tutorials/vp9-conversion.md

+# Parameters
+
+| Parameter | Function |
+| --- | --- |
+| `$INPUT` | the input video file |
+| `$CRF` | the target quality (33 is a good value, the smaller the better the video quality) |
+| `$THREADS` | the number of threads to use converting the video |
+| `$GOP` | the number of frames after which another inter frame is forced. You can seek between inter frames, so if you want to seek make this something like 5 * `$FPS` |
+| `$OUTFILE` | the target video file name |
+| `$WIDTH` | the target width when scaling. When setting the height to `-1` the aspect ratio is kept |
+
+Other regulations:
+
+The webm container only supports webvtt subtitles, so if you have embedded subtitle in the video `-c:s webvtt` tries to convert them into webvtt.
+
+# Converting a single file
+
+With one-pass:
+
+```bash
+ffmpeg -i $INPUT -c:v libvpx-vp9 -b:v 0 -crf $CRF -threads $THREADS -g $GOP -tile-columns 6 -frame-parallel 1 -c:s webvtt -speed 1 -c:a libopus -b:a 64k -f webm $OUTFILE
+```
+
+When doing a two-pass, the speed value can be changed and the audio conversion can be omitted during 1st pass:
+
+```bash
+ffmpeg -i $INPUT -pass 1 -passlogfile $PREFIX -c:v libvpx-vp9 -b:v 0 -crf $CRF -threads $THREADS -g $GOP -tile-columns 6 -frame-parallel 1 -speed 4 -an -f webm /dev/null
+ffmpeg -i $INPUT -pass 2 -passlogfile $PREFIX -c:v libvpx-vp9 -b:v 0 -crf $CRF -threads $THREADS -g $GOP -tile-columns 6 -frame-parallel 1 -c:s webvtt -speed 1 -c:a libopus -b:a 64k -f webm $OUTFILE
+```
+
+When scaling the video just add the following parameter:
+
+```bash
+-vf scale=$WIDTH:-1
+```
+
+# Merging several files into one
+
+To merge several files into one a "complex filter" has to be used. When this is used, no normal filter can be used (e.g. for scaling). To simply merge two files use the following parameters:
+
+```bash
+-i $INPUT1 -i $INPUT2 -filter_complex "[0:v:0] [0:a:0] [1:v:0] [1:a:0] concat=n=2:v=1:a=1 [v] [a]" -map '[v]' -map '[a]'
+```
+
+This takes the first video and audio streams from the first file and the first video and audio streams from the second file. It then concats the two (`n=2`) streams in one video (`v=1`) and audio stream (`a=1`) and exports them as `[v]` and `[a]`. Later it maps them to the output file.
+
+When trying to scale this video you can chain complex filters. This would look like this:
+
+```bash
+-i $INPUT1 -i $INPUT2 -filter_complex "[0:v:0] [0:a:0] [1:v:0] [1:a:0] concat=n=2:v=1:a=1 [v] [a]; [v]scale=$WIDTH:-1[v2]" -map '[v2]' -map '[a]'
+```
+
+# Known Errors and Workarounds
+
+## Invalid channel layout 5.1(side) for specified mapping family -1
+
+Either add an audio filter
+
+```bash
+-af "channelmap=channel_layout=5.1"
+```
+
+or enhance the complex filter with
+
+```bash
+;[a] channelmap=channel_layout=5.1 [a2]
+```
\ No newline at end of file