Blog de François Becker

Aller au contenu | Aller au menu | Aller à la recherche

mercredi 30 septembre 2015

Compress a MP4 video using ffmpeg

This is the two-pass command, using a 2048k video bitrate:

ffmpeg -y -i <input>.mp4 -c:v libx264 -preset medium -b:v 2048k -pass 1 -c:a libfdk_aac -b:a 128k -f mp4 /dev/null && ffmpeg -i <input>.mp4 -c:v libx264 -preset medium -b:v 2048k -pass 2 -c:a libfdk_aac -b:a 128k <output>.mp4

samedi 21 mars 2015

Automate Octave script execution on save on Mac OS X

Let's say the octave script you are working on is located in ~/Documents/script.m . On save, you want it to be automagically executed by octave, including persistent figures.

Let's create a watchdog script ~/Documents/

while true
  ATIME=`ls -al ~/Documents/script.m`
  if [[ "$ATIME" != "$LTIME" ]]
    echo "Running octave script"
    echo "script" > ~/Documents/OCTAVEPIPE
  sleep 1

Make it executable: chmod u+x ~/Documents/

Don't launch it yet. We are missing the OCTAVEPIPE pipe. Create it with mkpipe ~/Documents/OCTAVEPIPE

Launch the watchdog in a shell: ~/Documents/ or ./ if you are in the same directory.

In another shell, cd to /Documents/OCTAVEPIPE) | octave --interactive

Now edit and save your octave script, it shoud be executed after less than 1 second.

lundi 12 janvier 2015

Activer le protocole aptX sur Mac OS X

Activer le protocole aptX sur Mac OS X :

dimanche 17 août 2014

A Complementary Learning Methodology For Foreign Language Vocabulary

It can prove difficult to estimate its own vocabulary capacities in a foreign language. It can also be difficult to find which words are still missing to its own knowledge: there is a need for targeting the words which are the most useful in the foreign language.

One can find frequency analysis of text corpuses (or "corpora" for pedants) online for instance on the Wiktionary project, originating from the Project Guteberg which works on a collection of public domain books. The words are ordered by frequency, and one can browse the list in order to find the first words we don't know — yet. Once discovered, these words can be added to flashcard-type learning softwares — these flashcards could even be provided pre-built so that adding one is easier.

One can then evaluate its own vocabulary level to the number of most-used words known.

Instead of working on a student-class level, this personalized method looks like one of the best complementary method for brute-force vocabulary upgrade. Though for the classroom it can set student level targets.

jeudi 24 octobre 2013

De la startup

Quelques bonnes explications pour monter ou comprendre les startups.