--
You received this message because you are subscribed to the Google Groups "WebP Discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to webp-discuss...@webmproject.org.
To post to this group, send email to webp-d...@webmproject.org.
Visit this group at http://groups.google.com/a/webmproject.org/group/webp-discuss/.
For more options, visit https://groups.google.com/a/webmproject.org/d/optout.
If so, what is lossless WEBP for? Is it mainly for converting RAW or DNG images?
Hi team,I've been trying to convert JPEG photos to WEBP for a website. cwebp doubles the size of the files (typically from ≈ 4MB to ≈ 8 MB). Digging into this, I've learned that converting from one lossy format to another often increases the file size.
Since converting JPEGs seems to be a canonical use case, is there a guide or tips for the right settings? I initially used q 100 lossless. The cwebp guide says 100 will result in maximum compression for lossless.
Taking it down to lossy q 90 still results in massive file size increase.
If the answer is that the output webp quality setting has to equal the jpeg's quality setting, or be less than the jpeg's quality setting, then I have some subsidiary questions: 1) How do I determine the JPEG's quality setting?
2) Are these quality settings commensurable across JPEG and WEBP? (Is this a universal construct in image formats?)
3) Does matching the quality setting result in new loss of image data? e.g. 75 JPEG to 75 WEBP. I would've assumed that WEBP is just more efficiently packing the JPEG's data.
If it incurs another iteration of image data loss, then I'd have to weigh whether it's appropriate to use lossy webp, given archival concerns. If so, what is lossless WEBP for? Is it mainly for converting RAW or DNG images?
Hi Pascal,Thanks for your answers! On the issue of q 100, here is the text from your webp documentation page:
"In case of lossless compression (specified by the -lossless option), a small factor enables faster compression speed, but produces a larger file. Maximum compression is achieved by using a value of 100."It says 100 achieves maximum compression, not quality. I assume quality is held constant for lossless. Is the page in error? It's here: https://developers.google.com/speed/webp/docs/cwebp
Sorry for replying to a quite old question, but I am in the same situation as Joe was, I also tried to search around for this information with no luck, so my question is does converting JPEG to WEBP using lossy compression and with "jpeg_like" switch (to match the same quality of the JPEG compression) will guarantee there is no further loss in the image data?
Because generally converting from one lossy format to another lossy one will have image data loss during the compression, does this apply to JPEF to WEBP as well? Or it's like what Joe said: "I would've assumed that WEBP is just more efficiently packing the JPEG's data"On Tuesday, August 18, 2015 at 1:18:12 AM UTC+2 pascal.m...@gmail.com wrote:Hi Joe,On Thu, Aug 6, 2015 at 9:21 PM, Joe Duarte <songof...@gmail.com> wrote:Hi team,I've been trying to convert JPEG photos to WEBP for a website. cwebp doubles the size of the files (typically from ≈ 4MB to ≈ 8 MB). Digging into this, I've learned that converting from one lossy format to another often increases the file size.That's true for WebP but also JPEG: using q=100 is pretty much guarantied to make the file large for most input. You'd pretty much be using more bits to describe the same input information.Since converting JPEGs seems to be a canonical use case, is there a guide or tips for the right settings? I initially used q 100 lossless. The cwebp guide says 100 will result in maximum compression for lossless.I think you meant: maximum quality, not compression.Taking it down to lossy q 90 still results in massive file size increase.Yes, JPEG files are using encoded with quality q~=80, depending on the source (sometimes more, sometimes less).If the answer is that the output webp quality setting has to equal the jpeg's quality setting, or be less than the jpeg's quality setting, then I have some subsidiary questions: 1) How do I determine the JPEG's quality setting?Some jpeg-sniffer programs can do that. For instance, ImageMagick's 'identify' gives a lot of information:identify -verbose test.jpg | grep Quality(note: -verbose must come first. And it's "Quality", with a capital Q).2) Are these quality settings commensurable across JPEG and WEBP? (Is this a universal construct in image formats?)Not necessarily. But a '-jpeg_like' option was added to cwebp to try to make the two quality scales match in size (WebP being visually less distorted *if* the sourceis uncompressed enough).3) Does matching the quality setting result in new loss of image data? e.g. 75 JPEG to 75 WEBP. I would've assumed that WEBP is just more efficiently packing the JPEG's data.Yes indeed. Using the same -q value will typically results is a smaller file size for WebP, except maybe for very high quality (95+), where WebP will struggle a lot to reproduce some JPEG-friendly coefficient values.If it incurs another iteration of image data loss, then I'd have to weigh whether it's appropriate to use lossy webp, given archival concerns. If so, what is lossless WEBP for? Is it mainly for converting RAW or DNG images?Hope it helps,skal/
--
You received this message because you are subscribed to the Google Groups "WebP Discussion" group.
To unsubscribe from this group and stop receiving emails from it, send an email to webp-discuss...@webmproject.org.
To view this discussion on the web visit https://groups.google.com/a/webmproject.org/d/msgid/webp-discuss/561a1f5a-f94a-4ef7-a2a7-c3aeff6147bfn%40webmproject.org.