Re: JPEG file internal structure

Discussion in 'Digital Photography' started by kbhat@kaxy.com, Sep 3, 2007.

  1. Guest

    On Aug 31, 10:19 am, wrote:
    > Hi,
    >
    > I am looking for some information on the internal structure of a JPEG
    > file. I was playing with a handful of bits in a JPEG file as part of
    > a steganagraphy experiment and I saw something remarkable. I am able
    > to cause a lateral shifting in the file. It is as if someone has torn
    > the right quarter of the picture and pasted it on the extreme left. I
    > was certainly expecting loss of clarity in the picture, but seeing
    > such a remarkable phenomenon has left me completely dumbstruck. Any
    > suggestions what might be causing this?
    >
    > Thanks,
    > Bhat



    Reposting ....
     
    , Sep 3, 2007
    #1
    1. Advertising

  2. wrote:
    > On Aug 31, 10:19 am, wrote:
    >> I am looking for some information on the internal structure of a JPEG
    >> file.

    > Reposting ....


    Did you check http://jpeg.org?

    jue
     
    Jürgen Exner, Sep 3, 2007
    #2
    1. Advertising

  3. fup2 r.p.d; best group is probably comp.compression, though

    :

    >> I am looking for some information on the internal structure of a JPEG
    >> file. I was playing with a handful of bits in a JPEG file as part of
    >> a steganagraphy experiment and I saw something remarkable. I am able
    >> to cause a lateral shifting in the file. It is as if someone has torn
    >> the right quarter of the picture and pasted it on the extreme left. I
    >> was certainly expecting loss of clarity in the picture, but seeing
    >> such a remarkable phenomenon has left me completely dumbstruck. Any
    >> suggestions what might be causing this?


    If part of the bitstream is corrupted, the decoder takes up decoding
    at the next restart marker or wherever the bitstream makes sense to it
    again. Decoding resumes at an incorrect position because the decoder
    does not know how many pixels (MCUs) it skipped over. Hence,
    everything is shifted.

    Regards,
    Marco
     
    Marco Schmidt, Sep 4, 2007
    #3
  4. > If part of the bitstream is corrupted, the decoder takes up decoding
    > at the next restart marker or wherever the bitstream makes sense to it
    > again. Decoding resumes at an incorrect position because the decoder
    > does not know how many pixels (MCUs) it skipped over. Hence,
    > everything is shifted.


    assuming the bitstream is corrupted at only one spot, couldn't the decoder
    calculate an appropriate amount of padding after the entire bitstream is
    decoded?

    --
    Bx.C
     
    Bx.C / x87asm, Sep 5, 2007
    #4
  5. Bx.C / x87asm:

    >assuming the bitstream is corrupted at only one spot, couldn't the decoder
    >calculate an appropriate amount of padding after the entire bitstream is
    >decoded?


    Yes, that is possible. However, I doubt that any tool actually does
    that.

    Regards,
    Marco
     
    Marco Schmidt, Sep 8, 2007
    #5
    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. Martik

    DVD recorder file structure

    Martik, Sep 11, 2005, in forum: Computer Support
    Replies:
    4
    Views:
    1,973
    Martik
    Sep 11, 2005
  2. David

    Deleting Old File Structure

    David, Dec 1, 2006, in forum: Computer Information
    Replies:
    0
    Views:
    382
    David
    Dec 1, 2006
  3. fuji frontier 340 file structure question

    , Nov 6, 2005, in forum: Digital Photography
    Replies:
    2
    Views:
    611
    Mike Warren
    Nov 6, 2005
  4. Amit
    Replies:
    3
    Views:
    1,279
    Ed Ruf (REPLY to E-MAIL IN SIG!)
    Mar 17, 2006
  5. External Drives: File structure

    , Nov 22, 2006, in forum: Computer Support
    Replies:
    10
    Views:
    570
    philo
    Nov 23, 2006
Loading...

Share This Page