5

In this excellent answer I learned how to extract a series of individual frames from a movie clip node.

This solution suits me well - there is a python script, and a few nodes, and it runs automatically.

Is there a way I could extract the audio from a clip - either the whole thing, or possibly just between two frame number limits, using a similarly simple technique in Blender? For me simple is python and nodes - I am not an advanced user, and something I can automate and customize is the most useful in my case.

I've have tried to examine the attributes of the vidNode but didn't find anything that looked like audio:

S = bpy.context.scene
n = S.node_tree

vidNode = n.nodes['Movie Clip']

enter image description here

Community
  • 1
uhoh
  • 2,667
  • 2
  • 27
  • 56
  • 1
    What about ffmpeg: http://stackoverflow.com/a/27413824/3091066 ? – p2or Sep 03 '16 at 09:45
  • 2
    @poor Thanks! I'm sure there are several ways to do this, but I am looking for something that runs in a python environment - preferably either my anaconda installation or preferably in Blender. The more ways I learn to use capabilities of Blender, the more I can learn about Blender. If there is a way I can do this using Blender, I would like to understand how. Even if it's not the "best" way to do it, if it can be done here I'd really like to do it here. – uhoh Sep 03 '16 at 15:49
  • 1
    In my opinion your best bet is to use VSE. Even if I don't know its python capabilities I think it's the only part of Blender that can do audio things (except Game Engine). – Polosson Sep 12 '16 at 07:33

1 Answers1

3

Currently there isn't a node with that capability built in. The simplest that you might script is the following:

  • add the clip to the VSE
  • set start and endframe: bpy.context.scene.frame_start and .frame_end
  • call the mixdown operator: bpy.ops.sound.mixdown (see the docs for parameters)
  • reset frame start/end if you need to.

something like this.

import bpy

def main(filepath_source, filepath_destination, frame_start=0, frame_end=20):
    scn = bpy.context.scene

    scn.frame_start = frame_start
    scn.frame_end = frame_end

    if not scn.sequence_editor:
        scn.sequence_editor_create()

    sequences = scn.sequence_editor.sequences
    sequences.new_sound(
        name='drone_detroit', 
        filepath=filepath_source,
        channel=0, frame_start=0
    )
    bpy.ops.sound.mixdown(filepath=filepath_destination)   

fp_source = r"C:\Users\zeffi\Downloads\DRONE\0001-2964.avi"
fp_destination = r"C:\Users\zeffi\OneDrive\Documents\chopper_drone4.flac"

main(fp_source, fp_destination, frame_start=500, frame_end=560)

the sequences.new_sound, lifts the audio track off the .avi.

The mixdown operator will output the audio of whatever is between those two frames.

zeffii
  • 39,634
  • 9
  • 103
  • 186
  • zeffii to the rescue - OK that's great. This will drag me kicking and screaming into the world of the video sequence editor - which is a good thing. There's a typhoon coming so I'll be stuck indoors in need of a project, perfect timing! – uhoh Sep 12 '16 at 09:57