Compress images: "FAILED (couldn't decode image)"

For some of my images, the Compress Images test reports that it couldn’t decode:
“FAILED (couldn’t decode image) - http://s3.amazonaws.com/cantoni.org.blog/images/iphone.jpg

Does anyone know the cause of that message? The images exist at those links and have already been compressed with SmushIt.

Test results:
http://www.webpagetest.org/result/101025_97KA/1/performance_optimization/#compress_images

thanks,
Brian

I’ll take a look and see if I can figure out what is going on. Pagetest uses libjpeg to decompress the images and it’s possible there’s something specific to those images that it doesn’t like.

FWIW, SmushIt only does lossless compression of jpeg’s (stripping out meta data, etc). Where you’ll usually see benefit in jpegs is to compress them at a lower quality level (without sacrificing the image visually). We use “Photoshop save-for-web quality 50” as the baseline but you can usually do better than that (and testing on the iPhone image it dropped from 5.2KB to 4.5KB).

Thanks,

-Pat

Looks like the decoding problem is pretty widespread and is probably linked to the upgrade to Page Speed 1.9 (which swapped out the image libraries that Pagetest uses). I’m looking into it now and should have it fixed before the end of the day.

Thanks,

-Pat

Thanks for investigating that, and got the point about JPG quality – I’ve been focused on SmushIt and haven’t though to reduce the original quality setting.

ok, I just fixed the image decoding problem (fix one bug, introduce another - sigh). The fix should be deployed now so everything should be back to normal, sorry for the inconvenience.

-Pat

Confirmed it’s working fine now - thanks for the quick fix!