shrivel VS oxipng

I was trying to create PDF files from photos taken with a phone camera (aka docscanner) and tried to compress as good as possible, when I realized that PNG supports 1 bit mode, with a palette of only 2 entries...

Combined with adaptive thresholding (e.g. Sauvola) this resulted in extremely compressed but still readable images.

I wrote a little C# tool called shrivel for my requirements: https://github.com/sandreas/shrivel/blob/main/shrivel/Comman...

I have had good experiences with https://github.com/shssoichiro/oxipng . Although, I suspect this wouldn't give you nearly enough space savings as jpg.

If you do want the file as a PNG (for transparency and a common format that's well supported), but don't want it so huge, consider something like oxipng. https://github.com/shssoichiro/oxipng

oxipng, pngquant and svgcleaner — optimizing images

I wonder how `pngcrush` compares to `oxipng` (https://github.com/shssoichiro/oxipng).

Personally, I use `oxipng` if I want lossless compression. However, most of the time, I use `pngquant` instead, since it gives significant size reduction even at `99%` (I can't even distinguish between the original and reduced image).

    pngquant --quality=99 --ext=.png --force file.png

Depending on your workflow it might make sense to export PNGs directly from Affinity and then reduce their size with a utility like Oxipng, which uses all your cores to find the best algorithm for each particular image.

Assuming your frames are PNG files, you could use a lossless optimizer like optipng to try if their size can be reduced. I prefer oxipng, which is faster and multithreaded, and seems to have more active development.

Not sure what extensions of images you use, but if they’re PNG you could use oxipng: https://github.com/shssoichiro/oxipng