Hello. I have 6 channel AC3 files ( from mpeg2 files). I'm trying to make them stereo because only first 2 channels have sound (other are empty).
Videos are too quiet when playing.
I'm searching and searching but I found only about down-mixing 5.1 to stereo
But I'd like to NOT down-mix , only delete 4 empty channels and keep 1 and 2 channel untouched (WITHOUT re-encoding)
Does anyone know any tool to accomplish this task?
+ Reply to Thread
Results 1 to 21 of 21
-
-
Tested
Code:ffmpeg -i your51.ac3 -af "pan=stereo|c0=FL|c1=FR" output.ac3
This was done in a few seconds, at 230x speed.
That's why I think more about copying than recoding. -
Sorry, but:
Code:ffmpeg -i your51.ac3 -af "pan=stereo|c0=FL|c1=FR" output.ac3
Also tried:
Code:ffmpeg -i VTS_01_1.ac3 -af "pan=stereo|c0=FL|c1=FR" -acodec copy output.ac3
Filtergraph 'pan=stereo|c0=FL|c1=FR' was defined for audio output stream 0:0 but
codec copy was selected.
Filtering and streamcopy cannot be used together.
So probably FFmpeg can't do that. -
But what are You doing: Load AC3 -> select channels -> encode to WAVs -> load WAVs as channels -> encode to AC3
I'd like just:
Load AC3 -> select channels -> Save AC3
Instead of:
Load AC3 -> select/mix -> encode AC3->Save AC3
(this is good if you have true 5.1 and want to mix to stereo)
All examples from here (https://trac.ffmpeg.org/wiki/AudioChannelManipulation)
Also doing re-encoding.
When I was googling I found that many people ask similar questions, but no working solutions
And for example I was googling about more popular format: MP3. You can trim it, join two files without encoding but why not separate/join channels?
i know that after 2-3 little encoding quality won't change. But imagine 10 or 20 re-encodings.
If you just trim file and join trim and join... 200 times, quality you will have the same quality, because you can do it without re-encoding.
But try to split channels, join, split join,.. After first two or three times you won't hear difference. But after 20?
Is it technically possible, but no software available, or impossible? -
Most mp3 files are internally encoded in "joint stereo". Internally the two channels are mixed down to a mono channel (L+R) and a difference channel (L-R), then compressed. The only way to reproduce the original left or right channel is to decompress those two channels and then reverse the down mix.
Code:left = (L+R) + (L-R) = L+L right = (L+R) - (L-R) = R+R
-
It is expected the overall bitrate will decrease when you discard channels.
You can separate channels without re-encoding https://forum.videohelp.com/threads/398507-Top10-Commands-for-Lossless-Video-manipulat...e)#post2592759 -
-
Pure channel mapping detected message with the provided command means there is no recompression with ffmpeg.
AC3 5.1 384kbps to 192kbps stereo is the expected and best possible result. -
-
Hey butterw, I have to disappoint you, it's still doing recompression.
I tested this because I was curious and first thing I noticed compared to original audio is that this has additional 256 audio samples added like it's always doing when using ac3 encoder. Second is that spectrogram display is not the same for these two. New file is a little worse so to speak so this only means ffmpeg is using its own ac3 encoder and encoding 2 channels. -
You are likely correct, it seems ffmpeg -af forces a (quick) re-encode.
It's possible to uncompress to wav or re-encode to lossless (flac). -
I don't know how ffmpeg chooses the bitrate when you don't specify one, but it seems to choose a CBR so it's possible it could appear to be halving the bitrate by throwing one of the channels away, but I'm pretty sure any compressor will take advantage of the commonality between channels to compress, even a lossless one. For VBR, dual mono and single channel audio should be close to the same bitrate.
The easiest way to confirm 2 channels are exactly the same in Audacity is to open the "stereo" audio, split it into two mono tracks, pan them both centre, highlight one and select Edit/Invert. If they were identical they'll cancel each other out and you'll hear nothing but silence. Anything you do hear is effectively the difference between them.Last edited by hello_hello; 7th Apr 2021 at 16:09.
Avisynth functions Resize8 Mod - Audio Speed/Meter/Wave - FixBlend.zip - Position.zip
Avisynth/VapourSynth functions CropResize - FrostyBorders - CPreview (Cropping Preview) -
192kbps seems to be the default bitrate for ffmpeg stereo AC3.
-
Thinking about it, for AC3 it's probably the same as Aften.
http://aften.sourceforge.net/longhelp.html
Code:[-b #] CBR bitrate in kbps CBR mode is selected by default. This option allows for setting the fixed bitrate. The default bitrate depends on the number of channels (not including LFE). 1 = 96 kbps 2 = 192 kbps 3 = 256 kbps 4 = 384 kbps 5 = 448 kbps
Avisynth functions Resize8 Mod - Audio Speed/Meter/Wave - FixBlend.zip - Position.zip
Avisynth/VapourSynth functions CropResize - FrostyBorders - CPreview (Cropping Preview) -
to reencode to ac3 256kbps with ffmpeg:
-acodec ac3 -b:a 256k "out.ac3" -
Similar Threads
-
Converting a stereo 116kbps Opus track to AC3; what bitrate should I use?
By broom441 in forum AudioReplies: 5Last Post: 7th Jan 2021, 23:09 -
Keep core AC3 track or encode TrueHD to E-AC3?
By slap_monster in forum AudioReplies: 1Last Post: 4th Dec 2020, 10:45 -
How to Convert 2 channel stereo wave into DD5.1 AC3 448 / AC 640
By Bonie81 in forum User guidesReplies: 36Last Post: 2nd Feb 2018, 12:42