#How to sound like models?
1 messages · Page 1 of 1 (latest)
less of your specs or setting issues than the model itself, or try adjust pitch
so try another model
if you have trained the current model, make sure the dataset is consistent enough, not mixed with falsetto/roleplayed voice, or other voices & sounds
Can I train models that people already made?
And if so, how would I approach that?
Like all the models sound similar to the person I am downloading, but it definitely doesn’t fully sound like them. Like their accent is top notch though.
training requires processed dataset which mostly only belongs to the model maker
So is my only option finding another model and adjusting the settings such as pitch and shift?
if adjusting pitch doesn't help, then it's the model issue itself
I suppose there must be many other Juice Wrld models you could search for
One last question. Do you know if models come default to work at 0 pitch?
Or do I have to adjust them all manually to try to match the person I am mimicking?
Such as Stewie Griffin.
Because for all the models I use, I just keep them at 0.
may depend on the kind of voice, e.g. +12 for male to female