Description of the ART Image Compression Algorithm?

Discussion in 'Digital Photography' started by Richard Ballard, Jul 17, 2004.

  1. I have been comparing results compressing color digital imagery
    with the JPG, TIFF, and ART image compression algorithms.
    (ART is the compression algorithm used in Internet Explorer.)
    The ART algorithm apparently creates minor artifacts during
    compression (i.e., it is not a lossless algorithm), but it achieves
    significant compression while maintaining image quality.

    Does anybody know a reference discussing the ART algorithm's
    mathematics?

    Thanks in advance.

    Richard Ballard MSEE CNA4 KD0AZ
    --
    Consultant specializing in computer networks, imaging & security
    Listed as rjballard in "Friends & Favorites" at www.amazon.com
    Last book review: "Guerrilla Television" by Michael Shamberg
    Richard Ballard, Jul 17, 2004
    #1
    1. Advertising

  2. "Richard Ballard" <> wrote in message
    news:...
    []
    > (ART is the compression algorithm used in Internet Explorer.)
    > The ART algorithm apparently creates minor artifacts during

    []
    > Richard Ballard MSEE CNA4 KD0AZ


    Are you sure? I though it was an ancient AOL format?

    David
    David J Taylor, Jul 17, 2004
    #2
    1. Advertising

  3. In article <eE3Kc.1418$>,
    "David J Taylor"
    <-this-bit.nor-this-part.uk> writes:

    >"Richard Ballard" <> wrote in message
    >news:...
    >[]
    >> (ART is the compression algorithm used in Internet Explorer.)
    >> The ART algorithm apparently creates minor artifacts during

    >[]
    >> Richard Ballard MSEE CNA4 KD0AZ

    >
    >Are you sure? I though it was an ancient AOL format?


    "Ancient AOL" used a version of Microsoft Corporation's
    Internet Explorer as its Web browser. ART is an image
    compression algorithm implemented in Microsoft Corporation's
    Internet Explorer.

    ART appears to provide excellent compression of color
    digital imagery with few compression artifacts. The artifacts
    I have observed reflect subblock effects during compression,
    and disturb hue more than brightness. (Empirically, the
    compression artifacts resemble a minor rash on human skin in
    image areas transitioning between light skin and dark clothing.)
    In some cases ART appears to provide better compression than JPG
    (and much better compression than TIFF).

    ART is a nice algorithm. I am seeking a reference describing
    its mathematical basis. I am curious.

    Richard Ballard MSEE CNA4 KD0AZ
    --
    Consultant specializing in computer networks, imaging & security
    Listed as rjballard in "Friends & Favorites" at www.amazon.com
    Last book review: "Guerrilla Television" by Michael Shamberg
    Richard Ballard, Jul 17, 2004
    #3
  4. "Richard Ballard" <> wrote in message
    news:...
    []
    > "Ancient AOL" used a version of Microsoft Corporation's Internet

    Explorer as its Web browser. ART is an image compression algorithm
    implemented in Microsoft Corporation's Internet Explorer.

    Thanks, Richard. I knew about the AOL / MS-IE link, but I think ART
    predated it. Unfortunately I can't quickly find any references to confirm
    this.

    []
    > ART is a nice algorithm. I am seeking a reference describing its

    mathematical basis. I am curious.

    I didn't find any references about that, either!

    Cheers,
    David
    David J Taylor, Jul 17, 2004
    #4
  5. In article <xy8Kc.1545$>,
    "David J Taylor"
    <-this-bit.nor-this-part.uk> writes:

    >"Richard Ballard" <> wrote in message
    >news:...
    >
    >>"Ancient AOL" used a version of Microsoft Corporation's Internet
    >>Explorer as its Web browser. ART is an image compression algorithm
    >>implemented in Microsoft Corporation's Internet Explorer.

    >
    >Thanks, Richard. I knew about the AOL / MS-IE link, but I think
    >ART predated it. Unfortunately I can't quickly find any references
    >to confirm this.


    ART is not included in the image processing software packages
    I am familiar with. I wonder if ART is a Microsoft Corporation
    proprietary algorithm?

    >>ART is a nice algorithm. I am seeking a reference describing its
    >>mathematical basis. I am curious.

    >
    >I didn't find any references about that, either!


    If you find references I would appreciate your forwarding them.
    A relatively high compression low loss color digital imagery
    compression algorithm must have an interesting design.
    I consider computer logic an art form, and I enjoy browsing.

    Richard Ballard MSEE CNA4 KD0AZ
    --
    Consultant specializing in computer networks, imaging & security
    Listed as rjballard in "Friends & Favorites" at www.amazon.com
    Last book review: "Guerrilla Television" by Michael Shamberg
    Richard Ballard, Jul 17, 2004
    #5
  6. "Richard Ballard" <> wrote in message
    news:...
    []
    > ART is not included in the image processing software packages I am

    familiar with. I wonder if ART is a Microsoft Corporation proprietary
    algorithm?

    I think it has nothing to do with Microsoft:

    http://whatis.techtarget.com/fileFormatA/0,289933,sid9,00.html

    gives:

    "ART - AOL Image file compressed using the Johson-Grace compression
    algorithm"

    I think it predates AOL using Microsoft.

    Cheers,
    David
    David J Taylor, Jul 17, 2004
    #6
  7. In article <99aKc.1577$>,
    "David J Taylor"
    <-this-bit.nor-this-part.uk> writes:

    >"Richard Ballard" <> wrote in message
    >news:...
    >
    >>ART is not included in the image processing software packages I am
    >>familiar with. I wonder if ART is a Microsoft Corporation proprietary
    >>algorithm?

    >
    >I think it has nothing to do with Microsoft:


    <RB snipped URL>

    >gives:
    >
    >"ART - AOL Image file compressed using the Johson-Grace
    >compression algorithm"


    I have not accessed the URL you reference since it does not appear
    to discuss the ART algorithm's mathematics. Interesting, I have
    some knowledge of compression algorithms and I have _not_ heard
    either of Johnson (sic) or Grace in the open literature. In
    EE graduate school I studied statistical communications theory.
    Signal compression algorithms are relevant to statistical
    communications theory.

    Nevertheless, based upon my preliminary studies, ART appears to
    be a superior compression algorithm for color digital imagery and
    I hope to find a reference to its mathematical basics.

    Richard Ballard MSEE CNA4 KD0AZ
    --
    Consultant specializing in computer networks, imaging & security
    Listed as rjballard in "Friends & Favorites" at www.amazon.com
    Last book review: "Guerrilla Television" by Michael Shamberg
    Richard Ballard, Jul 17, 2004
    #7
  8. Richard Ballard wrote:
    []
    >> "ART - AOL Image file compressed using the Johson-Grace
    >> compression algorithm"

    >
    > I have not accessed the URL you reference since it does not appear
    > to discuss the ART algorithm's mathematics. Interesting, I have
    > some knowledge of compression algorithms and I have _not_ heard
    > either of Johnson (sic) or Grace in the open literature.


    No, neither have I.

    > In
    > EE graduate school I studied statistical communications theory.
    > Signal compression algorithms are relevant to statistical
    > communications theory.


    I have a similar background, in digital signal processing.

    > Nevertheless, based upon my preliminary studies, ART appears to
    > be a superior compression algorithm for color digital imagery and
    > I hope to find a reference to its mathematical basics.
    >
    > Richard Ballard MSEE CNA4 KD0AZ


    I expect quite a few people would be interested in that as well. Good
    luck with your search.

    Cheers,
    David
    David J Taylor, Jul 17, 2004
    #8
  9. Richard Ballard

    Guest

    In message <eE3Kc.1418$>,
    "David J Taylor"
    <-this-bit.nor-this-part.uk> wrote:

    >"Richard Ballard" <> wrote in message
    >news:...
    >[]
    >> (ART is the compression algorithm used in Internet Explorer.)
    >> The ART algorithm apparently creates minor artifacts during

    >[]
    >> Richard Ballard MSEE CNA4 KD0AZ


    >Are you sure? I though it was an ancient AOL format?


    The AOL .art files create *major* artifacts. A friend of mine was using
    AOL's free match.com service, and the pictures he uploaded looked like
    garbage on the screen. I suggested that they might have a maximum file
    size, and might over-compress them in JPEG format if you pass the
    threshold, and that he might try sending the same 400*300 JPEGs at
    various compression levels, but even the 12K ones were reduced to about
    half the file size. Turned out that his AOL account was set to use
    extra compression for non-broadband connections, and .art files were
    being substituted. They were simply *ghastly*; I have never seen images
    as bad as these .art files. Even JPGs made in irfanview or PSP the same
    size as the .art files were so much better. His 0.12MP 15K uploads
    looked better than the 50K .art files made from 100K 3.1MP originals.

    We changed the setting in his AOL account, but he still has no control
    over what the women see when they view his pictures, which look like
    garbage in .art format.
    --

    <>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
    John P Sheehy <>
    ><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
    , Jul 18, 2004
    #9
  10. Richard Ballard

    Guest

    In message <>,
    (Richard Ballard) wrote:

    >"Ancient AOL" used a version of Microsoft Corporation's
    >Internet Explorer as its Web browser. ART is an image
    >compression algorithm implemented in Microsoft Corporation's
    >Internet Explorer.


    AOL9 uses .art files substituted for .jpg files, by default, if you have
    dial-up. If .art is a good compression scheme, it certainly isn't being
    used right or the same way in AOL.

    I don't know if they substitute all JPEGs on all types of pages, but
    they do in their internal stuff.
    --

    <>>< ><<> ><<> <>>< ><<> <>>< <>>< ><<>
    John P Sheehy <>
    ><<> <>>< <>>< ><<> <>>< ><<> ><<> <>><
    , Jul 18, 2004
    #10
  11. Richard Ballard

    Matt Guest

    On 17 Jul 2004 13:37:36 GMT, (Richard Ballard)
    wrote:

    >In article <xy8Kc.1545$>,
    >"David J Taylor"
    ><-this-bit.nor-this-part.uk> writes:
    >
    >>"Richard Ballard" <> wrote in message
    >>news:...
    >>
    >>>"Ancient AOL" used a version of Microsoft Corporation's Internet
    >>>Explorer as its Web browser. ART is an image compression algorithm
    >>>implemented in Microsoft Corporation's Internet Explorer.

    >>
    >>Thanks, Richard. I knew about the AOL / MS-IE link, but I think
    >>ART predated it. Unfortunately I can't quickly find any references
    >>to confirm this.

    >
    >ART is not included in the image processing software packages
    >I am familiar with. I wonder if ART is a Microsoft Corporation
    >proprietary algorithm?


    Best I found so far...
    http://members.aol.com/oldnelly77/page9.htm
    (Converting from ART, and a mention of ONE program)

    http://members.aol.com/prncssvi/aol/artht.html
    How bad?


    --
    I may be dozzzy, but take the ZZZ's out to mail me
    http://www.junkroom.freeserve.co.uk/jvc2080.htm - 2x2x24 CD-RW troubles

    If you drop a cactus, don't try to catch it!
    Matt, Jul 20, 2004
    #11
    1. Advertising

Want to reply to this thread or ask your own question?

It takes just 2 minutes to sign up (and it's free!). Just click the sign up button to choose a username and then you can ask your own questions on the forum.
Similar Threads
  1. Glenn Nelson

    Kodak DX 4530 Compression Algorithm Question

    Glenn Nelson, Sep 24, 2003, in forum: Digital Photography
    Replies:
    4
    Views:
    448
    Ron Hunter
    Sep 24, 2003
  2. john chapman

    Best Image -- Image Size vs Compression

    john chapman, Aug 4, 2004, in forum: Digital Photography
    Replies:
    5
    Views:
    708
    john chapman
    Aug 9, 2004
  3. arcade
    Replies:
    0
    Views:
    427
    arcade
    Nov 30, 2006
  4. arcade
    Replies:
    1
    Views:
    634
    Paul Heslop
    Nov 30, 2006
  5. Andrew Mowat
    Replies:
    0
    Views:
    601
    Andrew Mowat
    Sep 14, 2004
Loading...

Share This Page