I Can't Find Any, I Only Found Old Tutorials

Oh no! The sample rate changer is gone for me too!

Me RN:

Oh, well. You can still select the track, click the "Tracks" dropdown menu, and click "resample".

Oh, and there's also this settings window that opens when you click the "Audio Setup" button

Does It Work On Training?

What do you mean?

I Mean, I Trained My Percy Model With The Wrong Sample Rate, And It Made The Results Slow

Remember to change the project sample rate to "22,050 Hz"

I Did, But It Keeps Putting It Back To The Default One

Are you sure these are the settings you apply?

I Am Trying The Audio Settings One

It Works, It's Now At 22050

Nice

https://www.mp3cutter.com/audio-converter Here's a useful site to convert voice clips to the correct format (converts 20 at a time)

https://audiomass.co/ There's also this audio editor that I use but unfortunately it doesn't have an option to make wav files to 22050 Hz (defaults to 44100 Hz)

js
Generating Mels
---------------------------------------------------------------------------
RuntimeError                              Traceback (most recent call last)
<ipython-input-10-df8aeedb909f> in <cell line: 3>()
      2 
      3 if generate_mels:
----> 4     create_mels()

3 frames
/content/TTS-TT2/stft.py in transform(self, input_data)
     82 
     83         # similar to librosa, reflect-pad the input
---> 84         input_data = input_data.view(num_batches, 1, num_samples)
     85         input_data = F.pad(
     86             input_data.unsqueeze(1),

RuntimeError: shape '[1, 1, 45281]' is invalid for input of size 90562

What this error means?

If you have trouble using any notebooks, make sure you use the correct notebooks, which are linked in the #notebooks channel, as those notebooks are the most up-to-date ones.

I was using fakeyou tacotron 2

Audio isn't mono

Uberduck's notebook has a option to automatically convert the audios

is it possible to control the "emotion" or inflections on tactron? I'm using the Uberduck synthesis notebook, I want certain aggressive inflections, but I can't always get them.

Yes, but on one condition:

you have to train a seperate model for each emotion

Well I notice when I use certain words like "evil" before the dialog that I want, it can change emotion - if I want a more aggressive tone. But it doesn't always come out the way I want.

that's tacotron for you, wu-ception! wheeze

But my point still stands: you have to train a seperate model for each emotion.

I was wondering was there little trigger words, to kinda stir the "emotion"