I would rather encourage you to help making the PNG format smaller instead of trying to establish another new web image format. The world does not need it.
Love it. Thought Jpeg2k was not a good way for the web, but jpeg is old, very old in ways that there has to be something better.., and WebP with Alpha Channel is a Ruler. But how is the Performance need in comparing to jpeg and png for render in browser or encode/decode.
I think will be a good idea to add a header with an index and embed several "versions" at different resolutions of the same image (with offsets and lengths). So from a small device (smartphone) people could access to a small version of the image (without needing of downloading the whole file), and a normal computer could skip small versions and get the biggest resolution. That will greatly speed up internet navigation from mobile devices and will reduce traffic data.
Wonderful article,thanks for putting this together! "This is obviously one great post. Thanks for the valuable information and insights you have so provided here. Keep it up!" How to seduce a woman
It would be nice to compare the same image with different compression levels, in JPEG and WebP. I mean, same image at 100K, at 50K, at 25K. Also do not use "Images in the left column are JPEG originals", use TIFF or PNG originals.
Will it have a progressive mode? For small screens on mobile devices sometimes its enough not to have the full resolution. A mobile client could decide its got enough of a picture when a certain resolution is reached.
Using scaled-down images for the sample gallery does not provide a useful comparison. If you want to limit the size of the web page, show crops from the full images, but show them at full scale.
I'm interested to see this file format used and to see how it compares. Will it perserve transparencies? Is this supposed to be a replacement for all web images?
And why would JPEG "originals" show DCT artifacts that are not present in the WebP examples? Were they both compressed from a lossless original, or is there a mistake in your presentation of the examples?
Really trustworthy blog. sesli Please keep updating with great posts like this one. sesli sohbet I have booked marked your site and am about to email it
to a few friends of mine that I know would enjoy reading.. sesli chat
tux - I'm afraid you are actually an immense retard and are completely wrong. The compression schemes used by JPEG and PNG are aimed at entirely different goals and therefore neither makes the other obsolete. Your retarded simplification of the differences doesn't demonstrate an understanding of their relative purposes. I'll outline it in layman's terms for your simple mind. JPEG is suited for photographic images where pixels don't fall into uniform blocks allowing lossy compression to have a minimal impact on perceived loss of quality. PNG is suited for vector web graphics and images requiring hard outlines and smooth gradients. It is a lossless format but is compressed using the deflate algorithm commonly employed by web servers among other things. In short, it would appear that you've used Photoshop a little, compared a few file sizes and naively think you understand the finer points of complex bitmap formats. Newsflash: you don't - get a clue before you expose your incredible stupidity next time.
A Stéréoscopic integration with only diff of pictures(compressed) soon ? And to WebM too ? That will be so cool ! I'd like deposed a patent for this...
@Craig I'm afraid you're an immense jerkass-wad, and that you wouldn't be able to say what you said out face to face without a computer. Newsflash: using 'newsflash' doesn't add any emphasis to your conclusion.
If you are going to make a web specific image format, why not embed a metadata tag containing a URL to a better (resolute/quality) image. You could create multi-scale images as a chain of multiple files (whereby the lowest quality image is selected by default or if some CSS setting suggests it as default). Then the http client can recursively request better images in the background and refine the page dynamically. I'm sure that this is not a novel idea.
Should add XMP or EXIF support for image metadata. Easy enough for anyone to do, since you're using RIFF, but it should be standardized, and for now, you guys control the standard.
I'd like to see: - support for XMP metadata - psychovisual optimizations in the WebM encoder - support for lossless encoding - using both WebM inter and intra coding for images (as in the hipix image format)
If Google new invention success then we may be preparing a farewell for JPEG, PNG, and GIF formats.
ReplyDeleteI would rather encourage you to help making the PNG format smaller instead of trying to establish another new web image format. The world does not need it.
ReplyDeletePNG has a totally different goal than this, so usurping PNG means nothing.
ReplyDeleteThis is JPEGs goal, so usurping JPEG would be desirable. Especially with a functional lossy alpha layer.
The advantage of JPEG to PNG is the file size. If you want small files, use JPEG. If you want transparency, accept a few bytes more. So what?
ReplyDeleteThank you WebP Team. This is very appropriate. Definitely add transparency.
ReplyDeleteLove it. Thought Jpeg2k was not a good way for the web, but jpeg is
ReplyDeleteold, very old in ways that there has to be something better.., and
WebP with Alpha Channel is a Ruler. But how is the Performance need in
comparing to jpeg and png for render in browser or encode/decode.
Will it support animation? Would be great to finally have an alternative to the antique GIF for animated images.
ReplyDeleteI think will be a good idea to add a header with an index and embed several "versions" at different resolutions of the same image (with offsets and lengths).
ReplyDeleteSo from a small device (smartphone) people could access to a small version of the image (without needing of downloading the whole file), and a normal computer could skip small versions and get the biggest resolution. That will greatly speed up internet navigation from mobile devices and will reduce traffic data.
Wonderful article,thanks for putting this together! "This is obviously one great post. Thanks for the valuable information and insights you have so provided here. Keep it up!"
ReplyDeleteHow to seduce a woman
It would be nice to compare the same image with different compression levels, in JPEG and WebP. I mean, same image at 100K, at 50K, at 25K.
ReplyDeleteAlso do not use "Images in the left column are JPEG originals", use TIFF or PNG originals.
Will it have a progressive mode? For small screens on mobile devices sometimes its enough not to have the full resolution. A mobile client could decide its got enough of a picture when a certain resolution is reached.
ReplyDeleteA good idea, but practically, you're talking about a decade before the older browsers will have reduced by enough to enable this to be usable.
ReplyDeleteI apreciate this idea, we need innovation rether than satisfying ourself with existing technology.
ReplyDeleteUsing scaled-down images for the sample gallery does not provide a useful comparison. If you want to limit the size of the web page, show crops from the full images, but show them at full scale.
ReplyDeleteI'm interested to see this file format used and to see how it compares. Will it perserve transparencies? Is this supposed to be a replacement for all web images?
ReplyDeletesounds a lot like .art files to me.
ReplyDeleteWhy do most of the WebP images show a significant color shift compared to the original images? Shouldn't that be considered a major bug?
ReplyDeleteAnd why would JPEG "originals" show DCT artifacts that are not present in the WebP examples? Were they both compressed from a lossless original, or is there a mistake in your presentation of the examples?
ReplyDeleteAm I seeing that certain details actually seem to crisp-up in the WebP version? If so, then that kind of cool.
ReplyDeleteHow is WebP with black & white?
I've got a site which is graphics-heavy with a stack of black and white .GIFs
Any chance it would be worth converting GIFs over to WebP?
Oh, and what kind of crunch power is required for this format? Is this going to drag on older machines?
Really trustworthy blog. sesli Please keep updating with great posts like this one. sesli sohbet I have booked marked your site and am about to email it
ReplyDeleteto a few friends of mine that I know would enjoy reading.. sesli chat
tux - I'm afraid you are actually an immense retard and are completely wrong. The compression schemes used by JPEG and PNG are aimed at entirely different goals and therefore neither makes the other obsolete. Your retarded simplification of the differences doesn't demonstrate an understanding of their relative purposes. I'll outline it in layman's terms for your simple mind. JPEG is suited for photographic images where pixels don't fall into uniform blocks allowing lossy compression to have a minimal impact on perceived loss of quality. PNG is suited for vector web graphics and images requiring hard outlines and smooth gradients. It is a lossless format but is compressed using the deflate algorithm commonly employed by web servers among other things. In short, it would appear that you've used Photoshop a little, compared a few file sizes and naively think you understand the finer points of complex bitmap formats. Newsflash: you don't - get a clue before you expose your incredible stupidity next time.
ReplyDeleteAs pointed out that, it will add extra accessibility to mobile Internet users.
ReplyDeleteThis comment has been removed by the author.
ReplyDeleteA Stéréoscopic integration with only diff of pictures(compressed) soon ?
ReplyDeleteAnd to WebM too ?
That will be so cool !
I'd like deposed a patent for this...
Proprietary? Or will Google provide it Open Source?
ReplyDelete@Craig I'm afraid you're an immense jerkass-wad, and that you wouldn't be able to say what you said out face to face without a computer.
ReplyDeleteNewsflash: using 'newsflash' doesn't add any emphasis to your conclusion.
WebP team read this and be ashamed: http://x264dev.multimedia.cx/archives/541
ReplyDeleteIf you are going to make a web specific image format, why not embed a metadata tag containing a URL to a better (resolute/quality) image. You could create multi-scale images as a chain of multiple files (whereby the lowest quality image is selected by default or if some CSS setting suggests it as default). Then the http client can recursively request better images in the background and refine the page dynamically. I'm sure that this is not a novel idea.
ReplyDeleteRegards,
Ryan Fogarty
מצדיע לכם
ReplyDeleteWell done
ReplyDeleteShould add XMP or EXIF support for image metadata. Easy enough for anyone to do, since you're using RIFF, but it should be standardized, and for now, you guys control the standard.
ReplyDeleteAlso, if you're adding alpha support to WebP anyway, alpha channel support in WebM (VP8 video) would be really nice.
ReplyDeleteI'd like to see:
ReplyDelete- support for XMP metadata
- psychovisual optimizations in the WebM encoder
- support for lossless encoding
- using both WebM inter and intra coding for images (as in the hipix image format)
well done!
ReplyDeleteAny word on the transparency support? Potential updates?
ReplyDelete