Inference Download Model

Easy GUI v2 (rejekts) - adapted to Mangio-RVC-Fork [With

extra features and fixes by kalomaze & alexlnkp]
[Link] your Model. Optional: You can
change the pitch here
[Link] Refresh or leave it at 0. Convert

Drop your audio here & hit the Reload button.

Index Settings

▼
Jisoo vocal 2.mp3 395.2 KB Download

Output Audio (Click on the Three Dots in the Right Corner to

OR Record audio. Download)

Record from microphone

0:06 -0:10

[Link] your audio.

Advanced Settings ▼
Refresh
./audios/Jisoo vocal 2.mp3
Optional: Change the Pitch Extraction Algorithm.
Extraction methods are sorted from 'worst quality'
to 'best quality'. mangio-crepe may or may not be
Text To Speech
▼

better than rmvpe in cases where 'smoothness' is

more important, but rmvpe is the best overall.

Wav2Lip pm dio crepe-tiny

▼

mangio-crepe-tiny crepe

harvest mangio-crepe

rmvpe

If >=3: apply median filtering to the harvested 3

pitch results. The value represents the filter
radius and can reduce breathiness.

Use the volume envelope of the input to 0.21

replace or mix with the volume envelope
of the output. The closer the ratio is to 1,
the more the output envelope is used:

Protect voiceless consonants and breath 0.33

sounds to prevent artifacts such as tearing
in electronic music. Set to 0.5 to disable.
Decrease the value to increase protection,
but it may reduce indexing accuracy:

Used for male to female and vice-versa conversions

[EXPERIMENTAL] Formant shi! inference

audio

Textbox

Success.
Using index:./logs/bangchan/added_IVF1424_Flat_nprobe_1_v2.index.
Time:
npy:0.38934874534606934s, f0:9.498378038406372s, infer:1.9001662731170654s

Batch Conversion
▼

Use via API · Built with Gradio