Du screencast sous GNU/Linux partie 3

Une solution m’est apparue : ffmpeg, que j’avais dénigré jusqu’ici, est en réalité la meilleure façon de faire un screencast sous GNU/Linux. Qualité vidéo irréprochable, post-traitement (dont une conversion en mp4 HD pour Youtube…), enregistrement des sons et du micro en parallèle, et la cerise sur le gâteau : tout est en ligne de commande.

Je me suis servi du script screencaster disponible ici sous licence GPL en le mixant un peu à ma sauce. Il reste donc sous licence GPL, mais n’a été testé que sous Ubuntu et dérivées jusqu’à présent. À ce titre, il nécessite libavcodec-extra-52 , mais aussi ffmpeg du dépôt Medibuntu. Le reste des dépendances s’installe d’un simple sudo apt-get install xdpyinfo grep head sed pacat parec sox.

Quant à l’utilisation, il suffit de le lancer avec l’option -h pour comprendre assez rapidement. L’option -w fait apparaitre un curseur, il n’y a plus qu’à cliquer sur la fenêtre de votre choix pour lancer l’enregistrement. Terminez alors l’enregistrement en tapant Entrée. Il existe une petite astuce pour éviter d’avoir à revenir sur la console pour arrêter l’enregistrement, et donc devoir couper la vidéo en post-prod : on peut arrêter l’enregistrement en configurant un raccourci clavier et en utilisant xdotool : xdotool search --onlyvisible --classname "gnome-terminal" key "Return".

#!/bin/bash
#
# screencaster.sh - script to make screencasts on Linux 
#
# For information about using this script:
# http://www.guillaume.litaudon.fr/blog/
#
# This program is free software: you can redistribute it and/or modify it
# under the terms of the GNU General Public License version 3, as published
# by the Free Software Foundation.
#
# This program is distributed in the hope that it will be useful, but
# WITHOUT ANY WARRANTY; without even the implied warranties of
# MERCHANTABILITY, SATISFACTORY QUALITY, or FITNESS FOR A PARTICULAR
# PURPOSE.  See the GNU General Public License for more details.
#
# You should have received a copy of the GNU General Public License along
# with this program.  If not, see <http://www.gnu.org/licenses/>.

# list of programs we depend on
progs="xdpyinfo grep head sed ffmpeg pacat parec sox"

# check for programs we depend on
result=0
for prog in $progs
do
  type -p $prog > /dev/null
  if (( $? != 0 )); then
    echo "Error: Cannot find required program '$prog'"
    result=1
  fi
done
if (( $result != 0 )); then
  exit 1
fi

screenSize="$(xwininfo -root | grep 'geometry' | awk '{print $2;}')" # default if we cant detect it
screenOffset="0,0" # default to top-left corner
frameRate="24" # default frame rate
baseName="`date +%Y%m%d`-`date +%H%M%S`-screencast" # default base filename for capture

# attempt to detect the dimension of the screen for the default
dimensions=`xdpyinfo | grep 'dimensions:' | head -1 | \
  sed -e 's/^.* \([0-9]\+x[0-9]\+\) pixels.*$/\1/'`
if [[ "$dimensions" =~ [0-9]+x[0-9]+ ]]; then
  screenSize=$dimensions
fi

# collect command line settings
while getopts 'hs:o:t:r:p:w' param ; do
  case $param in
    s)
      screenSize="$OPTARG"
      ;;
    o)
      screenOffset="$OPTARG"
      ;;
    t)
      timeToRecord="$OPTARG"
      ;;
    w)
      INFO=$(xwininfo)
      WIN_GEO=$(echo $INFO | grep -oEe 'geometry [0-9]+x[0-9]+' | grep -oEe '[0-9]+x[0-9]+')
      WIN_XY=$(echo $INFO | grep -oEe 'Corners:\s+\+[0-9]+\+[0-9]+' | grep -oEe '[0-9]+\+[0-9]+' | sed -e 's/\+/,/' )
      screenOffset="$WIN_XY"
      screenSize="$WIN_GEO"
      ;;
    r)
      frameRate="$OPTARG"
      ;;
    *)
      echo ""
      echo "$0 - records screencast"
      echo ""
      echo "$0 [options] [base-filename]"
      echo ""
      echo "options:"
      echo "    -h            show brief help"
      echo "    -s <size>     screensize to record as <width>x<height>"
      echo "    -o <offset>   offset off recording area as <xoffset>,<yoffset>"
      echo "    -t <time>     time to record (in seconds)"
      echo "    -w        window to record"
      echo "    -r        framerate (FPS)"
      echo ""
      exit 0
      ;;
  esac
done

shift $(( $OPTIND - 1 ))

# determine basename of files
if [ -n "$1" ] ; then
  baseName="$1"
fi

echo ""
echo "Size = $screenSize"
echo "Offset = $screenOffset"
echo "Rate = $frameRate"
echo "Filename = $baseName"

# get ready to start recording
echo ""
if [ -n "$timeToRecord" ]; then
  echo "Preparing to capture for $timeToRecord seconds."
else
  echo "Preparing to capture."
  echo "Press ENTER when finished capturing."
fi
sleep 3
echo ""

# start playing silence to make sure there is always audio flowing
pacat /dev/zero &
pidSilence=$!

# starts recording video using x11grab to make mpeg2video
ffmpeg -f alsa -ac 2 -i pulse \
       -f x11grab -r "$frameRate" -s "$screenSize" -i :0.0+"$screenOffset" \
       -acodec pcm_s16le -vcodec libx264  -vpre lossless_ultrafast \
       -threads 0 -y "$baseName.mkv" &
pidVideo=$!

#ffmpeg -y -an \
#  -s "$screenSize" -r "$frameRate" -f x11grab -i :0.0+"$screenOffset" \
#  -s "$screenSize" -r "$frameRate" -aspect 4:3 -vcodec mpeg2video -sameq \
#  -f mpeg2video "$baseName.mpeg" &
#pidVideo=$!

# starts recording raw audio
parec -d alsa_output.pci-0000_00_1b.0.analog-stereo.monitor --format=s16le --rate=44100 --channels=2 "$baseName.raw" &
pidAudio=$!

echo ""
echo "Video recording started with process ID $pidVideo"
echo "Audio recording started with process ID $pidAudio"
echo ""

# wait for recording to be done, either by timer or user hitting enter
if [ -n "$timeToRecord" ]; then
  sleep "$timeToRecord"
else
  read nothing
fi

# stop recordings
echo ""
echo "Terminating recordings ..."
kill -15 $pidVideo $pidAudio 
kill -15 $pidSilence
wait

# filter and normalize the audio
echo "" 
echo "Filtering and normalizing sound ..." 
sox --norm -s -b 16 -L -r 44100 -c 2 -v 0.33 "$baseName.raw" "$baseName.wav"  highpass 65 lowpass 12k

# Récupère le son du microphone
echo ""
echo "Extract audio from mkv"
ffmpeg -i "$baseName.mkv" -ar 44100 -ac 2 "$baseName-2.wav"

# encode video and audio into mp4 file
echo "" 
echo "Encoding to final Youtube format..." 
sox "$baseName.wav" "$baseName-2.wav" -m -p gain | ffmpeg -f sox -i - -i "$baseName.mkv" -s 1280x720 -aspect 16:9 -vcodec libx264 -vpre fast -crf 18 -threads 0 -acodec libfaac -ab 320k -y "$baseName.mp4"

#ffmpeg -isync -i "$baseName.wav" -i "$baseName.mpeg" -acodec mp2 -ab 192k -vcodec copy "$baseName.avi"

# convert to ogg - to turn on uncomment next three lines
#echo""
#echo "convert to theora"
#ffmpeg2theora "$baseName.avi" -o "$baseName.ogv"

# convert avi to flv - to turn on uncomment next three lines
#echo""
#echo "convert to theora"
# ffmpeg -i "$baseNamee.avi" -ab 56 -ar 44100 -b 200 -r 15 -s 320x240 -f flv  "$baseNamee.flv"

echo ""
echo "DONE! Final media written in file $baseName.mp4"

echo ""
exit 0