"the pdf-to-anim series" (the drawing-to-animation pipeline)
i made an animation compositing system
around october 2025;
with imagemagick, potrace, and ffmpeg
that required only drawing on paper
and completely side-stepping animation software,
at least for drawing.
for each page of the pdf:
raster ← page at 150 dpi # one image per drawing
one_bit ← raster, 2 colours # black lines, white paper
svg ← trace(one_bit) # the white simply stops existing
png ← render(svg) # transparent where the paper was
repeat holds[page] times: # the csv is the timing
frames.append(png)
encode(frames, 24 fps) # mp4 to watch, gif, prores 4444 to composite
you can add a pdf scan of a drawing,
then a set of frame holds in a simple one-column-no-header csv,
and even a background for some of them;
you can export as a video, or as separate drawings
the list is as such;
| images | |
|---|---|
| ./pdf-svg-png-white.sh [input].pdf | processed raster drawing, white background |
| ./pdf-svg-png.sh [input].pdf | processed raster drawing, transparent |
| ./pdf-to-svg.sh [input].pdf | processed vector graphic drawing, transparent |
| videos | |
| ./pdf-to-mp4.sh [input].pdf [holds].csv | mp4 video export solo, white background |
| ./pdf-to-mp4-w-gif.sh [input].pdf [holds].csv | mp4 video export, white background, w/ gif export |
| ./pdf-to-alpha_mov.sh [input].pdf [holds].csv | alpha-channel (transparent) video export |
| ./pdf-to-superimpos.sh [input].pdf [holds].csv [background].{jpg, png} | mp4 video export with a background of your choice |
you can also run this engine here in the browser,
with nothing installed and nothing kept on the server:
→ pdf-to-anim, in the browser
as a companion to the pdf-to-anim series,
i made the program "spotcol"
to do colour fills on drawings that have been run
through the pdf-to-anim system
a program made with raylib
where you choose the colour and fill it in where needed
soon i'll even make a neural network for it with pufferlib
to automate colour filling;
hopefully it can be trained to change colours according to a given drawing
you can provide the drawings made from the image outputs
and then collate them together into a video when done
the repos are public here on gitlab → pdf-to-anim, spotcol
if you have any questions about it,
message me on x @ tebillusassort