Hi marcus,
i tried your xml code and it loaded just fine.
Now i can scale gamma values just working on the
xml values. My problem starts when i try try to
use the grid thing.
For what i could understand reading the OSG code,
the grid is passed to the servers as a texture coordinate
4 the GrabBackground textures.so i assume that having this
XML the mapping is:
<grid>
<row><point>0.0 0.0</point><point>0.5 0.0</point><point> 1.0
0.0</point></row>
<row><point>0.0 0.5</point><point>0.5 0.5</point><point>1.0
0.5</point></row>
<row><point>0.0 1.0</point><point>0.5 1.0</point><point>1.0
1.0</point></row>
</grid>
1,0 1,0.5 1,1
*--------*---------*
| | |
| | |
| | |
*--------*---------*
| | |
| | |
| | |
*--------*---------*
0,0 0,0.5 0,1
or at least i hope so. Now, if i alter one of the values like:
(the 0.3 in the lower middle value)
<grid>
<row><point>0.0 0.0</point><point>0.5 0.3</point><point> 1.0
0.0</point></row>
<row><point>0.0 0.5</point><point>0.5 0.5</point><point>1.0
0.5</point></row>
<row><point>0.0 1.0</point><point>0.5 1.0</point><point>1.0
1.0</point></row>
</grid>
i'd expect:
*\ / *
| \ / |
| \ * / |
| | |
*--------*---------*
| | |
| | |
| | |
*--------*---------*
where the middle point is lowered by 0.3 texture units in the vertical
axis and is in the same texture unit in the horizontal.
instead i get the full quad a weirdly distorted.
If i change the coordinate to -0.3 (i know it's absurd) i get something
like:
*\ /*
| \ / |
| \ *-----*/ |
| | |
*---------*---------*
| | |
| | |
| | |
*---------*---------*
so the texture coordinate appear duplicated and translated.
I looked in the OSGDisplayCalibration.cpp file, and tried
both absolute and relative tex coords, but they work
just like the same... Is it a bug or there is still something
that i can't understand?
I know that my ascii art isn't really good... so
if u can't understand, i'll manage to put some screenshots
online to give you a better view of the situation.
Thanks a lot,
enrico
Il ven, 2005-05-20 alle 12:24, Marcus Roth ha scritto:
> Hi,
>
> try to use the following calibration file. You should see some
> distortions and the colors red and blue should be swapped.
> Change m1 with your cluster server name.
>
> <?xml version="1.0"?>
> <displaycalibration>
> <server name="m1">
> <colormatrix>
> 0 1 0 0
> 1 0 0 0
> 0 0 1 0
> 0 0 0 1
> </colormatrix>
> <gamma>
> 1
> </gamma>
> <gammaramp>
> <color> 0 0 0</color>
> <color> 1 1 1</color>
> </gammaramp>
> <grid>
> <row><point>0 0</point><point>200 0</point><point>399 0</point></row>
> <row><point>0 150</point><point>200 150</point><point>399 150</point></row>
> <row><point>0 299</point><point>200 299</point><point>399 299</point></row>
> </grid>
> </server>
> </displaycalibration>
>
>
> Marcus
>
> Enrico Borrione wrote:
>
> >Hi marcus,
> >i tryied to make some experiment with the display calibration,
> >as you suggested, but i can't get on top of it.
> >Using the source files and some examples i've found,
> >i tried to run a simple cluster, with one client and
> >two servers. To set up the calibration i use this xml file:
> >
> ><?xml version="1.0"?>
> ><displaycalibration>
> ><server name="test1">
> > <gamma> 0.5f </gamma>
> ></server>
> ><server name="test2">
> ></server>
> ></displaycalibration>
> >
> >as you can see i try just to alter the gamma correction
> >of one of the servers, while the other remains unchanged.
> >I load this via:
> >
> > std::filebuf fb;
> > fb.open ("test_due.xml",std::ios::in);
> > std::istream os(&fb);
> >
> >
> > mwin->loadCalibration(os);
> >
> >
> >//DisplayCalibrationPtr calibPtr = DisplayCalibration::create();
> >//
> >mwin->getCalibration()[0]->calibrate(mwin,(RenderAction*)mgr->getAction());
> > std::cout << "GAMMA1 " << mwin->getCalibration()[0]->getGamma() <<
> >std::endl;
> > fb.close();
> >
> > and for what i can see the xml is correctly
> > parsed. (the Gamma corresponds).
> >The problem is that the corrisponding server window remains all black,
> >no matter what parameter i pass in it. Why? What's wrong?
> >I tried also a more direct approach with:
> >
> >mwin->getCalibration()[0]->calibrate(mwin,(RenderAction*)mgr->getAction());
> >
> >but it just segmentation faults (i admit i haven't experimented much
> >this second way).
> >
> >Any suggestion? thanks Enrico
> >
> >Il ven, 2005-05-13 alle 14:49, Marcus Roth ha scritto:
> >
> >
> >>Hi,
> >>
> >>it ist possible to render the contents of a cluster server into a
> >>texture and the do color correction and
> >>geometrical correction to this texture. This is done by the
> >>DisplayCalibration class.
> >>
> >>A clusterwindow has a multi field clusterCalibration. This field
> >>can hold references to a DisplayCalibration object. A display
> >>calibration can hold a grid that
> >>can be used to do distortion.
> >>
> >>Marcus
> >>
> >>Enrico Borrione wrote:
> >>
> >>
> >>
> >>>Hi Dirk,
> >>>i am still struggling after the second viewpor problem.
> >>>As you may remember my final goal is to make it run on
> >>>my cluster, to simulate the geometric distortion of a
> >>>cilindrical screen and the ovaerlapping between the projections.
> >>>
> >>>In order to achieve this objective, i was thinking about
> >>>adding the second viewport trick to my cluster servers,
> >>>so half of the application works as a slave of the client
> >>>(rendering its part of the scene), and half of the app
> >>>uses the obtained buffer to texture the target mesh and
> >>>obtaining my desired effect.
> >>>I did a little experiment of mixing two programs (the
> >>>render to texture and cluster server) but i can't get
> >>>but a crash after another. In a few words I create
> >>>a second viewport, add a TextureGrabBackground and
> >>>add all this to the main window of the server:
> >>>
> >>>winid = glutCreateWindow(name);
> >>>
> >>>window = GLUTWindow::create();
> >>> window->setId(winid);
> >>> window->init();
> >>>.
> >>>.
> >>>.
> >>>
> >>>server = new ClusterServer(window,name,connectionType,"");
> >>> // start the server
> >>> server->start();
> >>>.
> >>>.
> >>>.
> >>>
> >>>ViewportPtr vp2 = Viewport::create();
> >>>beginEditCP(vp2);
> >>> vp2->setCamera(cam);
> >>> vp2->setRoot(tr);
> >>> vp2->setBackground(tg);
> >>>
> >>>
> >>> vp2->setLeft(0);
> >>> vp2->setRight(1);
> >>> vp2->setBottom(0);
> >>> vp2->setTop(1);
> >>>endEditCP(vp2);
> >>>
> >>>
> >>>beginEditCP(window);
> >>> window->addPort(vp2);
> >>>endEditCP(window);
> >>>
> >>>
> >>>I noticed that the program crashes as soon as i add the second
> >>>viewport to the window.
> >>>Probably i am misunderstanding something crucial in the
> >>>rendering pipeline of OSG :)
> >>>Does anyone has some clue to offer?
> >>>
> >>>Many thanks,
> >>>enrico
> >>>
> >>>Il lun, 2005-05-09 alle 21:59, Dirk Reiners ha scritto:
> >>>
> >>>
> >>>
> >>>
> >>>> Hi Enrico,
> >>>>
> >>>>On Fri, 2005-05-06 at 10:54 +0200, Enrico Borrione wrote:
> >>>>
> >>>>
> >>>>
> >>>>
> >>>>>Hi dirk,
> >>>>>after playing with the example you pointed me for a bit,
> >>>>>i understood the basic principles...
> >>>>>Now I am trying to change the textured quad in the
> >>>>>second viewport with a simple geometry (as you suggested).
> >>>>>So... I modelled a half tube with blender,
> >>>>>exported it in wrml along with the texture coordinates
> >>>>>and tried to use it as the *screen* of the application.
> >>>>>Surprise surprise, it isn't working, i just see the
> >>>>>solid background. My best guess is that the matrixcameraptr
> >>>>>is pointing in the wrong direction, so i pointlessy looked for a
> >>>>>way to move the camera.
> >>>>>I used to query the simple scene manager for the navigator
> >>>>>and then update the projection matrix all by myself, but
> >>>>>in this case i think the navigator is linked to the first vieport...
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>Yup.
> >>>>
> >>>>
> >>>>
> >>>>
> >>>>
> >>>>>So: how can i update the camera position of the second vp?
> >>>>>and the target of the look at? :)
> >>>>>
> >>>>>
> >>>>>
> >>>>>
> >>>>The easiest would probably be to create your own Camera and use that in
> >>>>the second viewport, which should work just fine.
> >>>>
> >>>> Dirk
> >>>>
> >>>>
> >>>>
> >>>>
> >>>
> >>>--
> >>>This SF.Net email is sponsored by Oracle Space Sweepstakes
> >>>Want to be the first software developer in space?
> >>>Enter now for the Oracle Space Sweepstakes!
> >>>http://ads.osdn.com/?ad_id=7393&alloc_id=16281&op=click
> >>>_______________________________________________
> >>>Opensg-users mailing list
> >>>[email protected]
> >>>https://lists.sourceforge.net/lists/listinfo/opensg-users
> >>>
> >>>
> >>>
> >>>
> >>
> >>--
> >>This SF.Net email is sponsored by Oracle Space Sweepstakes
> >>Want to be the first software developer in space?
> >>Enter now for the Oracle Space Sweepstakes!
> >>http://ads.osdn.com/?ad_id=7393&alloc_id=16281&op=click
> >>_______________________________________________
> >>Opensg-users mailing list
> >>[email protected]
> >>https://lists.sourceforge.net/lists/listinfo/opensg-users
> >>
> >>
> >>
> >>
> >
> >
> >
> >--
> >This SF.Net email is sponsored by Oracle Space Sweepstakes
> >Want to be the first software developer in space?
> >Enter now for the Oracle Space Sweepstakes!
> >http://ads.osdn.com/?ad_id=7412&alloc_id=16344&op=click
> >_______________________________________________
> >Opensg-users mailing list
> >[email protected]
> >https://lists.sourceforge.net/lists/listinfo/opensg-users
> >
> >
>
>
>
> --
> This SF.Net email is sponsored by Oracle Space Sweepstakes
> Want to be the first software developer in space?
> Enter now for the Oracle Space Sweepstakes!
> http://ads.osdn.com/?ad_id=7412&alloc_id=16344&op=click
> _______________________________________________
> Opensg-users mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/opensg-users
>
>
-------------------------------------------------------
This SF.Net email is sponsored by Yahoo.
Introducing Yahoo! Search Developer Network - Create apps using Yahoo!
Search APIs Find out how you can build Yahoo! directly into your own
Applications - visit http://developer.yahoo.net/?fr=offad-ysdn-ostg-q22005
_______________________________________________
Opensg-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/opensg-users
