Sorry, I was under the impression that simply extracting and using the original images as they are doesn’t count as data mining.I thought data mining refers to when you combine multiple extracted images (like different expressions) to create something new...
Please do not use screenshots as a way to extract anything from a video, you can get the frames losslessly and directly from the mkv using ffmpeg or VLC.
Since they said they downloaded it, I assumed they meant they took a screenshot from the media player itself, not the screen. However, there is one pixel missing from the vertical dimension, so this isn't an ideal crop.
Also, I fixed the source. The root path shouldn't be the source in this case since you can navigate back to it easily even with a more direct URL. Users should not have to go through a bunch of folders to find the file if it can be avoided.
I couldn't see any significant visual differences, but https://duplicatebooru.zipfiled.info/ indicates that this is a webp with 100 quality while the Twitter version is a jpg with 92 quality. The webp assets are consistently higher-quality than the jpgs found on Twitter or elsewhere, and I assume it's likely the case for these as well even though I can't tell visually.
This was obtained by downloading the MKV file of the Chaos;Child anime menu volume 0 and screenshotting it. If anyone knows a better way to crop it and turn it into a PNG, they can find the file at https://drive.sci-adv.cc
Please do not use screenshots as a way to extract anything from a video, you can get the frames losslessly and directly from the mkv using ffmpeg or VLC.
