This Python package enables you to extract an instrumental from any public Youtube video (or from a playlist, file or from a folder) using Deezer's Spleeter AI.
Import Stem from Any2Stem.classes and create a Stem object. Currently Stem provides 4 ways to extract instrumentals.
Example code
import os
import sys
from Any2Stem.classes import Stem
# Gets the current execution path
execution_path = os.path.dirname(os.path.realpath(
os.path.abspath(sys.argv[0])
))
def main():
# Creates a stem object with path being the execution path
stem = Stem(path=execution_path)
# Extracts a video
stem.extract_from_yt(
"https://www.youtube.com/watch?v=2N--RorISzo",
output_path=os.path.join(execution_path, "audio_output")
)
# Opens the last added folder
stem.open_last_folder()
if __name__ == "__main__":
main()Functions
| Function | Description |
|---|---|
extract_file |
Extracts one file |
extract_from_folder |
Extracts all files in a folder |
extract_from_yt_playlist |
Iterates through all videos in a given playlist and extracts them using extract_from_yt |
open_last |
Opens last added song |
open_last_folder |
Opens last added song folder |
Parameters
kwargs will always be passed to extract_file.
__init__ parameters |
Description |
|---|---|
path |
Working path |
stems_type |
Will be passed to Spleeter.separate.Separate, default = "spleeter:2stems" |
bit_rate |
Will be passed to Spleeter.separate.Separate, default = 320 |
codec |
Will be passed to Spleeter.separate.Separate, default = "mp3" |
extract_file parameters |
Description |
|---|---|
file_path |
Path to file which should be extracted |
output_path |
Output path |
print_progress |
Wether the current progress should be printed or not |
automatic_cleanup |
Wether the tmp_path should be cleared or not |
extract_from_folder parameters |
Description |
|---|---|
folder_path |
Path to folder which contains the files |
output_path |
Output path |
skip_same_names |
If True, songs will be skipped, if the name already exists in output_path |
**kwargs |
Will be passed to extract_file |
extract_from_yt parameters |
Description |
|---|---|
link |
Link to Youtube video |
output_path |
Output path |
tmp_path |
Temporary path where downloaded and extracted videos will be saved |
print_progress |
Wether the current progress should be printed or not |
skip_same_names |
If True, songs will be skipped, if the name already exists in output_path |
fast_forward |
If True, songs will be skipped, if the title (Youtube video title) already exists in tmp_path |
**kwargs |
Will be passed to extract_file |
extract_from_yt_playlist parameters |
Description |
|---|---|
playlist_link |
Link to Youtube playlist |
print_progress |
Wether the current progress should be printed or not |
**kwargs |
Will be passed to extract_file |
You can do whatever you want but you have to give credits.
I'll add more to this documentation, but I haven't much time currently, if you want to help me, just contact me.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent