Finetuning problem for scrolling panorama installation


    Sep 18 2006 | 7:14 pm
    Hello all,
    I'm looking for a way to correct a little error that is occurring in
    an installation.
    The movielink below is a registration of this installation, the error
    is visible about 3/4ths into the movie.
    Look for : the little bar on top going away, leaving just the big
    panorama in the background visible. The scrolling goes smooth except
    for a little hickup
    on the right side, where one panel of the panorama flashes on top
    before going away again.
    Below is a part of the patch for this installation. The problem is
    this : I move 4 panels left>right or right>left containing part of
    the panorama. Once they are offscreen, they are moved to the other
    side to create endless scrolling.
    The way these panels move is regulated by the sensor in the rotating
    chair. Since this sensor detects a given number of concrete steps in
    the rotation, I interpolate the motion of the panels using a little
    abstraction I made called 'bounce' (also below).
    Now, this 'bounce' abstraction uses the bline object, this all works
    well to move the panels normally, but when whey have to move to the
    other end after going offscreen, the bline object would move them
    from say point -100 to 100, going through every step in between and
    thus passing through the projected image. Since this is not desired,
    I detect when the pictures x-position is > than a certain nr. and if
    so, set the new position to bline "immediately".
    I don't know where something exactly goes wrong, whether it is in the
    'immediate' setting of the new location or with the detection which
    opens the gate to send the location to 'immediate' or not. I just
    thought I'd post it.. perhaps someone has an "aaah that's why"-
    moment... since all those I had didn't solve it.
    I'm afraid that this might all still be a bit vague but I can only
    ask and cross my fingers :D
    I've sent the patches in a new mail.
    Rgds,
    Brecht.

    • Sep 18 2006 | 7:48 pm
      I havent seen the patches (did they bounce), but that looks really
      nice. Id love to try it myself.
      Im assuming you are using openGL for this (it looks quite smooth)? if
      the panels are set to automatic 0, maybe simply stop banging them
      until the bline has finished placing them?
      v a d e //
      www.vade.info
      abstrakt.vade.info
      On Sep 18, 2006, at 3:14 PM, Brecht Debackere wrote:
      > Hello all,
      >
      > I'm looking for a way to correct a little error that is occurring
      > in an installation.
      > The movielink below is a registration of this installation, the
      > error is visible about 3/4ths into the movie.
      > Look for : the little bar on top going away, leaving just the big
      > panorama in the background visible. The scrolling goes smooth
      > except for a little hickup
      > on the right side, where one panel of the panorama flashes on top
      > before going away again.
      >
      > http://www.marnixdenijs.nl/downloads/Beijing_Accelerator_TENT_WEB.mov
      >
      > Below is a part of the patch for this installation. The problem is
      > this : I move 4 panels left>right or right>left containing part of
      > the panorama. Once they are offscreen, they are moved to the other
      > side to create endless scrolling.
      >
      > The way these panels move is regulated by the sensor in the
      > rotating chair. Since this sensor detects a given number of
      > concrete steps in the rotation, I interpolate the motion of the
      > panels using a little abstraction I made called 'bounce' (also below).
      > Now, this 'bounce' abstraction uses the bline object, this all
      > works well to move the panels normally, but when whey have to move
      > to the other end after going offscreen, the bline object would move
      > them from say point -100 to 100, going through every step in
      > between and thus passing through the projected image. Since this is
      > not desired, I detect when the pictures x-position is > than a
      > certain nr. and if so, set the new position to bline "immediately".
      >
      > I don't know where something exactly goes wrong, whether it is in
      > the 'immediate' setting of the new location or with the detection
      > which opens the gate to send the location to 'immediate' or not. I
      > just thought I'd post it.. perhaps someone has an "aaah that's why"-
      > moment... since all those I had didn't solve it.
      > I'm afraid that this might all still be a bit vague but I can only
      > ask and cross my fingers :D
      >
      > I've sent the patches in a new mail.
      >
      > Rgds,
      > Brecht.
      >
    • Sep 18 2006 | 8:07 pm
      Yep. They bounced. I've put the text here : http://
      www.visualantics.net/BA_accelerator.txt
      It is indeed open GL, current resolution of the photos is 1024x768,
      but 1600x1200 worked smoothly as well (I will send the specs of the
      PC on which it is running, otherwise "smooth" is a pretty vague
      notion :) ) . Your solution might just as well allready work
      though... it'd require some repatching, but it's a good way to do it.
      I was thinking of using the layer possibilities in 1.6, but
      unfortunately it's running on PC so no such luck.
      grtz.
      Brecht.
      On Sep 18, 2006, at 9:48 PM, vade wrote:
      > I havent seen the patches (did they bounce), but that looks really
      > nice. Id love to try it myself.
      >
      > Im assuming you are using openGL for this (it looks quite smooth)?
      > if the panels are set to automatic 0, maybe simply stop banging
      > them until the bline has finished placing them?
      >
      >
      >
      >
      > v a d e //
      >
      > www.vade.info
      > abstrakt.vade.info
      >
      > On Sep 18, 2006, at 3:14 PM, Brecht Debackere wrote:
      >
      >> Hello all,
      >>
      >> I'm looking for a way to correct a little error that is occurring
      >> in an installation.
      >> The movielink below is a registration of this installation, the
      >> error is visible about 3/4ths into the movie.
      >> Look for : the little bar on top going away, leaving just the big
      >> panorama in the background visible. The scrolling goes smooth
      >> except for a little hickup
      >> on the right side, where one panel of the panorama flashes on top
      >> before going away again.
      >>
      >> http://www.marnixdenijs.nl/downloads/Beijing_Accelerator_TENT_WEB.mov
      >>
      >> Below is a part of the patch for this installation. The problem is
      >> this : I move 4 panels left>right or right>left containing part of
      >> the panorama. Once they are offscreen, they are moved to the other
      >> side to create endless scrolling.
      >>
      >> The way these panels move is regulated by the sensor in the
      >> rotating chair. Since this sensor detects a given number of
      >> concrete steps in the rotation, I interpolate the motion of the
      >> panels using a little abstraction I made called 'bounce' (also
      >> below).
      >> Now, this 'bounce' abstraction uses the bline object, this all
      >> works well to move the panels normally, but when whey have to move
      >> to the other end after going offscreen, the bline object would
      >> move them from say point -100 to 100, going through every step in
      >> between and thus passing through the projected image. Since this
      >> is not desired, I detect when the pictures x-position is > than a
      >> certain nr. and if so, set the new position to bline "immediately".
      >>
      >> I don't know where something exactly goes wrong, whether it is in
      >> the 'immediate' setting of the new location or with the detection
      >> which opens the gate to send the location to 'immediate' or not. I
      >> just thought I'd post it.. perhaps someone has an "aaah that's
      >> why"-moment... since all those I had didn't solve it.
      >> I'm afraid that this might all still be a bit vague but I can only
      >> ask and cross my fingers :D
      >>
      >> I've sent the patches in a new mail.
      >>
      >> Rgds,
      >> Brecht.
      >>
      >