#How to get the voice Similar
1 messages · Page 1 of 1 (latest)
the VOICE SAMPLE is what you used to make the clone? then the other two are the output from the clone?
yes
haroon.mp3 is the sample i used to clone a voice
426b* and voice sample are generated from the models
426 is genrated from api
and the other one from web app
You should make the clone again in my opinion, there is a lot of noise on the track so the clone is going to sound similar to the data you provided.
Also try and get more than one minute if you can and make sure you select the option for "remove noise"
@bleak heart Can we schedule a meeting. we are trying to build a product which is gonna use ElevenLabs api mostly. So i have a couple of questions which i would like to ask
@wooden kelp I am just a moderator here in discord, you can ask the questions here and myself or someone else should be able to help.
Yes, the voice clone will sound very similar to the sound of the voice used in the training data.
then why there's so difference and i have also attached the settings i am using
Are you just using one minute of audio to clone the voice?
3 different clips of 1 min
Okay, well unfortunately the IVC is not always great at recreating accents. Maybe the sample could be a bit clearer as well in my opinion, try doing some edits in audacity or something like that to try and remove them before cloning..
Instant Voice Cloning vs Profession Voice Cloning which is much better quality but they need around 1-2 hours of "crystal clear" professionally recorded audio to be effective. Try GPT and tell them what you want to achieve, to be honest I would re-record the audio again though. If you dont have a mic and are just suing your phone you can put your phone in a thin sock or have some other thin material infront of the mic, it sounds weird but it can really help..
lol let me test it out