Sur d’autres sites (7155)

• How to save synthesized audio during streaming distribution

  7 avril 2022, par its-ogawa

  I would like to stream a composite of microphone audio and digital sound sources by ffmpeg and save the delivery to an m3u8 file.

  


  Below are the commands I have actually tried.

  


  ffmpeg -rtbufsize 100M -f dshow -i video=<my webcam="webcam">:audio=<my microphone="microphone"> -re -stream_loop -1 -i <my sound="sound" source="source"> -filter_complex "[0]aformat=sample_fmts=fltp:sample_rates=44100:channel_layouts=stereo,adelay=2100|2100,volume@voice=volume=10dB[voice],[1]aformat=sample_fmts=fltp:sample_rates=44100:channel_layouts=stereo,azmq,volume@bgm=volume=0.2[bgm],[voice][bgm]amerge=inputs=2[out]" -map 0:v -map [out]:a -f mpegts -flush_packets 0 udp://XXX.XXX.XXX.XXX:XXX?pkt_size=1316 -f hls -hls_time 5 hls.m3u8&#xA;</my></my></my>

  


  I have played an m3u8 created this way, but I cannot hear the digital sound source that I am supposed to have synthesized.
Even though the synthesized audio was audible when streamed.

  


  Perhaps you need to set up something like -map 0:v -map [out]:a when saving to the m3u8 file, in which case the Output with label 'out' does not exist in any defined filter graph, or was already used elsewhere. message appears and does not work.

  


• How to convert mp3 data to wav data ?

  12 mai 2023, par Yali

  I have a wav audio file and i extracted data from that wav using python pydub module and i got this data

  


  [-139 18 -215 34 -196 6 -295 -31 -301 -35 -211 13 -93 47
-60 39 -58 7 -17 2]

  


  (this is first 10 data i got more than 1 million data)

  


  from pydub import AudioSegment
import numpy as np

song = AudioSegment.from_file("test.wav")
extract_data = np.array(song.get_array_of_samples())
print(extract_data[:10])


  


  then i converted wav to mp3 using that module and again extracted data from mp3 file and i got this data

  


  [-108 7 -193 24 -223 11 -239 -31 -248 -43 -203 -10 -101 23
-14 24 10 15 24 16]

  


  (this is first 10 data i got more than 1 million data)

  


  song = AudioSegment.from_file("test.wav")
song.export("test.mp3")
mp3_song = AudioSegment.from_file("test.mp3")
extract_data = np.array(mp3_song.get_array_of_samples())
print(extract_data[:10])


  


  and again i converted mp3 to wav now i got mp3 data instead of wav data.

  


  mp3_song = AudioSegment.from_file("test.mp3")
mp3_song.export("test1.wav", format="wav")

song = AudioSegment.from_file("test1.wav")
extract_data = np.array(song.get_array_of_samples())
print(extract_data[:10])


  


  My point is how to convert mp3 data to original wav data ?

  


  please help me,

  


  Thanks.

  


• CMD Batch Variable Won't Save FFprobe Output

  15 juillet 2017, par Matt McManis

  I have an CMD Batch Script that will convert a folder of mp4 videos to webm.

  You will need :

  • FFmpeg/FFprobe installed and set in Environment Variables to run from CMD.
  • A folder with an mp4 for FFprobe to parse.

  ---

  To make it easy, this is only the first part of the script, showing the Video Bitrate variable.

  Here is a full script, just replace the paths.
  https://pastebin.com/raw/3ng77Exz

  How the Script works :

  • Loops through all videos in folder
  • Has FFprobe parse the Video’s Bitrate and save it to %V and
    %vBitrate%.
  • Has FFmpeg use %V. Such as -b:v %V will become the parsed value
    -b:v 9401k.
  • Converts each video from mp4 to webm using the parsed Bitrate

  ---

  Problem

  I can’t get FFprobe’s Output to save to the variable. I’ve come up with a workaround, having it first save the bitrate value to a temp file, then import that to the %vBitrate% variable.

  Example : (%V > tmp_vBitrate) & SET /p vBitrate= < tmp_vBitrate.

  ---

  Works

  Temp File Variable

  cd "C:\Users\Matt\Videos\" && for %f in (*.mp4) do ffprobe -i "C:\Users\Matt\Desktop\Test\%~f" -select_streams v:0 -show_entries stream=bit_rate -v quiet -of csv="p=0" & for /f "tokens=*" %V in ("ffprobe -i "%~f" -select_streams v:0 -show_entries stream=bit_rate -v quiet -of csv=p=0") do (echo ) & (%V > tmp_vBitrate) & SET /p vBitrate= < tmp_vBitrate & del tmp_vBitrate & for /F %V in ('echo %vBitrate%') do (echo %V)

  ---

  Does Not Work

  Memory Variable

  cd "C:\Users\Matt\Videos\" && for %f in (*.mp4) do ffprobe -i "C:\Users\Matt\Desktop\Test\%~f" -select_streams v:0 -show_entries stream=bit_rate -v quiet -of csv="p=0" & for /f "tokens=*" %V in ("ffprobe -i "%~f" -select_streams v:0 -show_entries stream=bit_rate -v quiet -of csv=p=0") do (echo ) & SET vBitrate=%V & for /F %V in ('echo %vBitrate%') do (echo %V)

  ---

  Testing It

  Run the first command. When it is finished, type echo %vBitrate% in CMD and press Enter. You’ll see the bitrate of the last mp4 file parsed.

  Do the same for the second command and you’ll see it doesn’t work.

  ---

  Solution

  I would like to get rid of the Temp File Variable and get the second command to work.

  (%V > tmp_vBitrate) & SET /p vBitrate= < tmp_vBitrate to just SET vBitrate=%V.

  Maybe this whole thing can be simplified ? Am I using the variables wrong ?