Hi, Eric
thank you very much for your answer.
I agree whith you that I will always have this stepping effects.
But I mean, that my image must looks at high resolution better.
I mean that Canvas3D must take the configuration of my graphics hardware
from:
GraphicsConfigTemplate3D gct3d =
Hey Eric and Olga,
This 'stepping' is known as aliasing, or more fondly as 'jaggies'. There are
blending techniques to reduce the effect, called anti-aliasing. This is an
expensive operation and most hardware does not support it yet.
This is really a problem with drawing lines or edges that
Andreas,
The solution to your problem is to extend the canvas and override two
methods which control the size of the canvas. (This applies to Canvas and
Canvas3D). I'm not sure if this is a lightweight/heavyweight problem or a
problem with layout.
Add the following code to your canvas subclass
Hi Andrew,
Andrew Phelps wrote:
> The solution to your problem is to extend the canvas and override two
> methods which control the size of the canvas. (This applies to Canvas and
> Canvas3D). I'm not sure if this is a lightweight/heavyweight problem or a
> problem with layout.
> Add the follo
hello,
In the right side you put a JPanel and inside the Canvas3D and you have
to set minimumSize to the Jpanel
olivier Fabregue
[EMAIL PROTECTED]
===
To unsubscribe, send email to [EMAIL PROTECTED] and include in the body
Andreas,
I understood your problem.
It is not the splitframe that you need worry about. The splitframe is
passing any resize callback to the canvas, but for some reason the canvas's
internal minimum size is being set to the canvas's size every time the
canvas is enlarged, thus preventing you fro
Hi Andrew,
sorry for beeing stupid! :-( I should have first try out what you said.
Andrew Phelps wrote:
> I understood your problem.
> It is not the splitframe that you need worry about. The splitframe is
> passing any resize callback to the canvas, but for some reason the canvas's
> internal
Ouch that sounds like a bad one, glad to hear you guys are hard at work. In
my case, if I remember it correctly, I did not have a transform group. In
this case I just had a bunch of shapes under a BranchGroup, and I was
detaching the BranchGroup. I solved the problem by turning off compilation
A while back I fought with the settings to make transparency textures
work. I could easily place the texture on the polygon, but the places
where the alpha value of the texture was 0, it showed the color of the
polygon. It didn't make sense to me that I could see the underlying
polygon at all.
Hi all,
Can anyone tell me whether or not the following features are supported by
Java3D:
* Collision detection
* Shadowing
* Camera imaging such as:
* simulation of depth of field effects, i.e., blurred effects
when images are out of focus
* simulati
Hmmm,
Two ways:
1) Interesting way but probably not what you want is to set the color of the
vertex to include an alpha component and fade across your vertices, using
modulate with the texture. I know this works in opengl, I assume it does in
Java3d.
2) The next way is to use a transparent pol
Java 3D performs view frustum culling, only sending geometry that is
within the viewing frustum to the graphics accelerator.
We have done research into general purpose occlusion culling, but it
isn't currently in Java 3D (nor is it currently planned for the 1.3
version).
--
Kevin Rushforth
Java
> * Collision detection --> yes. But you might have to do some work
> to handle "preemptive cd", in other words, check for collisions before
> letting the user move instead of waiting for a behavior to trigger with
> the collision event.
>
> * Shadowing -> not real time casted shadows
Hi,
I am trying to fix a directional light source at eyepoint, ie going into screen at all
times.
This is so when I rotate image on screen, light source is still remaining fixed and
always
pointing into screen.
The problem know is directional light source follows the object as I rotate it with
Thanx david !
-->-Original Message-
-->From: Yazel, David J. [mailto:[EMAIL PROTECTED]]
-->Sent: dinsdag 19 september 2000 16:36
-->To: [EMAIL PROTECTED]
-->Subject: Re: [JAVA3D] Advanced Features
-->
-->
-->> * Collision detection --> yes. But you might have
-->to do some work
-->
Thank you!
So to dig a bit further, what is the level of granularity for your fov
culling: Shape, Geometry Array, polygon or vertice? If you have a shape 3d
that has 20k points and one vertice of that overlaps the view frustum, is it
all sent? Are there two stages of culling, one by shape then
Try putting the light in the same transform group as the view, don't put the
light in the transform/branch group of your geometry.
Dave Yazel
> --
> From: Bradley Crouch[SMTP:[EMAIL PROTECTED]]
> Reply To: [EMAIL PROTECTED]
> Sent: Tuesday, September 19, 2000 10:53 AM
How to use quaternion to do rotation about an arbitrary axis in Java3D?
e.g: Quat4f
__
Do You Yahoo!?
Talk to your friends online with Yahoo! Messenger.
http://im.yahoo.com
===
Hello,
Quite a bit of the javax.media.j3d.* and javax.vecmath.* classes would
probably be implemented as native code. Still, is the Java source code for
those classes available anywhere ? Any amount, however small, of
non-native code would be helpful.
Thanks in advance,
Regards,
Manu.
I mentioned that my polygon was already fully transparent to support the
alpha channel texture. The alpha component of the polygon does not seem
to affect the transparency of the texture when the texture has an alpha
channel.
I guess I could also be more specific. When I say "fade-out", I mean
Someone posted once that you cannot rotate around an arbitrary point in
Java3d. You must rotate around the origin and then translate. I have not
personally confirmed this.
Dave Yazel
> --
> From: Siau Tan Long[SMTP:[EMAIL PROTECTED]]
> Reply To: Discussion list for Java 3D
Hi,
One way I can think of is to add the light source and the target transform
group for the behaviors as separate branches of the scene graph. That is,
the light source should not be a direct descendant of the transform group
that is affected by PickRotateBehavior, etc.
I assume that the mouse
Ah! hehe, I thought you meant over distance. I have not "faded out" a
polygon over time. This is what I would try: do not make the polygon
transparent, instead modulate a completely opaque texture on your polygon.
Since we are modulating, changes to the colors of the polygon should mixed
with t
I'm playing with using BillBoard... whenever I set the
bb.setAlignmentAxis(0f, 0f, 1f )
to select the Z-axis, the object DISAPPEARS. If I take the axis a
LITTLE off the Z:
float D = 0.001f;
bb.setAlignmentAxis( 0f, D, Math.sqrt(1f - D*D) );
It works fine. What's wrong with using EXACTLY
Hi Dave
Thanks for the quick reply.. now that is service..
I will try the solution, once again thanks..
Brad
> Try putting the light in the same transform group as the view, don't put
> the light in the transform/branch group of your geometry.
>
> Dave Yazel
==
I will quote someone here:
QUOTE (Hongkun Wang):
"Hi , Dave Yazel
Thanks for your help.
I already resolve the trees rendering. I am using:
(1) Put all trees in a OrderedGroup, and use SharedGroup to save memory.
(2) Figure out all trees according 'Z' coordinate in viewplatform coordinate
SYS.
> From: Norihito Hiruma <[EMAIL PROTECTED]>
> > From Doug Gehringer <[EMAIL PROTECTED]>
> > When making many changes to the scene graph outside of a behavior there
> > is no way for Java3D to know when you are done and Java3D will probably
> > update before you are ready.
> > Stopping/starting t
Manu Eapen Varghese wrote:
>
> Hello,
>
> Quite a bit of the javax.media.j3d.* and javax.vecmath.* classes would
> probably be implemented as native code. Still, is the Java source code for
> those classes available anywhere
http://www.j3d.org/faq/running.html#source
Short answer is - no, and ne
I can't switch to an opaque texture. For example, a texture of a tree
*MUST* have an alpha component. I've faded-out polygons over time
before... no big deal. Just adjust the TransparencyAttributes of the
Appearance.
By the way, I am *modulating* the texture with the object. This allows
for *
Yeppers. Looks like you are on to something here. Sorry I could not have
been more help.
Dave
> --
> From: J. Lee Dixon[SMTP:[EMAIL PROTECTED]]
> Reply To: Discussion list for Java 3D API
> Sent: Tuesday, September 19, 2000 12:59 PM
> To: [EMAIL PROTECTED]
> Subjec
> How to use quaternion to do rotation about an arbitrary axis in Java3D?
> e.g: Quat4f
A good introduction to quaternions is at:
http://www.gamasutra.com/features/programming/19980703/quaternions_01.htm
To rotate about an arbitrary axis, AxisAngle4f is probably a better bet, the
fields of Axis
Warning! I do this quite a bit in one of my Java 3D apps. I found that the
Quat4f get normalized regardless of the initial coordinates I use. So I end
up setting the length of the quaternion after I instantiated it.
Note that *all* vectors can be thought of as having one "end" at the origin
an
Hi Doug,
I don't understand this technique completely. The docs on Behavior.postID() are
very terse.
"Post the specified Id. Behaviors use this method to cause sequential scheduling of
other behavior object"
That's it. Is it a priority system? What number = first? Why does using postId()
> I'm playing with using BillBoard... whenever I set the
>
> bb.setAlignmentAxis(0f, 0f, 1f )
>
> to select the Z-axis, the object DISAPPEARS.
The alignment axis is relative to the eye. The eye looks along the Z-axis, so
that if you align to that the Billboard is always edge-on, which is why it
I have my "avatar" roaming the x-y plane, z being up. When I set the
alignment axis to *nearly" the z-axis, this axis goes through all the
transforms to get it into eye coordinates... which is almost the y-axis
in eye coordinates. I don't believe the alignment axis is relative to
the eye-coordin
I don't think Lee is having a good day, we keep misunderstanding his
scenarios, and then we suggest things he has already tried. :(
> --
> From: J. Lee Dixon[SMTP:[EMAIL PROTECTED]]
> Reply To: Discussion list for Java 3D API
> Sent: Tuesday, September 19, 2000 2:50 P
I'm using Java3D 1.2 (have tried both OpenGL and DirectX), and am having the
following problems with collision detection:
* The time at which a collision is reported is inaccurate (the parts might
be visually apart by a significant amount). What capabilities need to be
enabled to get ACCURATE col
Yeah, I guess. I kinda feel like I'm delving into parts of Java3D that
other ppl aren't using. Oh well, I'm happy that I'm helping to make
Java3D more robust. Although, I am trying to put out a product...
:)
Really need to be able to draw pictures!! ;)
Thanks for all the help & patience...
How can I specify a figure to follow a curve trajectory
between to points?
Thanks in advance.
--
Alejandro Sanchez
Tel (02293)15640639 - 14 de julio 417
Tandil - Buenos Aires - Argentina
===
To unsubscribe, send email to [EM
John,
The answer is you can't.
What I believe most of us are doing in order to have accurate collision
detection is to use picking to detect collisions. For instance in our
application the avatar sends out a pickray in the direction it is
walking *before* it moves to make sure it's a valid move
hi all!!!
Has anyone tried storing a J3d object (Shape3D, for ex) in a relational
database like Oracle8i?
Is it possible, first of all???
Thanx.
Ajit
Apt #2, 911, Franklin Avenue,
Normal, IL 61761,
USA.
(Phone) 001-309-438-3743.
Get free email and a permanent address at http://www.amexmail
> From: "Shawn Kendall" <[EMAIL PROTECTED]>
> I don't understand this technique completely. The docs on Behavior.postID()
> are very terse.
>
> "Post the specified Id. Behaviors use this method to cause sequential
> scheduling of other behavior object"
The basic idea is that Behavior.postID() i
John Wright wrote:
> What I believe most of us are doing in order to have accurate collision
> detection is to use picking to detect collisions. For instance in our
> application the avatar sends out a pickray in the direction it is
> walking *before* it moves to make sure it's a valid move.
Th
Ajit Dharmik wrote:
> Has anyone tried storing a J3d object (Shape3D, for ex) in a
> relational database like Oracle8i?
>
> Is it possible, first of all???
No, it is not possible because you cannot serialise any part of the J3D
scenegraph. You will need to create your own geometry format/table
s
This is a report -- I'm not confident that my first post went through.
(Sorry if you have already seen this.)
Any suggestions would be greatly appreciated. Feel free to e-mail me
directly at [EMAIL PROTECTED] Thanks!
-Original Message-
From: Scott S. Lynde [mailto:[EMAIL PROTECTED]]
Se
The other way to handle this is to use multiresolution grids. You place
your objects into your world and "register" them into your own data
structure. Registering them is basically placing the scene graph "piece"
into the grid (and of course you can use a more 3 dimensional octree for
inside).
Dear Mr. Couch,
Could you give me some collision detection examples, or where can I get source code
about collision detection?
Thanks a lot
Jiang Minhua
- Original Message -
From: "Justin Couch" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Wednesday, September 20, 2000 7:42 AM
S
Canoe wrote:
> Could you give me some collision detection examples, or where can I get source code
>about collision detection?
Unfortunately I don't have any handy and j3d.org doesn't seem to have
any links to tutorials etc. I thought my Raw J3D tutorial had a whole
bunch of code there (I remem
48 matches
Mail list logo