vipsdisp VS WebPShop

if you're using linux (which it doesn't seem so) there's also vispdisp https://github.com/jcupitt/vipsdisp which is based on https://github.com/libvips/libvips which will likely take over how images are decoded in the future for everything, at least methodology wise.

Yes, I based it partly on visual programming system that Canon made back in the 90s. I loved the interactive development, but I really disliked that style of visual programming. Lines between boxes seems very hard to reason about, to me.

nip2 tries to do something similar, but using expressions, like a spreadsheet. Unlike graphical plugs and wires, expressions allow things like copy-paste, smart refactoring, referential transparency, and progressive rewriting.

nip2 is ancient now (I finished it way back in 2002, I think). I've started a rewrite for gtk4 -- the image display window is here:

https://github.com/jcupitt/vipsdisp

It's quite fancy -- it has a sparse pyramid of image tiles which get composited to your display by the GPU, and computed asynchronously by a libvips pipeline as you zoom and pan around. You should get a smooth 60 fps even with enormous (eg. 300,000 x 300,000 pixel) images.

Grab a the webp plugin for Photoshop: https://developers.google.com/speed/webp/docs/webpshop

Knock yourself out mate

Webp is a newer format and adoption hasn't reached universal levels yet, but there's a free (official) Photoshop plugin: https://developers.google.com/speed/webp/docs/webpshop. It compresses well and can store lossless data.

This fixes it

Try the free Squoosh.app from Google, you can see before and after at the same time. Ideally you'd use webp as the format, it's supported by most browser and WP since 5.8. They also have a webp plugin for Photoshop which uses the same algorithm, meaning that you can try with the before and after on their site and test different percentages, and then replicate or automate it in Photoshop.

most common would be to use Photoshop with Google's WebPShop plugin