Subject:
|
Re: LDGLite bug report (and POV preview question)
|
Newsgroups:
|
lugnet.cad.dev
|
Date:
|
Thu, 6 Sep 2001 16:06:12 GMT
|
Viewed:
|
630 times
|
| |
|
|
In lugnet.cad.dev, Lars C. Hassing writes:
> If you want to make a previewer for finding a good camera setting
> you should adjust the camera location and look_at positions.
> When done, specify these two coordinates to L3P using -cc and -cla.
I'm not sure if I *want* to do this, I was just asking if it would
be useful. I asked because I already added -cc and -ca (not sure
what -cla is) to ldglite so I could move the camera enough to test
my recent bug fixes. Since I tend to add things to the code pretty
much at random, I thought I'd see if this was worth pursuing before
I get distracted by some other neat idea.
> As mentioned at the introduction of L3Lab (Feb 2000), my plan is to
> enable you to rotate/pan/zoom the scene until you have a good view,
> and then generate the POV code. I believe LeoCAD is capable of this.
> I still need to figure out the best user-interface for rotate/pan/zoom
> and I need to find the time...
Yeah, I played with this in ldglite at one time. I found some nice
quaternion camera moving code and pasted it in. I kept it separate
from the ldraw view matrix though because at that time I didn't know
what to do about the shearing in the default view.
This is the interface I ended up with. It needs work, but I like the
fact that it's modeless. And I'm a big fan of the keyboard. :)
The mouse spins the model about its origin.
Arrow keys turn the viewpoint camera 5 degrees left, right, up, down.
Alt Arrow keys roll the camera left, right, up, down.
Ctrl Arrow keys move the camera left, right, up, down.
PGUP and PGDN move the camera in and out.
> > Anyways, I suspect I have just about everything I need built in with
> > the L3 parser. The only iffy part is the auto-centering part of L3P.
> > I'm not sure exactly what coordinates that uses: the bounding box, or
> > something else?
>
> The automatic camera positioning algorithm is rather complex and involves
> solving six equations with six unknowns.
> The basic idea is described in
> http://home16.inet.tele.dk/hassing/l3p.html#l3globe
Ahh! The auto-position is based on the intersection of the first level
bounding boxes and the view frustrum. That helps. I'm not sure I'm
smart enough to figure out what the 6 unknowns are. The frustrum has
4 sides and you could add front and back clip planes. Gotta think
about this...
Don
|
|
Message has 1 Reply:
 | | Re: LDGLite bug report (and POV preview question)
|
| Should we change the default view to be something that is (what is the word..."only made from rotation matricies and does not produce shearing"? det(A)=1?). It bugs me that the default is flawed that way, although there are some other (...) (23 years ago, 6-Sep-01, to lugnet.cad.dev)
|
Message is in Reply To:
| | Re: LDGLite bug report (and POV preview question)
|
| Don Heyse wrote... (...) ;-) (...) If you want to make a previewer for finding a good camera setting you should adjust the camera location and look_at positions. When done, specify these two coordinates to L3P using -cc and -cla. As mentioned at the (...) (23 years ago, 4-Sep-01, to lugnet.cad.dev)
|
21 Messages in This Thread:
  
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
This Message and its Replies on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
