WebP: Google's new attempt at speeding up the web

WebP: Google's new attempt at speeding up the web

Summary: Anything to make the web faster is good in my books -- but I have a question. Why are we worried about fractionally speeding up the web when we live in a broadband world that's just getting faster and faster anyway?

SHARE:
TOPICS: Browser, Google
23

In their latest attempt to make the web even faster, Google has devised their own image format to rival JPEG. Since forever, JPEG has been doing a great job of compressing photographic images -- it doesn't support transparency, like other formats, but it does what it was meant to do very well.

Google has determined that JPEG isn't the best format for that anymore. WebP is a new compression that achieves an impressive average 30% reduction in file size for images. Google will be proposing an update for WebKit that will make viewing those images in a browser actually possible.

Now, I'm not complaining here -- anything to make the web faster is good in my books -- but I have a question. Why are we worried about fractionally speeding up the web when we live in a broadband world that's just getting faster and faster anyway? The web is speeding up far more quickly from just network speeds improving than tweaking compression on files that are usually only used to show photographs online.

Sure, Sites like Flickr might see notable speedups -- with each page showing several jpeg images typically -- but most webpages largely use png or gif images (when you can't do something in CSS).

I'll stop whining now. WebP is still cool, and like I said before, anything to make the web faster is good in my books.

Topics: Browser, Google

Kick off your day with ZDNet's daily email newsletter. It's the freshest tech news and opinion, served hot. Get it.

Talkback

23 comments
Log in or register to join the discussion
  • JPEG2000 Flopped

    Seems to me JPEG2000 addresses this and flopped.
    guihombre
    • RE: WebP: Google's new attempt at speeding up the web

      @guihombre
      google is just trying to own everything, and wit the release of the majority of people who use google Chrome hating it, becuase htere is next to no personalization available, or really anyworth, they have to do something. picture compressions would be a good way to do that. if not for the web as a whole but for here own Photo sharing services. i doubt it has much of anything to do with saving network traffic at all
      Ez_Customs
      • RE: WebP: Google's new attempt at speeding up the web

        @Ez_Customs snagging. thanks for sharing! <a href="http://www.chanelhandbagsreplica.org/chanel-flap-bags-c-4.html">chanel flap bags</a> <a href="http://www.chanelhandbagsreplica.org/chanel-rose-bag-c-3.html">chanel rose bags</a> <a href="http://www.chanelhandbagsreplica.org/chanel-shoulder-bag-c-7.html">chanel shoulder bags</a>
        3shao
  • RE: WebP: Google's new attempt at speeding up the web

    It's not so much about the speed of one web page (though, that's certainly part of it), it's about reducing overall network usage, allowing any network that these images are served up over to gain previously wasted bandwidth. It's more about reducing the overall percentage of traffic. If you reduce you own web page by say 25% in network use, most users wouldn't notice if it already loaded quickly, but if you're serving up 1000000 pages a day, you're going to pay for those bytes. If you reduced your byte load by 25%, you reduce your spending too. Not to mention, there are fewer network traffic jams and they don't last as long.
    Software Architect 1982
    • RE: WebP: Google's new attempt at speeding up the web

      @Digital Video Expert Yes, I agree on a small scale it might not feel like any big difference but Google is always thinking in terms of scale. When you multiply that number up by millions it does make a difference.
      jack of daniels
    • RE: WebP: Google's new attempt at speeding up the web

      It's also a somewhat dangerous industry precedent to compromise when given the choice between streamlining a process or product and letting it sit static. It would be like developing software where 33% of the memory nabbed by the program is never released until the OS kills it, and being OK with that because a lot of users are running many-core CPU's w/ at least 4GB of memory.
      R3fleX
  • RE: WebP: Google's new attempt at speeding up the web

    In the mobile web world, speeds are lower, and caps are limited. This will clearly help in that case. Also, in developing countries, speeds are still capped depending on how much you pay. Its 64k/128k/256k tiers. At 64 or even 128k, this represents a significant speedup.

    Also, think about the amount of data moving across the whole WWW network. This will help a lot with that as well. Add to that the disk(space) requirements on websites and caches, backups and the ripple effects of reducing data usage by 30% are very high.
    ALISON SMOCK
  • Thanks, again Garett

    for another horribly and very uninformative copy & paste job from some other site. Every weekend now. Keep'em coming.
    Jared Neale
    • RE: WebP: Google's new attempt at speeding up the web

      @Jared Neale If you don't like Garett's articles, why do you continue to read them? Also, you should really proof read your comments before submitting them- especially when you're criticizing someone's writing skills.
      ddferrari
      • I was wondering the same...

        @ddferrari / Jared Neale
        Ok, so the guy doesn't like how or what Garett writes. Errr, so why bother reading it in the 1st place, or moaning about it in the 2nd? It might be a 'copy & paste', but I hadn't read it until now...
        naibeeru
      • RE: WebP: Google's new attempt at speeding up the web

        @ddferrari Doesn't look like I have to worry about it now. Garett seems to be gone. Thanks zdnet, finally!

        I am not criticizing the blog, folks....just the author. There's a difference. He isn't the only author of the blog. You people are idiots. For real. So, you get your kicks out of criticizing me. I criticize a professional and you criticize me? This is one of the many problems with America. There is ONE legitimate source to criticize on this blog and that is the author. The rest is just you and your ego. Just read and comment on the topic/article's contents/writing and move on.
        Jared Neale
  • RE: WebP: Google's new attempt at speeding up the web

    Google - still acting silly. Trying to grab the spotlight with a minor fix to a major problem. Actually, not a fix.
    rufioorljdfslai
    • RE: WebP: Google's new attempt at speeding up the web

      @rufioorljdfslai what are you saying? A 30% size reduction is huge. I know it will speed up my pages. And it will really speed up the web in the mobile world. Where bandwidth is scarce and most people pay by byte.
      Jimster480
  • RE: WebP: Google's new attempt at speeding up the web

    Yay! Let's all ignore the standards, provide a very small fix, and speed up the web, you know, like South Korea did, with superspeed web for nearly everyone! Nothing like adding a new format that no one needs.
    rufioorljdfslai
  • Don't get what the big deal is.

    So WebP sacrifices a bit of detail and a bit lower contrast in lower contrast areas when compared to JPEG. You get a smaller file. WebP seems to be a bit better on the adaptive part when compared to JPEG keeping compression low in areas of high frequency/high contrast while increasing compression in areas of low frequency detail/low contrast.

    But the JPEG image is a bit better for the extra memory. Looks like different trade offs that might have a real world 10-15% or so difference once the quality in all areas are set the same.

    Adoption faces a HUGE uphill battle.

    Nothing new here.
    Bruizer
  • It's the users, not the format

    Most users have no clue about reducing image file size. While their camera mya be taking shots 3000 pixels or more wide, they only need a fraction of that when displaying on the web or sending. Most users have no idea about compression, JPEG quality and don't understand that PNGs are superior for line drawings and JPEGs for photos. Some people even send BMPs around.<br><br>We could reduce the amount of data transmission by education and making sure all email clients, web or otherwise, do a reduction in size and quality of attached images as a default.<br><br>Use a new image format put together by an advertising company? - I don't think so. We already have JPEG 2000 available and MNG to replace GIFs, if we could convince people to use them.
    tonymcs1
    • RE: WebP: Google's new attempt at speeding up the web

      @tonymcs@... <br><br> - OK if it is an 'optional' reduction is size/quality. Sometimes the high detail is truly needed. Most people are selfish and wouldn't care if presented with the altruistically worded choice.. They are also lazy and won't bother with a workaround if it is forced on them as a default but will resent it. It could be presented in a way that makes it seem a benefit to the user instead of a sacrifice. How about having the client ask the user if they want to speed up the transfer or save bandwidth? - <br><br>What is a reasonable size for a crummy webcam image? 640x480 or 1024x768? But my sister would have a fit if she sent me a picture of her cat or the new baby that is less than the highest quality her digital camera will take. <br><br>I wonder if the bandwidth we are discussing is taken up more by streaming videos from youtube, annoying e-mail chain letters full of .PPS files and/or 10 or 20 separate images, and ridiculously wasteful HTML-by-default e-mails with stationery and piles of code, - than by all the cat and baby pictures in the world.
      opcom
  • transparency

    It's really the transparency support that would be the benefit. But, realistically, we don't need another format. We need designers who aren't lazy hacks, creating Web pages that are one giant bitmap.

    Everyone thought that "broadband" was going to erase all concerns over page size, but now they're burned by the explosion of mobile use. We're right back to dial-up speeds, folks. Have fun re-doing your sites.
    dgurney
  • RE: WebP: Google's new attempt at speeding up the web

    I'm willing to use it if:
    0.) the format is open
    1.) the browser 'plugin' is open (no spying)
    2.) there is a free converter to deal with legacy images
    3.) it is not 'locked' to a browser, there is a standalone viewer

    The point it, it has to be universally usable outside a browser to really catch on. such as - M$ Paint, gimp, etc. will eventually be able to open it as updates come along.
    opcom
  • Opacity and multiple images per file?

    Better compression is nice but somewhat less useful if it doesn't offer opacity. Just offering an optional mask layer would work fine.

    A good way to store multiple images in a single file would be very useful too. Cleaner than trying to mangle them into a single image and clip them back out. Just packing them into a zip file would work.
    MikeFM