As far as I've tracked this down, when uplading an image XF first checks the constraints (eg. max allowed dimensions/filesize), rotates the image if necessary and saves the (probably modified) image.
Afterwards, it does generate the thumbnail from this image.
@Chris D
Is there any particular reason (apart from processing resources?) for doing it this way instead of generating the thumbnail straight from the original?
Doesn't this approach negatively affect thumbnail quality (the "full size" image that is used to generate the thumbnail might have already been resized and/or rotated and saved as JEPEG@80%)?
Afterwards, it does generate the thumbnail from this image.
@Chris D
Is there any particular reason (apart from processing resources?) for doing it this way instead of generating the thumbnail straight from the original?
Doesn't this approach negatively affect thumbnail quality (the "full size" image that is used to generate the thumbnail might have already been resized and/or rotated and saved as JEPEG@80%)?
