Suggestion for rapid_videocr: Image Format Support

[chuducnam]

Hi SWHL!

Firstly, rapid_videocr is fantastic – fast and effective!

Just wanted to share a common scenario many of us face. While Video Sub Finder (VSF) is useful, it can sometimes struggle with accurately capturing subtitles, especially if caption positions are inconsistent or there are motion graphics/fade effects. This often forces us into manual "frame-by-frame" timing using scripts (e.g., with MPV) to get correct timecoded images, which usually output as .jpg or .png.

The challenge is that rapid_videocr seems tightly coupled with VSF's output (often .jpeg). For those of us with limited coding skills, constantly trying to tweak the source code to switch between .jpeg (for VSF folders) and .jpg (for our manually captured images) is quite a pain point.

It would be a massive help if rapid_videocr could natively support common formats like .jpg and .png. This would make it much more flexible and user-friendly for these common workarounds.

No pressure at all, but it'd be a brilliant enhancement if you get a chance to look into it.

Thanks for building such a great tool!

Best,
chuducnam

[SWHL]

In rapid_videocr>=3.0.2, different image formats are already supported, including {". jpg", ". JPEG", ". png", ". gif", ". bmp", ". tif", ". webp"}. You can install the v3.0.2 and try again.

Enjoy it.