Force left audio channel to both speakers

[encodingmonkey]

Hi, I hope this is in the right place.

I have a video file that only has audio coming out from the left channel. I was wondering if there was anything in ffmpeg that could force the sound form the left channel to both of the channels? Thanks!

[bat999]

I have a video file...

mediainfo?

[encodingmonkey]

ffmpeg -i '/media/INT03-STG/pappyland-01.mp4'
ffmpeg version 0.10.3 Copyright (c) 2000-2012 the FFmpeg developers
built on Jun 13 2012 15:57:20 with gcc 4.6.3
configuration: --enable-libass --enable-gpl --enable-libfaac --enable-libmp3lame --enable-libopencore-amrnb --enable-libopencore-amrwb --enable-libtheora --enable-libvorbis --enable-libvpx --enable-libx264 --enable-nonfree --enable-version3 --enable-x11grab
libavutil 51. 35.100 / 51. 35.100
libavcodec 53. 61.100 / 53. 61.100
libavformat 53. 32.100 / 53. 32.100
libavdevice 53. 4.100 / 53. 4.100
libavfilter 2. 61.100 / 2. 61.100
libswscale 2. 1.100 / 2. 1.100
libswresample 0. 6.100 / 0. 6.100
libpostproc 52. 0.100 / 52. 0.100
Input #0, mov,mp4,m4a,3gp,3g2,mj2, from '/media/INT03-STG/pappyland-01.mp4':
Metadata:
major_brand : mp42
minor_version : 0
compatible_brands: mp42mp41
creation_time : 2012-09-21 18:04:09
Duration: 00:28:25.81, start: 0.000000, bitrate: 20187 kb/s
Stream #0:0(eng): Video: h264 (Main) (avc1 / 0x31637661), yuv420p, 1280x720 [SAR 1:1 DAR 16:9], 19994 kb/s, 29.97 fps, 29.97 tbr, 29970 tbn, 59.94 tbc
Metadata:
creation_time : 2012-09-21 18:04:09
handler_name : Mainconcept MP4 Video Media Handler
Stream #0:1(eng): Audio: aac (mp4a / 0x6134706D), 48000 Hz, stereo, s16, 189 kb/s
Metadata:
creation_time : 2012-09-21 18:04:09
handler_name : Mainconcept MP4 Sound Media Handler
At least one output file must be specified

[bat999]

Try this:-

ffmpeg -i '/media/INT03-STG/pappyland-01.mp4' -vcodec copy -acodec libfaac -ab 220k -ar 48000 -ac 1 output.mp4

[Chris K]

You can use audio channel mapping. Add this to the command line given by bat999, just before the output.

-map_channel 0.1.0 -map_channel 0.1.0

This should map the input left channel to both output channels.

[sanlyn]

The sound will be muffled and "dead".

[encodingmonkey]

thanks, works like a charm!

[sanlyn]

Of course it did. But not "like a charm". More like duct tape. You have two mono signals in almost exact phase-cancellation relationship with each other. Lifeless audio, muffled midrange, anemic bass. There are other ways to do it, but they take longer. Anyway, enjoy.

[Cornucopia]

@sanlyn,
IF 2 signals had near-complete phase cancellation, they would exhibit the kind of symptoms you describe - WHEN COMBINED. But the OP only has 1 signal and is trying to distribute it to 2 channels, not combine it. If the distribution by ffmpeg works correctly, they will be duplicates and do exactly as they should. If ffmpeg works incorrectly with phase, it would likely be COMPLETE & EXACT phase opposites. Even then, unless you recombine the signals (which if you did in that case it would be total SILENCE), the only place it is probably "combining" is at your ears, where the sound would just be "spacey".

And, unless I'm misreading the ffmpeg nomenclature, I see nothing to suggest that those signals will have any phase differences.

Scott

[sanlyn]

The fact that "the phasing between the two isn't any different" is the problem.

I don't know how ffmpeg would work that operation, so I can't say with certainty that there is now a phasing problem. I had some similar videos a while back with only the right channel working in the DVD recorder. Transferred the video to MPEG2 on a computer, extracted the single-channel stereo to WAV and used Audacity to fix it. At first, all I did in Audacity was copy the right channel to the left channel. Totally dead, lifeless sound resulted, very muffled. Out of frustration I went back to the original stereo file and used Audacity to separate the two channels (i.e., Audacity was now looking at 2 separate mono channels, not a left/right stereo file, with the left mono channel unpopulated). Then I copied the right mono channel into the left mono channel. Finally I selected both channels and hit the "convert to stereo" switch. Apparently Audacity does a phase fix when the two mono's are converted to stereo. This brought dynamics back and you could even hear some remnants of the lost left channel. Of course the new output was "mono", but apparently Audacity worked a phase change on the conversion to "stereo".

I don't have any other explanation for the difference I heard. About that time I did find some hardware/software device that was reported to retrieve a lost stereo channel by analyzing phase relationships within the surviving channel and recreating stereo effects from that. But it cost well over $2K USD and required auxiliary hardware -- so that wouldn't be a solution I'd recommend.

[budwzr]

Any "re-creation" of a lost channel would be speculation. Like re-creating the other half of a pizza. What if it has anchovies on the missing half?

Also, it's quite common to find instruments on only one channel, especially in the older stuff, where they were trying to create an orchestra-like sound. Didn't the piccolos come in from the side? Hahaha.

And re-creation wouldn't work well if anything is panned over time.

The concept is interesting though. Extrapolation or Interpolation I believe is the term. I've always been intrigued by the possibilities of fractal geometry.

Sony has a new approach to sound visualization called Spectral Layers that looks interesting. Apparently it has visual spatial depth, or some newer way of picking through frequencies. I'm already invested in Ableton Live, and a bunch of VST plugs, but newcomers might enjoy it.

[arsesra]

I have a similar problem but would like to know if it's possible to do this losslessly without ffmpeg re-encoding the audio track, either by copying the left channel to the right, or by dropping the silent channel to leave a mono track.

I can't figure out how to do this in ffmpeg, and Audacity involves re-encoding. Any help would be much appreciated. Thanks!

[budwzr]

Decompress to PCM first, then do a "mono mix". Time fer turkey..gobble gobble.

[arsesra]

As the original audio stream's an aac from an mp4, the decompression would still be lossy though, no?

Enjoy your thanksgiving!

[manono]

As the original audio stream's an aac from an mp4, the decompression would still be lossy though, no?

No.

[Cornucopia]

De-compression is ALWAYS lossless (unless a interpolation process like resize or colorspace conversion is included).

So let's say,
original LPCM wav = 100%
AAC compression/encoding = 70%

so the AAC file has 70% of the quality of the original.

De-compression = 100%

The decompressed stream has 100% of it's source (the 70% AAC), so that's 100% * 70% = 70%. The process is lossless, but the result is still less quality because the source has ALREADY and IRREVOCABLY been compromised.

Scott