I have a project with an asset library of large .jpg files that I’d like to compress into the .webp format.
Some have compressed fine, while others will not compress at all. Either in batches or individually.
The originals were all formatted the same. I used a ‘default’ 3200x2000 px file size and let Webflow handle them. I did this for a different website recently, and it compressed them all with no problems.
Does anyone know of a workaround or tips for best practices, before I re-import all the originals at a smaller size and try again?
It has compressed one file over 1.4 MB absolutely fine, but not some of the smaller ones. Odd.
If an asset fails to convert, it may already be compressed as much as possible. You won’t be able to convert fully compressed files.
source - Webflow University
I read that but maybe misinterpreted it. To me, that would mean that some of my .jpgs are already heavily compressed. When they aren’t, in this instance.
Or is it just that .webp can do no better than the current .jpg compression on some images? And therefore, I should leave them.
In my head, webp is some sort of supernatural format that shaves a ton of data from jpgs.
Personally, I don’t use Webflow’s batch conversion since it nukes the original file, which can lead to older browsers (like older iPhones) not displaying any images. Instead I use SQUOOSH or other batch tools to handle this myself for optimum performance and quality, then build backwards compatible display using the picture element and custom code. I prefer control.
You could test out various compression settings yourself with SQUOOSH in your browser. If it does a better job on a file, you could save that and upload it to assets. There is no real magic with WebP. In fact AVIF blows it away. When I use image CDNs instead of Webflow, I end up automatically serving AVIF if the browser supports it, then gracefully degrade from Webp to jpg or png.
SQUOOSH (web app from Google) note: You can download it locally, and you can even run batch processes.
I prefer other tools in my image processing chest for batch work, but this one is pretty good.
Thanks for info Jeff, I’ll check it out.