I think that has something to do with code though, and it's hard for us to change things here on that level. This site might not even be capable of doing that.
Moved from post #315296:

Jennifer003b said:
The weird thing is...WebP is a free format and it's right now (not tomorrow, not in a year) supported by ALL major browsers (see https://www.keycdn.com/support/webp-browser-support ) so the only "excuse" not to use it it's because someone wants to pay more for his server's bandwidth (or doesn't want to bother to update the site) which is perfectly fine by me I don't pay my internet connection based on how much I use it so...
If it's not supported already, it probably won't be. The owner doesn't have anyone they allow to work on the site code. We use a fork of Moebooru and they haven't been willing to make it public.

And as I mentioned, using a webp as wallpaper without recompression might not be possible either.

I don't think pixiv supports webp either. If they do, it's certainly not common to find any there. So it's not like we're forcing contributors to convert any significant number of images just to upload here.

There's certainly no problem with supporting it. It's just not likely to happen.
Zolxys said:
mattiasc02 said:
It's better if you press "use quality settings from original image"
No. If you're cropping or rotating, that's not true. That is unless you're just cropping a multiple of the block size (usually 16 pixels).

Otherwise you'll just lower the quality even further.
Not true. I have an example:

Let's take this image for example.

The original quality of this image tooken directly from Twitter is at 85 (as most images compressed by twitter). I check the jpeg number using IMGonline. Now, I got that image, cropped and rotated, and have both versions:

cropped:
exporting it at 85 and exporting it at 100

rotated:
exporting it at 85 and exporting it at 100

There is no difference in quality within the image itself. However, there IS a difference in file size and the jpeg number. But that don't mean jack shit, is my point.

Don't get me wrong, however. I also do understand that sometimes, the artist/handler of the image can be a fucking trickster (like with this. Number is at 100 but it got a shit ton of artifacts anyways).

This is why it's best to just export at original settings, since it keeps the file size the same as the original, and also exporting it at a higher jpeg does nothing but up the file size and that's problematic.

Sorry for late response. I've been thinking about responding, but not executing cuz my life is SOOOO emotional and stressful right now </3 I feel like an abandoned princess peach
mattiasc02 said:
Now, I got that image, cropped and rotated, and have both versions:

cropped:
exporting it at 85 and exporting it at 100

There is no difference in quality within the image itself. However, there IS a difference in file size and the jpeg number. But that don't mean jack shit, is my point.
Here are this differences highlighted. Mattias, you need to get a better monitor or something, the differences in quality are obvious.

Edit: that being said I think even the "original" from twitter has an unacceptable level of artifacting.
traz64 said:
Here are this differences highlighted. Mattias, you need to get a better monitor or something, the differences in quality are obvious.
What's that?

But not when you zoom in. Both have the same amount of artifacts in it. At least when I use my naked eye.

edit:
traz64 said:
Edit: that being said I think even the "original" from twitter has an unacceptable level of artifacting.
I know that. I was just using that as an example.
Sample from 85
Sample from 100

Switch back and forth between the two. You should see that the artifacts are far more prominent in the 85.
Oh yeah... it DID