Useful multimedia manipulation

lectures/useful-multimedia-manipulation/.gitattributes

@@ -0,0 +1,3 @@
+PYTHON[[:space:]]PRESENTATION[[:space:]]RVA[[:space:]]2020[[:space:]]JULY.key filter=lfs diff=lfs merge=lfs -text
+*ppts filter=lfs diff=lfs merge=lfs -text
+PYTHON[[:space:]]PRESENTATION[[:space:]]RVA[[:space:]]2020[[:space:]]JULY.pptx filter=lfs diff=lfs merge=lfs -text

lectures/useful-multimedia-manipulation/.gitignore

@@ -0,0 +1,4 @@
+figures
+abstract
+*.html
+mailmap

lectures/useful-multimedia-manipulation/README.rst

@@ -0,0 +1,29 @@
+Python for Quick, Useful Multimedia Manipulation: Anecdotes from a Python Programmer
+=====================================================================================
+
+One of Python's many killer features is that a programmer can create snippets of code, and organically and coherently join them into larger collections of module code. In this presentation, I describe and demonstrate relatively simple Python tools -- for image, video, and audio manipulation -- that I use every day at work and outside work.
+
+* automatically cropping out whitespace in images (surprisingly easy to do).
+
+* creating movies from a sequence of generated images.
+
+* conversion of movie clips, either files or YouTube clips, into animated GIFs (useful where the online service does not allow for video animations from movie files, such as GitHub_ or `Read the Docs`_).
+
+If there's time or interest, I can even describe and demonstrate how to retrieve and label music you might find, all within Python.
+
+Demonstrations live in the ``demonstrations`` subdirectory, and each demonstration is its own directory within ``demonstrations``, and each demonstration has its own ``README.rst``. Here are the four demonstration directories with description.
+
+1. ``autocropping_images``: autocropping a PNG image and a PDF image.
+
+2. ``movie_image_demos``: converting a sequence of images into an MP4 file.
+
+3. ``movie_gif_demos``: converting an MP4 file and a YouTube clip into animated GIFs.
+
+4. ``making_music_youtube``: using a tool, `plex_music_songs`_, that takes metadata from MusicBrainz_ and the YouTube clip using `youtube-dl`_, into an M4A file.
+
+.. _GitHub: https://github.com
+.. _`Read the Docs`: https://www.readthedocs.io
+.. _CloudConvert: https://cloudconvert.com
+.. _`plex_music_songs`: https://plexstuff.readthedocs.io/plex-music/cli_tools/plex_music_cli.html?highlight=plex_music_songs#plex-music-songs
+.. _MusicBrainz: https://musicbrainz.org
+.. _`youtube-dl`: https://rg3.github.io/youtube-dl

lectures/useful-multimedia-manipulation/demonstrations/autocropping_images/.gitignore

@@ -0,0 +1,2 @@
+iwanttobelieve_cropped.png
+cumulative_plot_emission_cropped.pdf

lectures/useful-multimedia-manipulation/demonstrations/autocropping_images/README.rst

@@ -0,0 +1,29 @@
+INSTRUCTIONS
+=============
+
+1. Go to these locations:
+
+   a. autocropping lossy images: `nprstuff.core.autocrop_image.autocrop_image <autocrop_image_>`_
+
+   b. autocropping PDFs: `nprstuff.core.autocrop_image.crop_pdf_singlepage <autocrop_image_pdf_>`_
+
+2. Cropping a PNG image:
+
+   .. code-block:: console
+
+      autoCropImage --input=iwanttobelieve_uncropped.png --output=iwanttobelieve_cropped.png
+
+   Compare ``iwanttobelieve_uncropped.png`` and ``iwanttobelieve_cropped.png`` in a browser.
+
+3. Cropping a PDF image:
+
+   .. code-block:: console
+
+      autoCropImage --input=cumulative_plot_emission_uncropped.pdf --output=cumulative_plot_emission_cropped.pdf
+
+   Compare ``cumulative_plot_emission_uncropped.pdf`` and ``cumulative_plot_emission_cropped.pdf`` in a PDF viewer.
+
+
+.. _`autocrop_image`: https://github.com/tanimislam/nprstuff/blob/f67e719ba4f2ca7120774937d27cb1adbb51c933/nprstuff/core/autocrop_image.py#L25
+
+.. _`autocrop_image_pdf`: https://github.com/tanimislam/nprstuff/blob/f67e719ba4f2ca7120774937d27cb1adbb51c933/nprstuff/core/autocrop_image.py#L193

lectures/useful-multimedia-manipulation/demonstrations/making_music_youtube/.gitignore

@@ -0,0 +1 @@
+*.m4a

lectures/useful-multimedia-manipulation/demonstrations/making_music_youtube/README.rst

@@ -0,0 +1,34 @@
+INSTRUCTIONS
+=============
+
+1. Go to these locations:
+
+   a. showing Musicbrainz_ getting metadata (given an artist, all the albums + all songs of artist): `plexstuff.plexmusic.plexmusic.get_music_metadata <get_music_metadata_>`_
+
+   b. showing choose the correct YouTube clip: `plexstuff.plexmusic.plexmusic.youtube_search <youtube_search_>`_
+
+   c. showing part of downloading from `youtube-dl`: `plexstuff.plexmusic.plexmusic.get_youtube_file <get_youtube_file_>`_
+
+2. ``plex_music_songs`` documentation at `this website`_.
+
+3. Show animation of getting songs, ``plex_music_songs_download_artist_songs.mp4``, with video viewer.
+
+4. Show how to get ``All I Need`` by ``Air``
+
+   .. code-block:: console
+
+      plex_music_songs -a Air -s "All I Need" --musicbrainz
+
+   Which can take a bit of time.
+
+.. _MusicBrainz: https://musicbrainz.org
+
+.. _`get_music_metadata`: https://github.com/tanimislam/plexstuff/blob/37cfb9f9e52864d8bdd6a2e154dc93b48ff2c908/plexstuff/plexmusic/plexmusic.py#L411
+
+.. _`youtube_search`: https://github.com/tanimislam/plexstuff/blob/37cfb9f9e52864d8bdd6a2e154dc93b48ff2c908/plexstuff/plexmusic/plexmusic.py#L888
+
+.. _`youtube-dl`: https://rg3.github.io/youtube-dl
+
+.. _`this website`: https://plexstuff.readthedocs.io/plex-music/cli_tools/plex_music_cli.html?highlight=plex_music_songs#plex-music-songs
+
+.. _`get_youtube_file`: https://github.com/tanimislam/plexstuff/blob/37cfb9f9e52864d8bdd6a2e154dc93b48ff2c908/plexstuff/plexmusic/plexmusic.py#L848

lectures/useful-multimedia-manipulation/demonstrations/making_music_youtube/plex_music_songs_download_artist_songs.mp4

@@ -0,0 +1 @@
+../../figures/plex_music_songs_download_artist_songs.mp4

lectures/useful-multimedia-manipulation/demonstrations/movie_gif_demos/.gitignore

@@ -0,0 +1 @@
+*.gif

lectures/useful-multimedia-manipulation/demonstrations/movie_gif_demos/README.rst

@@ -0,0 +1,37 @@
+INSTRUCTIONS
+=============
+
+1. Go to these locations in `nprstuff.core.convert_image <convert_image_>`_.
+
+   a. mp4togif: `nprstuff.core.convert_image.mp4togif <mp4togif_>`_
+
+   b. youtube2gif: `nprstuff.core.convert_image.youtube2gif <youtube2gif_>`_
+
+2. Lower level FFMPEG_ command comes from `this website`_.
+
+3. MP4 to animated GIF:
+
+   .. code-block:: console
+
+      convertImage movie -f covid19_conus_cases_04072020.mp4
+
+   Creates animated GIF ``covid19_conus_cases_04072020.gif``, open in browser.
+
+4. YouTube Clip to GIF:
+
+   .. code-block:: console
+
+      convertImage youtube -u "https://www.youtube.com/watch?v=R-pmYwr8zbU" -o "lucas_bros.gif" -q high
+
+   Creates animated GIF ``lucas_bros.gif``, open in browser.
+
+
+.. _`convert_image`: https://github.com/tanimislam/nprstuff/blob/master/nprstuff/core/convert_image.py
+
+.. _mp4togif: https://github.com/tanimislam/nprstuff/blob/807a3cba7e8bfd6ded70cdea3083cd9c9494e438/nprstuff/core/convert_image.py#L150
+
+.. _youtube2gif: https://github.com/tanimislam/nprstuff/blob/807a3cba7e8bfd6ded70cdea3083cd9c9494e438/nprstuff/core/convert_image.py#L135
+
+.. _`this website`: http://blog.pkh.me/p/21-high-quality-gif-with-ffmpeg.html
+
+.. _FFMPEG: https://ffmpeg.org

lectures/useful-multimedia-manipulation/demonstrations/movie_image_demos/.gitignore

@@ -0,0 +1 @@
+*.mp4

lectures/useful-multimedia-manipulation/demonstrations/movie_image_demos/README.rst

@@ -0,0 +1,26 @@
+INSTRUCTIONS
+=============
+
+1. Go to this gist: `demo_create_movie_sequence.py`_.
+
+2. Instructions on the FFMPEG_ syntax to make an MP4 movie from a sequence of images come from `this website`_.
+
+3. MP4 movie of MCMC demo:
+
+   .. code-block:: console
+
+      python3.7 demo_create_movie_sequence.py --prefix=img --output="mcmc_images.mp4" --dirname="mcmc_animation_images" --fps=10
+
+   Open ``mcmc_images.mp4`` with a video viewer. This is 10 frames per second.
+
+4. MP4 movie of cumulative COVID-19 cases in continental United States:
+
+   .. code-block:: console
+
+      python3.7 demo_create_movie_sequence.py --prefix="covid19_conus_cases_04072020." --output="covid19_conus_cases_04072020.mp4" --dirname="covid19_conus_cases_04072020_imagefiles" --fps=5
+
+   Open ``covid19_conus_cases_04072020.mp4`` with a video viewer. This is 5 frames per second.
+
+.. _`demo_create_movie_sequence.py`: https://gist.github.com/tanimislam/406a1379e746c9882c101f656a6da949
+.. _FFMPEG: https://ffmpeg.org
+.. _`this website`: https://hamelot.io/visualization/using-ffmpeg-to-convert-a-set-of-images-into-a-video/

lectures/useful-multimedia-manipulation/demonstrations/movie_image_demos/demo_create_movie_sequence.py

@@ -0,0 +1,89 @@
+#!/usr/bin/env python3
+
+"""
+I developed this Python CLI script to demonstrate creating an MP4 movie from an image sequence.
+I follow instructions from https://hamelot.io/visualization/using-ffmpeg-to-convert-a-set-of-images-into-a-video. I make an MP4 movie that is 5 FPS with psychovisual quality of 25.
+"""
+import os, sys, numpy, subprocess, glob, re, logging, time
+from PIL import Image
+from distutils.spawn import find_executable
+from argparse import ArgumentParser
+
+def create_movie_from_sequence( prefix, output_file_name, dirname = os.getcwd( ), fps = 5 ):
+    """
+    Creates an MP4 movie from a sequence of PNG images. The output file name must end in .mp4.
+
+    :param str prefix: the beginnning base name of all the PNG images. If `foo` is the prefix, then will look for images that are like `foo0001.png`, `foo0002.png` and so forth. To make life simpler, the prefix *MUST* be alphanumeric.
+    :param str output_file_name: the name of the output file. Must end in .mp4.
+    :param str dirname: optional argument. The directory that contains the image sequence. Default is the current working directory.
+    :param int fps: frames per second. Default is 5.
+    :returns: `True` if successful, `False` otherwise.
+    :rtype: bool
+    """
+    time0 = time.time( )
+    assert( os.path.isdir( dirname ) ) # is a directory
+    assert( os.path.basename( output_file_name ).endswith( '.mp4' ) ) # ends with mp4
+    assert( re.match( '^[a-zA-Z]+', prefix ) is not None ) # alphanumeric
+    assert( find_executable( 'ffmpeg' ) is not None )
+    assert( fps >= 1 ) # fps must be positive
+    ffmpeg_exec = find_executable( 'ffmpeg' )
+    #
+    ## now sequence of images
+    sorted_filenames = sorted(
+        filter(lambda fname: re.match('.*[0-9]+\.png', os.path.basename( fname ) ) is not None and
+               os.path.basename( fname ).startswith( prefix ),
+               glob.glob( os.path.join( dirname,'%s*.png' % prefix ) ) ) )
+    if len( sorted_filenames ) == 0:
+        print( 'ERROR, COULD FIND NO IMAGE SEQUENCE.' )
+        return False
+    def is_divis_2( fname ):
+        img = Image.open( fname )
+        if img.size[0] % 2 != 0: return False
+        if img.size[1] % 2 != 0: return False
+        return True
+    try:
+        assert(all(filter(is_divis_2, sorted_filenames))) # all widths and heights div by 2
+    except:
+        print( "ERROR, NOT ALL IMAGES HAVE WIDTHS + HEIGHTS DIVISIBLE BY 2.")
+        return False
+
+    #
+    ## 
+    num_base_10 = 1 + int( numpy.log10(len(sorted_filenames)))
+    sequence_ffmpeg = '%%%02dd' % num_base_10 # this is tricky!
+    input_ffmpeg_image_string = '%s%s.png' % ( os.path.join( dirname, prefix ), sequence_ffmpeg )
+    logging.info( 'got here, %s.' % input_ffmpeg_image_string )
+    #
+    ## now run the ffmpeg command
+    command_to_process =  [
+        ffmpeg_exec, '-y', '-r', '%d' % fps, '-f', 'image2', '-i', input_ffmpeg_image_string,
+        '-vcodec', 'libx264', '-crf', '25', '-pix_fmt', 'yuv420p',
+        output_file_name ]
+    logging.info( 'COMMAND TO RUN: %s.' % ' '.join( command_to_process ) )
+    proc = subprocess.Popen( command_to_process, stdout = subprocess.PIPE,
+                            stderr = subprocess.STDOUT )
+    stdout_val, stderr_val = proc.communicate( )
+    logging.info( 'STDOUT_MESSAGE.' )
+    logging.info( '%s\n' % stdout_val )
+    logging.info( 'TOOK %0.3f SECONDS TO RUN TO COMPLETION.' % (
+        time.time( ) - time0 ) )
+
+if __name__=='__main__':
+    parser = ArgumentParser( )
+    parser.add_argument( '--prefix', dest='prefix', type=str, required = True,
+                        help = 'The prefix to the sequence of PNG images.' )
+    parser.add_argument( '--output', dest='output', type=str, required = True,
+                        help = 'The name of the MP4 output file.' )
+    parser.add_argument( '--dirname', dest='dirname', type=str, default = os.getcwd( ),
+                        help = 'The directory containing the image sequence. Default is CWD.')
+    parser.add_argument( '--fps', dest='fps', type=int, default = 5,
+                        help = 'Frames per second of movie. Default is 5.' ) 
+    parser.add_argument( '--info', dest='do_info', action='store_true', default = False,
+                        help = 'If chosen, then print out INFO debug logging.' )
+    args = parser.parse_args( )
+    logger = logging.getLogger( )
+    #
+    if args.do_info: logger.setLevel( logging.INFO )
+    status = create_movie_from_sequence(
+        args.prefix, args.output, dirname = args.dirname, fps = args.fps )
+