-
Notifications
You must be signed in to change notification settings - Fork 13
/
.env
49 lines (34 loc) · 1.47 KB
/
.env
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
# Public url from Google Colab, only required if running main_colab.py.
# It should end with .ngrok-free.app
COLAB_URL=
# Experiment name and name of model to use in weights folder
MODEL_NAME=
# whether to pitch the converted voice up or down
# use -12 for female to male, 12 for male to female, 0 for no change. Adjust lower or higher if needed.
PITCH_CHANGE=12
# Closer to 0 mimicks volume of original vocals. Can make volume sound more natural.
# Closer to 1 will be more consistent loud volume.
VOLUME_ENVELOPE=0.25
# crepe to use gpu for best quality
# harvest for ??? (not working)
# pm for fast conversions
PITCH_EXTRACTION_ALGO=crepe
# From 0 (Don't use index file) to 1 (Use index file, more bias towards sound quality/accent of training dataset)
# Used to reduce/resolve the timbre leakage problem. Too high would cause artifacting.
INDEX_RATE=0.88
# which gpu to use
GPU_INDEX=0
### PUSH TO TALK KEY ###
# which key to hold to start recording voice when speaking
MIC_RECORD_KEY=f
# If this is set, the script will hold down this key while playing audio.
# For apps like Valorant with no Open Mic functionality, must be different from MIC_RECORD_KEY
INGAME_PUSH_TO_TALK_KEY=
### AUDIO DEVICE IDS ###
# run the following command to obtain the IDs for your audio devices
# runtime\python.exe AniVoiceChanger\get_audio_devices.py
# The mic you will be speaking into
MICROPHONE_ID=
# The input device to your speakers
# Allows you to hear your converted voice
SPEAKERS_INPUT_ID=