WebP: 39.8% faster than JPEG

A WebP image of a man breathing fire

Google has used their insanely smart engineers to create an image compression algorithm that’s just as good as JPEG but 39.8% smaller. It’s called WebP and it’s pronounced “weppy”. You can create WebP images in Acorn, PixelmatorImageMagickLeptonica and XnConvert. If you use Photoshop, you can also install the WebP plugin.

Here’s a gallery comparing WebP to JPEG.

The problem is it’s currently only supported by Chrome and Opera, but if all of us in the web community make enough noise, we might succeed in getting it to be adopted by all major browsers. Please voice your support of WebP by …

  1. Commenting below. If we can get 500+ comments, then we’ll send this post to our contacts on the IE, Firefox and Webkit teams
  2. Re-tweeting this post
  3. Telling your followers, friends and clients about WebP

If we all voice our support of this awesome new image format, the web might just speed up by about 38.9% :)

If you can’t see the image above, then your browser doesn’t support WebP. Come on and help us get this adopted by all browsers!

Free Workshops

Watch one of our expert, full-length teaching videos. Choose from either HTML, CSS or Wordpress.

Start learning

Treehouse

Our mission is to bring affordable Technology education to people everywhere, in order to help them achieve their dreams and change the world.

Comments

364 comments on “WebP: 39.8% faster than JPEG

  1. WebP has so many technical merits over JPEG as highlighted in the article above and on the WebP homepage itself. It’s time for progress.

    • I heard about webp a while back and then it all went quiet I’m glad you’ve written this post, surely it’s a mo-brainer cheers :)

  2. Personally, I dont see the benefits outweighing the additional JS/hack clutter to support fallbacks for ‘older’ browsers.
    Best of luck guys!

  3. This sounds huge. If we could reduce all image files by 40% imagine what that would mean in terms of bandwidth and energy savings!

  4. Cool, just imagine what they can do with the new movie format if they could do this to an image format.  Good stuff.

    • Converting from one lossy format to another, poorly supported lossy format is a surefire recipe for making your images look rubbish. I would highly recommend against it, even if Ryan Carson is tooting WebP’s horn.

  5. Mobile devices need this. I mean why not support it? Just do it.

  6. I could notice and got really impressed with how fast the image at the top of this blog post loaded even before reading the title and knowing what this post was all about… As a designer I’m totally sold on that! And It would be great to see the browsers’ developers adopting also this technology and take the internet to the next level.

  7. I don’t understand a new image format is needed, when you can get so much more out of standard JPEG if you use it right – see jpegmini.com.

      • They are coming… Currently the site enables you to evaluate the quality of JPEGmini by processing a single photo at a time.  In a few weeks we will launch a web service enabling you to process full albums.

      • They are coming… Currently the site enables you to evaluate the quality of JPEGmini by processing a single photo at a time.  In a few weeks we will launch a web service enabling you to process full albums.

    • You had me interested. Then I went to jpegmini with a random banner ad that I quickly saved from another tab. I compressed the image and I’ve got to admit the loss of quality was huge.

      I think what people are interested in, is lossless compression.

    • You had me interested. Then I went to jpegmini with a random banner ad that I quickly saved from another tab. I compressed the image and I’ve got to admit the loss of quality was huge.

      I think what people are interested in, is lossless compression.

  8. Don’t forget it lacks ICC-support (any designer who stands behind this is not a designer worth his/hers title). No alpha channel, and supposedly Google didn’t do the whole benchmark thing very well (the had to revise their first version as their numbers were pure crap). 38.9% is just a random number, is holds no real value since it’s variable, you could just as well say 11.4% or 65.2%, it all depends on the image.

    But as it stands now it’s not a better format than JPEG, it’s actually worse in terms of usable functions for a better web,  it’s faster (and as a guy who scores 99-100/100 on page speed tests I really like speed), but you can’t sacrifice everything for speed, sometimes accuracy has to be a goal too.

    • Tozz in the politest way that’s a VERY introverted perspective. We’re hosting over 100,000 user submitted images and the space saving would be invaluable. If you don’t need to use WebP for your designs – just don’t. It’s a shame to bash a technology that has a multitude of advantages to big websites just because it isn’t fitting your design toolbox.

      • How is it introvert? I’m saying there’s no point in having a format that’s smaller but doesn’t support core features for correct display. You have no idea how much space you would save so I don’t understand how you can say it’s invaluable, seems quite drastic if you ask me.
        We’re currently at 2million images in S3 (so out potential space saving would be beyond invaluable then, no?), it’s not like I don’t know the concept of disk space.

        I have already acknowledged the single benefit, smaller images in some cases, but the cost is way too high, dropping ICC support is just a brain dead decision in this wide gamut day and age. What are those other multitude of advantages you speak of?

          • Perhaps a better understanding of some of the fundamentals before you issue your “thumbs up”? Tozz is correctly stating some issues that are easily overlooked by, for example, people who don’t understand what gamut means.

            The trouble with supporting a new image format is that even if it is not very widely used, you still have to support that format in your browser for years to come. This adds bloat to development as well as runtime memory.

          • I’m starting to feel like an old person. You have the head start on me with regard to the technical side and I’m not even going to pretend to want to understand the underlying tech further. Gamut, ICC, la-la-la-la.

            We just want to serve images faster and save disk space so do I believe a) a couple of geezers on an Internet forum or, b) the brightest engineers in the world’s most profitable internet company.From a business perspective it’s a slam dunk. Reputation is everything. I’ll stick with the “thumbs up” for WebP.

          • Replying to John due to threading issues…

            I would also question why someone who hasn’t performed basic analysis and enhancements on their sites want a smaller image format when they have incredible gains waiting in front of them with readily available tools. Someone who doesn’t employ browser caching, compression and spriting should really start in that end.

            Google has plenty of failures under their belt, anyone remember Wave? You can’t blindly just accept something Google publishes as the next best thing (I’m a long time Android user and I use a lot of Google products, so it’s not a personal bias against Google).

        • I’m just curious: How good is the ICC profile support (v2, v4) across browsers/OS ? I couldn’t really find anything useful on the interwebs and you seem quite knowledgable 

          • It’s actually horrible, and I don’t think a profile-less image format will help it at all.

            Safari does v2 (they apparently dropped v4 recently) (Win/OS X)
            Firefox does v2 (Win/OS X)
            IE9 does v2 and v4 (I read somewhere that it’s not true due to some conversion to sRGB being done or whatever it was, it does however pass the test at http://www.color.org/version4html.xalter on my custom profiled system).

            Opera and Google seems ignorant beyond belief when it comes to colors (Chrome supposedly have v2 support with a custom switch but I’ve never gotten it to work).

    • Google’s revised numbers are absolutely solid and quite astonishing. Have you read the recent report? http://code.google.com/speed/webp/docs/webp_study.html

      It’s a sweeping statement to say “any designer who stands behind this is not a designer worth his/her title”. Really?

      • Nice copy paste out of context, really, was it necessary?
        “Don’t forget it lacks ICC-support (any designer who stands behind this is not a designer worth his/hers title)”
        That was what I wrote, why did you feel it necessary to totally butcher my sentence?

        I’d say that someone who doesn’t care about color should call themselves a designer. Color profiles are crucial for a correct color rendition. The format also lacks support for metadata.

        A new image format should be more well thought out than this, Google stared blindly at speed (how about cpu-consumption?) and forgot all the other parts that makes an image. I’m happy Mozilla rejects semi functional formats, maybe Google will go back to the drawing board make the format actually usable.

        And again, stop with out of context pastes, it looks really horrible.

          • So you’re not allowed to have an opinion about something without having something else to replace it with?

            It’s not whining, it’s well grounded logic thinking based on years of work with photography, the web and software development. The difference is quite apparent for someone who’s not out for meaningless bashing.

            WebP is an attempt to solve one part of the speed aspect of images (it does nothing for latency, something that’s usually a bigger problem than file sizes). In this attempt it totally drops the ball on everything else though, which is a horrible tradeoff. Now if Google actually addressed these issues and added ICC v2, v4, alpha layers and metadatasupport it would be very interesting and I’m sure Mozilla would agree, but right now they’re taking the sane and logic stance. Adding an obviously unfinished image format to Firefox makes no sense.

          • You make a very good argument, and one I do agree with. But something to
            consider is, unless I missed this, Thinkvitamin is not suggesting
            getting rid of the JPEG format, but adding another option for sites
            where speed & decent photography is more important than all the
            things you noted. I can imagine large e-commerce sites would love this,
            but not smaller, nich e-commerce sites frequented by creative types.

            Or maybe I’m just wrong — you both make good points though.

          • Seems I have to reply to myself, guess it’s a nesting thing.

            I understand the suggestion is not to drop support for JPEG, but I still think you should require a certain maturity in something as essential as image formats before asking for companies to support it.
            If it was a simple patch with guaranteed future compatibility I would be all for it, but it seems like this is not the case. I think Mozilla has a sound and valid stand point, just like Microsoft has with HTML5 and how they refuse to implement the unfinished specs, I consider WebP to be unfinished just as much as I consider WebSockets to be.

            I’m currently building a photography-related site and I see a use case for WebP in thumbnails and JPEG for full size images, whenever it becomes an viable solution.

          • Todd,

            We have focused on quality and compression first. Check out Opera’s success with it: 
            http://my.opera.com/chooseopera/blog/on-a-horse-opera-turbo-to-the-rescue
            and
            http://googlesystem.blogspot.com/2011/04/opera-turbo-uses-webp-to-compress.html
            Work on meta and ICC is ongoing. I encourage you to give us more feedback on the webp-discuss mailing list. We have shared our thoughts on the roadmap and added alpha support in the experimental branch, and your feedback would be great.

          • After seeing the examples posted by Richard I’m convinced this is not a format to compare against JPEG. The loss of detail is extreme to the point that the images aren’t even comparable, at all.
            It’s like comparing MP3 at 128kbps CBR against FLAC, sure, one of them is smaller, but it also lacks a good amount of sound.

            It would be interesting to see WebP comparisons that retain at least a similar level of detail and how much of a size decrease you can expect then.

          • I agree with thinkdesign. I build websites that are commonly accessed by folks who only have access to dial up, and every byte counts to make their web experience bearable. While I’d stick with higher quality stuff on pages accessed mostly by people in Toronto, WebP would allow me to move toward a more image rich environment for our clients in the North.

    • The lack of ICC support and alpha channels certainly makes JPEG a better format for certain tasks. For the masses of user generated content, where the user doesn’t care about ICC and isn’t looking for an alpha channel, I will gladly make the trade down on features.

    • I kinda miss the point of ICC for a web graphics format at all, I mean, lets say webp is RGB123 and just live with that. Why not remove features, so that the decoder would be less complex?

      No metadata isn’t cool though, I’ll agree with that.

        • I think that we should configure the device, not the media.

          Good point, the bpp count will continue to grow (and 8 has stayed for long enough). Of course, the JPEG has the 12-bits in the specs, but that wasn’t in the original. So, maybe if the webp will stay truly open, it will adapt when the wider dynamic range monitors are more widely spread.

  9. This sounds amazing! Would mean I wouldn’t even need to use ‘slush it’s anymore! Will save time and money for mobile users!

  10. This sounds amazing! Would mean I wouldn’t even need to use ‘slush it’s anymore! Will save time and money for mobile users!

  11. I think WEBP standard should be supported by others so that web become faster.

  12. Mozilla seems to take a conservative stance and will not include WebP unless it makes some improvements. And I can understand that approach. The web doesn’t need a new half baked image format. That said, a lot of good stuff is in the pipeline for WebP, like alpha channel, and in the end this will be a kick ass image format that will be included by all browser vendors.

  13. I support this initiative however I too feel Alpha support would make me a whole lot more interested…. Maybe even some form of simple animation support.

  14. Why not? A great format must have great support for the developers comunity, so raise this to IE, Webkit and Mozilla!!

  15. i like the web!it’s a fashionable band! thank you for sharing!

  16. Great idea Ryan… Everyone should be on board to make the web faster!!

  17. Yes, yes and yes again, that’s all I have to say. The real issue is getting rid of the non supporting browsers though… lets hope that issue doesn’t take as long as IE6 to go away.

  18. that would sooo save my life, I’d post the 500 comments myself if I had the time ;)

  19. Count me in! Sadly I’m still a bit confused and disappointed to see resistance on this (and other forums) from the some sections of the design community.

    With a profitable holiday portal serving over 100,000 user submitted images, any speed/space saving is a major win and we’d love to switch to WebP. Some of the creatives who are resistant could perhaps think a little further than Photoshop? Massive applause for his initiative – it’s a real solution to real world problems.

    • Exactly! Everyone seems to think that the suggestion is to replace your well thought out and carefully crafted design elements with WebP images. That really doesn’t seem to be the use case for WebP. Just like with most anything, the right tool for the right job.

      Crappy user generated content = WebP
      Wedding photography = JPEG

    • The problem, of course, is that adoption won’t be quick, however much you want to rush it. Sometimes we complain a lot about the slowness of processes we aren’t involved in, like government, or technology standards, but the sad truth is that if everybody’s that slow, we just have to accept that it’s going to a take a while before you can be confident clients can handle WebP. It took years for PNG to be adopted, and that had the killer feature of alpha. WebP offers worse quality (currently) and somewhat better compression than JPEG, but it doesn’t help us do anything new. Do you really think it will be quicker to adopt than PNG? You see, until that point, you’re going to have to keep two copies of all your images if you want to serve up WebP, which is hardly a space saving.

      These things are just very slow, which is why the designers and programmers want to pick the best formats which are worth creating a legacy of content with. We can do better, and we should do it right before forcing browsers to support a new format basically for ever.

  20. I support the continued development of a new image format but we’re a long long way from wide spread support. Sounds like there are a few issues need to be addressed before vendors will get 100% behind this.

  21. Introducing any kind of new file type for the web is a major headache for developers. Mainly because it adds extra complexity to workflows and means there’s lots more thinking to be done about providing consistent support. Even if all of the major browser vendors started to support WebP tomorrow, there’d still be the need for a polyfill of some kind, and every single WebP file would need a JPEG alternative — that means oodles of extra work, more difficult QA and an extra element of fragility.

    I’m all for better image quality and smaller file sizes, but don’t think that WebP is the silver bullet. There’s a lot that can be done with more efficient JPEG and PNG compression to reduce file size and improve image quality (I think there are also questions about whether WebP is actually much better than JPEG anyway). I’d much rather have seen Google put some effort into tools which help us improve the creation of images in legacy formats, before trying to create an entirely new format from scratch. It smacks of idealism.

    • I think the primary use for WebP would be for low-quality user generated content (think Facebook profile pictures). WebP would be a massive boon for that. Support would be simple as well, upload a picture then show a default “Your browser can’t display this, please install x,y or z”. Most browser makers would want to be on the x, y or z list I imagine.

      I don’t think anyone is suggesting this be a replacement for JPEG, the lack of meta data and ICC support means that JPEG has it’s place.

      • My reading of the original post was that it was indeed advocating WebP as a replacement for JPEG. Fragmentation of standard formats and standards can end up messy.

        • I completely agree about fragmenting standards. I am sure the creators of WebP would love to see it replace JPEG, doesn’t mean it has too. I for one would not mind an efficient, faster standard for use when I want something that is faster and maybe not as feature rich.

  22. Adding support will not do any harm. It will rather open one more option of Image Format that can be used on the Web. People will have a choice to choose between the speed and quality as required.

  23. Go WebP!
    What about mobile browsers?It makes even more sense on mobile browsers as you might have a limited connectivity and every kb saved is welcome!

  24. In the days of broadband internet, I fail to see the need for a better compression method. Coders are, after all, at the moment busy with turning as many images as possible into one ‘sprite’, just to reduce the number of http requests, even if this means increasing the image size. If I’m using a 2Mbps connection, which is in itself not as fast since a long time now, what’s the difference between a 50KB image and a 70KB image?

  25. I, for one, welcome our new image format of choice. Roll on a full spectrum of browser support!

  26. I, for one, welcome our new image format of choice. Roll on a full spectrum of browser support!

  27. Sounds good, although they’ll be the inevitable “not supported in widely used browsers” argument. And then we’d need possibly another polyfill script to deliver .webp to some browsers, and a fall-back .jpg to others. Messy.

    Questions could be: could .jpg ever be improved to match .webp in terms of compression? Or is it a case that .jpg is what it is – with file size to quality levels set in stone?

    Is there also any movement on transparency in images? Faster equivalent .jpg formats are all well and good, but the most flexible image type, but with often the largest file size is the png32. I’d like to see a widely supported file size improvement on images with alpha transparency.

    All that said, I am in favour of the premise of .webp. If you don’t drum up noise about new things they may never become a reality. That also being said, I can’t see .webp being a a “real” choice and consideration for a long time to come …

  28. I like turtles… In other breaking news… Yes please, a faster web is a happier web. Will Microsoft ever support it though?

  29. I find that those 500 comments you’re requesting is for something else… this is not the way to make a point. Get a petition or something for that… 
    Comments count is biased because replies shouldn’t count.

  30. @facebook-906400362:disqus But We can clearly see with all of Google’s other products they bring it out fast and then do multiple iterations and bring out updates quicker it also allows for other developers to tinker with it and figure how to maybe do something better. Google is mostly concerned with setting a solid base on which to grow on. I believe that it should be adopted if not fully at least partially to give it the momentum to continue to grow. New innovative ideas are usually first frowned upon on because as humans we try to resist change when we feel that we are alright with a certain concept already. But without change we wouldnt be where we are technologically now. Yes its not exactly the best at the moment but neither was android when it first came out. Google just has a way of turning what they touch to gold and the promises coming from the WebP front is enough to make me salivate about what it could be a year from now 

    • Sure, if it was as simple as that. Updating support for image formats is usually not a simple thing to do, especially not if a lot has changed. Mozilla has clearly stated that they will look at the format again when it matures, which I am all for, I just think it’s way too early to put resources into it.
      If Google promised ICC, Alpha channels and Metadata I would be more exited, right now they don’t say anything, just “it’s somewhat faster to download” (I haven’t seen any numbers on CPU-time required to render). Not good enough, especially not from Google.

  31. How about mentioning that Mozilla has stated that they will not be supporting it?

  32. I had to do our monthly knowledge review session this morning and before I started my speech I asked a question if anyone heard about WebP…I was so happy to be the only one in our office who knew about it. So now our whole office is well educated about it.

  33. This is a great idea. weppy is easy to use and speeding up page loads benefits everyone from desktop to device. It will doubtless spark some new layout ideas for achievable image heavy designs too which can only be a good thing. Fight for it folks!

  34. I’ve heard of this new format previously and it sounded really interesting. I’m going to trust Ryan’s done his research on this one and say I’d like this format to be supported by more browsers.

    Surely if the format initially has flaws it can be incrementally improved over time until it satisfies all requirements?

  35. Hyper link in article didn’t get me where I needed to be. http://telegraphics.com.au/sw/product/WebPFormat

  36. As a direct replacement for JPEGs WebP sounds like an sensible idea. Even so, careful consideration when compressing along with judicious usage of a compression tool like smush.it usually gives pretty good results with JPEGs. Maybe not 38% worth but good enough to be going along with.

    Chucking another format with minimal support into the ring seems a little bit of a heavy handed solution. I’d prefer to see Google putting their energy into developing a format that supports alpha transparency etc. and that could go head to head with PNG. Cutting PNG size down 40% – now that would be grand!

  37. As a direct replacement for JPEGs WebP sounds like an sensible idea. Even so, careful consideration when compressing along with judicious usage of a compression tool like smush.it usually gives pretty good results with JPEGs. Maybe not 38% worth but good enough to be going along with.

    Chucking another format with minimal support into the ring seems a little bit of a heavy handed solution. I’d prefer to see Google putting their energy into developing a format that supports alpha transparency etc. and that could go head to head with PNG. Cutting PNG size down 40% – now that would be grand!

  38. I would like to see more support for webP compression and made the standard across all browsers.

  39. What I love about Google is they are willing to tackle long-standing problems in a simple fashion by being very opinionated with their approach.  WebP won’t make everybody happy, won’t solve every single image problem ever, and it doesn’t have to – we need a new alternative that addresses some long-standing problems even if not all of them and even if the newly created problems can be dealt with in other manner.  In short: go WebP, go!

  40. What I love about Google is they are willing to tackle long-standing problems in a simple fashion by being very opinionated with their approach.  WebP won’t make everybody happy, won’t solve every single image problem ever, and it doesn’t have to – we need a new alternative that addresses some long-standing problems even if not all of them and even if the newly created problems can be dealt with in other manner.  In short: go WebP, go!

  41. Even if we’re talking about a technology that may take some time to achieve full adoption, it’s good to see some innovation happening in an otherwise stagnant arena. I say keep pushing for it, keep voicing concerns about the technology and if it’s a relevant enough problem for the community as a whole, then changes to the spec can be made. The greatest thing about the web is that it never stops evolving.

  42. Even if we’re talking about a technology that may take some time to achieve full adoption, it’s good to see some innovation happening in an otherwise stagnant arena. I say keep pushing for it, keep voicing concerns about the technology and if it’s a relevant enough problem for the community as a whole, then changes to the spec can be made. The greatest thing about the web is that it never stops evolving.

  43. Put me in for WebP support on other browsers. It may not be everything to everybody, but it is a step in the right direction.

  44. I’d like to see this supported. I’d still love to see alpha channel, metadata improvements, but 40% reduction in file sizes is worth something.

  45. I’d like to see this supported. I’d still love to see alpha channel, metadata improvements, but 40% reduction in file sizes is worth something.

  46. Looks great, like an awesome option for thumbnailing scripts or sites that host a large number of images.

    Where I’d get really interested is in testing for browser support for weppys on the client side, and serving up weppys vs. jpegs, in a dynamic AJAX setting where part of the request could be “what images i support” and the returned data gets to either send weppys or jpegs based on the browser support…

    The lack of alpha bugs me (but hey, we have PNGs for that) – Obviously the engineers are focussing on “good enough” meanwhile making impressing savings in space

  47. sounds excellent, especially for mobile sites where bandwidth is at a preium.

  48. Its a step in the right direction, for large images such as photographs, not something i would use in the actual design of the website, the chances of anything below IE10 supporting this are also 0.

  49. @ Erik,

    Why couldn’t you have taken a higher road and been more helpful? If you know something better, take the teacher approach. Manners are really just a tried and tested way of being able to communicate better. No one likes a grouch, particularly one that isn’t helpful or courteous. Whatever point you might have had, however relevant, is undermined by your approach to the topic. Go vent somewhere else. Take this quote to heart:

    “All I ask of you is one thing: please don’t be cynical. I hate cynicism- it’s my least favorite quality and it doesn’t lead anywhere. Nobody in life gets exactly what they thought they were going to get. But if you work really hard and you’re kind, amazing things will happen.” 
    – Conan O’Brien

    As for WebP, you and I both know that this is only the beginning of a better compression method for images. It’s been a long time coming and I’m excited to see what will happen to the new format. 

    • There’s another quote I like, “like the devil reads the bible”.
      If you want everyone online to be evil and cynical, they will appear that way, but I ask you to not put words in my mouth or attempt to downplay me just because you don’t agree with my opinions. I wasn’t venting anything, I was offering a sound standpoint based on experience and facts (which many readers appear to agree with), there’s plenty of “+1″ posts here which in my opinion helps absolutely no one.
      The only cynicism in my posts are what you plant there yourself, that’s no fault of mine. It’s not cynical to factually disagree with something, just for future reference.

      Better compression methods are welcomed by everyone, but as JPEG2000 showed, it isn’t just about compression.

  50. Go-go-go!
    When did proposing options (that actually do something to help) became something “not to do”?

  51. Go-go-go!
    When did proposing options (that actually do something to help) became something “not to do”?

  52. This would be great if we could get all browser makers to support this. I’m just discouraged by the length of time it took to get all aspects of PNG fully supported.

  53. I think this could be a great addition to our toolkits. We should make noise while browser makers are in a listening mood, as they have been of late. I only hope it’ll take a lot less time than getting all aspects of PNG supported across the board. :)

  54. “Promote us or we else we won’t send this information to people who probably already know about it!!!!!!”

    Really Carsonified? I seriously appreciate some of your content (thus why I subscribe to RSS) but at times you are one of the biggest troll blogs and make no attempt to even hide that the post is largely about getting you more money. (See recent “Five reason why you need to be at The Future of Mobile” as an example)

  55. Sorry to point it out, but this isn’t necessarily the right move. Technology moves on very quickly; WebP isn’t outstanding, it’s just a bit newer. Whether or not to adopt WebP is a question of comparing it against other current technology, not pointing out that last gen technology, is, in fact, last gen. There are just a few image formats which are universally acceptable, and adding another one to that list is a huge challenge. It means basically supporting it indefinitely. Why is WebP better than JPEG2000, or PGF? This is a huge decision to make, and it’s well worth being a bit more informed about it. As it is, the concerns over features, compression ratio, patents, better alternatives, speed, and visual quality all suggest we should hold back judgement for a bit. Some of those will certainly be cleared up (patents in particular will be rapidly assessed, with every expectation that the format turns out to be fine), but there is simply no reason on the face of it why VP8 frames are the best compression technology available to us at the moment. Simply because Google has announced it does not mean it is mature, or superior; it just gets more press. So, contra a few other commenters, it is certainly not “retarded” or similar to have that caution.

    For a technical look at the opposing view, see http://x264dev.multimedia.cx/archives/541

  56. Decrease in filesize => decrease in wait time => better internet. Go WebP!!!!

  57. Decrease in filesize => decrease in wait time => better internet. Go WebP!!!!

  58.  We need to support the best standard, not just Google’s standard.  This is a very one sided view and you never compare it to  JPEG XR, which is a better spec.   Don’t just be a shill for Google.  There is a reason why Mozilla won’t throw their weight behind WebP and you never discuss that reason.

    -1 support.

  59.  We need to support the best standard, not just Google’s standard.  This is a very one sided view and you never compare it to  JPEG XR, which is a better spec.   Don’t just be a shill for Google.  There is a reason why Mozilla won’t throw their weight behind WebP and you never discuss that reason.

    -1 support.

  60. Your use of the number 39.8% is ignorant and misleading. Google claim WebP files are on average 25-34% smaller. The compression depends on the kind of image and so varies. When you try to claim the benefit is known to the precision of one decimal place you are indicating that you are either very ignorant, are wilfully misleading, or both.

  61. Your use of the number 39.8% is ignorant and misleading. Google claim WebP files are on average 25-34% smaller. The compression depends on the kind of image and so varies. When you try to claim the benefit is known to the precision of one decimal place you are indicating that you are either very ignorant, are wilfully misleading, or both.

  62. I am confused. Is it 39.,8% faster or smaller (smaller does not always means faster).

  63. Brilliant. Imagine the individuals with slower connections able to load websites that much faster without sacrificing quality! I love seeing technology unfold before our eyes. Now if only it were supported by all other browsers. However, I won’t hold my breath for Internet Explorer.

  64. Brilliant. Imagine the individuals with slower connections able to load websites that much faster without sacrificing quality! I love seeing technology unfold before our eyes. Now if only it were supported by all other browsers. However, I won’t hold my breath for Internet Explorer.

  65. As a image compression enthusiast (yes, there is such a thing) – I would totally support this implementation. Plus, the size savings is hard to ignore.

  66. While I agree this isn’t the best solution for every situation, it certainly has its advantages in a lot of situations. It will take education on our parts to make sure we are using it in the correct scenarios, but I like having another option where speed and/or space is a concern. I would like to see ICC-support though…

  67. I’ve long abandoned using JPEG’s in my webpages*—I’ve since switched to PNG since it supports transparency. But, man, Google this is uber awesome! Speed it is!

    *Except those from my point-and-shoot camera which is hopelessly tied to JPEG.

  68. All browsers should embrace this new image algorithm, it’s in everybody’s best interest. The only way to keep technology moving forward is through a community effort.

  69. All browsers should embrace this new image algorithm, it’s in everybody’s best interest. The only way to keep technology moving forward is through a community effort.

  70. Please add native support for WebP format for all operating systems and browsers

  71. Please add native support for the WebP image format for all browsers and operating systems.

  72. Please add native support for the WebP image format for all browsers and operating systems.

  73. Come on guys.It’s not about if JPEG is fine or we don’t need another image format. It’s about optimitazion. Formats bring solutions. that’s all. let’s celebrate this as an evolution for a better web :)

  74. Come on guys.It’s not about if JPEG is fine or we don’t need another image format. It’s about optimitazion. Formats bring solutions. that’s all. let’s celebrate this as an evolution for a better web :)

  75. I want WebP support as well.  Please pass this along to the powers that be! lol!  And what an argument we had here!

  76. GOOOO for webp! Just optimized a 342kb PNG to a 45kb WEBP, and they look the same!

  77. And you’re an insanely stupid blogger because first you write 38.9% faster and then it’s suddenly 38.9% smaller.

  78. Seems I have to reply to myself, guess it’s a nesting thing.

    I understand the suggestion is not to drop support for JPEG, but I still think you should require a certain maturity in something as essential as image formats before asking for companies to support it.
    If it was a simple patch with guaranteed future compatibility I would be all for it, but it seems like this is not the case. I think Mozilla has a sound and valid stand point, just like Microsoft has with HTML5 and how they refuse to implement the unfinished specs, I consider WebP to be unfinished just as much as I consider WebSockets to be.

    I’m currently building a photography-related site and I see a use case for WebP in thumbnails and JPEG for full size images, whenever it becomes an viable solution.

  79. After seeing the examples posted by Richard I’m convinced this is not a format to compare against JPEG. The loss of detail is extreme to the point that the images aren’t even comparable, at all.
    It’s like comparing MP3 at 128kbps CBR against FLAC, sure, one of them is smaller, but it also lacks a good amount of sound.

    It would be interesting to see WebP comparisons that retain at least a similar level of detail and how much of a size decrease you can expect then.

  80. Exactly! Everyone seems to think that the suggestion is to replace your well thought out and carefully crafted design elements with WebP images. That really doesn’t seem to be the use case for WebP. Just like with most anything, the right tool for the right job.

    Crappy user generated content = WebP
    Wedding photography = JPEG

  81. Replying to John due to threading issues…

    I would also question why someone who hasn’t performed basic analysis and enhancements on their sites want a smaller image format when they have incredible gains waiting in front of them with readily available tools. Someone who doesn’t employ browser caching, compression and spriting should really start in that end.

    Google has plenty of failures under their belt, anyone remember Wave? You can’t blindly just accept something Google publishes as the next best thing (I’m a long time Android user and I use a lot of Google products, so it’s not a personal bias against Google).

  82. I agree with thinkdesign. I build websites that are commonly accessed by folks who only have access to dial up, and every byte counts to make their web experience bearable. While I’d stick with higher quality stuff on pages accessed mostly by people in Toronto, WebP would allow me to move toward a more image rich environment for our clients in the North.