick reply
it doesnt take any space yea, google colab is a cloud platform, you are basically using google's pcs online, so yea its normal you have to download everytime again bc its like you go somewhere and use library's pc just as an example
doesnt use your pc memory/ram
whats the differance between local and public url
local url is the link that runs locally on the machine, but as you cant really see the google pc screen, you need to click the public link which is the only one u can access from colab, so theres really no other difference
on "download voice models">"huggingface or Google drive link: what do i need to paste where do i find the link?
this is what am trying as a test "the rock" and i downloaded it
and now i have it saved on my pc
@scarlet apex
Ayo? @cunning anchor level 3 !!! 
no you dont have to save it on your pc, its on cloud
you can search:
- #1175430844685484042
- #🔍│find-models
- https://weights.gg/
- https://huggingface.co/models (but watch out cus in hugging face there arent only rvc ai voice models)
- https://applio.org/models
find any rvc voice model
and paste the download link in there
dont download it on ur pc
ok i want this
and i clicked download
conected to my google acc
and it saved on my pc
@scarlet apex
for weights.gg links its kinda different yea, you have to download them, put them in your google drive, make the link visible to everyone and then put that link in ilaria rvc
how do i put it in my google drive @scarlet apex
https://drive.google.com/drive/ , click new, load file, select the model .zip, then after it loaded click the 3 dots next to it, click share, and put it to everyone who has the link
Access Google Drive with a Google account (for personal use) or Google Workspace account (for business use).
wym by put it to everyone who has the link?
with the 3 dots i have move>star msg>rename>details
or u mean the big blue shere botton
you should see, share>share then 2 options:limited or everyone who has the link, and you should choose the 2nd
send a screenshot
i did i can hear the rocks voice
she used a voice audio file
but am using text to speach
so the "search feature ratio" do i just leave it?
she changed it to 0
yea you can use that too
okay i think this is my last question
how do i add emotion to the caracters voice (angry,sad,fear.....) and how do i make him sing
its also called as the index Rate, it determines the level of influence of model's .INDEX file:
Higher values will apply more of the .INDEX's characteristics(so like the tones, accent).
Lowering it can reduce artifacting.
by tts? well you can just try to make him to, but tts doesnt really have much options, you could also try to use other tts #1212368971127590922 and put the tts generated output as the input in illaria rvc as some may have better emotions like bark tts, but other than that you cant do much, also with singing you mean still by tts or like ai cover?
both tts and ai cover
and btw when am done with the caracter i can just delete the file on my google drive correct?
@scarlet apex
yes
for tts you can just try but as i said tts dont got much emotion,
for ai cover, its better you follow this https://docs.aihub.wtf/essentials/how-to-make-ai-cover/
Have the audio file of your song ready, & let's extract the vocals from it with an audio isolation software.
is there a video about this?
this for ai covers https://www.youtube.com/watch?v=wh4cLMLLkj0
Finally an updated tutorial!
Keep in mind this video was dubbed and subbed by an AI, see this as a test since mistakes can be present.
Support mehere: https://ko-fi.com/ilariaowo
Colab:
https://colab.research.google.com/drive/16LkwvFZeudTpUOsE_6bMjOq2qkxFo8Hr?usp=sharing
Discord:
https://discord....
for ilaria rvc
in the docs it also tells u how to separate instrumentals and vocals
which in the video isnt deeply explained so thats why i gave you that guide
yw