Yeah I see (because as I remember now in Away Lite it is Vector3D instead
Number3d.)BUT my misunderstanding here is the following:
...var faces:Vector.<Number3D>;

and
... intersectP = _ray.getIntersect(_dir, pixel, faces[l], faces[l+1],
faces[l+2]);

Do you convert the faces v0,v1,v2 positions to Number3D before that
function? . Because here your faces already represented as Array of
Number(s)3D ?

In my  case i was doing :
_ray.getIntersect(_ray.dir,_ray.orig,_sf.faces[i].v0.position,_sf.faces[i].v1.position,_sf.faces[i].v2.position);





On Wed, Aug 25, 2010 at 5:52 PM, Fabrice3D <[email protected]> wrote:

> no, but should be like directly usuable in there with no or minor changes.
> Its just an helper class not dependent of anything.
>
> On Aug 25, 2010, at 4:47 PM, Michael Iv wrote:
>
> Was that Away3DLite implementation ?
>
> On Wed, Aug 25, 2010 at 5:40 PM, Michael Iv <[email protected]> wrote:
>
>> Thanks a lot Fabrice for your help, i will check this out. !
>>
>>
>> On Wed, Aug 25, 2010 at 5:34 PM, Fabrice3D <[email protected]> wrote:
>>
>>> here a ruw sample from the quick demo I've sended you earlyer.
>>> where white was no hit, black hit. Say very simple primary ray raytracer.
>>>
>>> private function someImage():void
>>> {
>>>  _tracemap.lock();
>>>  var intersectP:Number3D;
>>>   var i:int;
>>> var j:int;
>>> var k:int;
>>> var l:int;
>>>  var r:int = 255;
>>> var g:int = 255;
>>> var b:int = 255;
>>>
>>> var startx:int = _view.x-Math.round(_width*.5);
>>> var starty:int =_view.y-Math.round(_height*.5)
>>>  var pixel:Number3D;
>>>  var meshlength:int = _meshes.length;
>>>  var faces:Vector.<Number3D>;
>>> var faceslength:int;
>>>  for(i = 0; i<_width;++i){
>>> startx++;
>>> for(j = 0; j<_height;++j){
>>>  pixel = _view.camera.unproject(startx, starty+j);
>>>
>>> _dir.x = pixel.x;
>>>  _dir.y = pixel.y;
>>> _dir.z = pixel.z;
>>> _dir.normalize();
>>>
>>> for(k = 0; k<meshlength;++k){
>>>  faces = _meshes[k];
>>>  faceslength = faces.length;
>>>  for(l = 0; l<faceslength;l+=3){
>>>  intersectP = _ray.getIntersect(_dir, pixel, faces[l], faces[l+1],
>>> faces[l+2]);
>>> RayDebugger.traceRayFromDir(pixel, _dir, 2500);
>>>  if (intersectP != null)
>>> _tracemap.setPixel(i, j,  r << 16 | g  << 8 | b );
>>> }
>>>  }
>>> }
>>> }
>>>  _tracemap.unlock();
>>>
>>> [... some uninterresting code...]
>>>
>>> Fabrice
>>>
>>>
>>>
>>> On Aug 25, 2010, at 4:06 PM, Michael Iv wrote:
>>>
>>> Got you , no I haven't seen the FITC presentation yet . By the way ,still
>>> can't get ray.getIntersect() returning the data besides null.
>>> If you could show some source code if a test run it could be
>>> nice.Although I am almost sure I do it just as it has to be .
>>>
>>> Thanks a lot .
>>> Michael.
>>>
>>> On Wed, Aug 25, 2010 at 5:00 PM, Fabrice3D <[email protected]> wrote:
>>>
>>>> As usual works just fine for me :))
>>>> I haven't added it exactly for what I suppose you want to do.
>>>> It was added to allow the retrieving of uv's,map,meshname from a hittest
>>>> result performed by the BSP.
>>>> Look at BSPHittest code if you want to see how the class was used.
>>>> (If you have followed the livestream from FITC where Rob showed the FPS
>>>> demo we've builded, then you know now how the rockets were builded.)  :)
>>>>
>>>> If you ensure that the position you test on is a result of a plane/ray
>>>> hitest where the position is on the "plane" of the triangle
>>>> then it should result in what you need: that is being used as test point
>>>> in triangle. UV returned would be null if not inside.
>>>>
>>>> I wouldn't be honest not adding that I think very soon make some speed
>>>> compairs with the ray intersect implementation I have in Prefab replaced
>>>> with this class...
>>>> so the class might evolve bits.
>>>>
>>>> Fabrice
>>>>
>>>>
>>>> On Aug 25, 2010, at 3:12 PM, Michael Iv wrote:
>>>>
>>>> hey Fabrice , is BaryCentricTest is like Ray.getIntersect() but  for UV
>>>> coords >?
>>>>
>>>> Doest it work fine ? (Unlike Ray.getIntersect() ;)  );
>>>>
>>>> BaryCentricTest
>>>>
>>>>
>>>> On Wed, Aug 25, 2010 at 3:33 PM, Fabrice3D <[email protected]> wrote:
>>>>
>>>>>
>>>>> http://code.google.com/p/away3d/source/browse/#svn/trunk/fp10/Away3D/src/away3d/tools/utils
>>>>>
>>>>> here you go.
>>>>>
>>>>> Fabrice
>>>>> On Aug 25, 2010, at 2:18 PM, Bobby wrote:
>>>>>
>>>>> > Hi Edwin,
>>>>> >      I am using the away3d_2_5_0 version. I have the away3d_3_5_0
>>>>> > also with me. But in both these I cannot find utils folder with the
>>>>> > tools package. How can I overcome this error. I am using the
>>>>> > downloaded zip files from the Away 3D site. I am not using SVN. Is
>>>>> > that causing the problem. CAN ANYONE HELP ME !!!!!!!!!!!!!!
>>>>> >
>>>>> > Thanks in Advance
>>>>> > Bobby Francis Joseph
>>>>>
>>>>>
>>>>
>>>>
>>>> --
>>>> Michael Ivanov ,Programmer
>>>> Neurotech Solutions Ltd.
>>>> Flex|Air |3D|Unity|
>>>> www.neurotechresearch.com
>>>> http://blog.alladvanced.net
>>>> http://www.meetup.com/GO3D-Games-Opensource-3D/
>>>> Tel:054-4962254
>>>> [email protected]
>>>> [email protected]
>>>>
>>>>
>>>>
>>>
>>>
>>> --
>>> Michael Ivanov ,Programmer
>>> Neurotech Solutions Ltd.
>>> Flex|Air |3D|Unity|
>>> www.neurotechresearch.com
>>> http://blog.alladvanced.net
>>> http://www.meetup.com/GO3D-Games-Opensource-3D/
>>> Tel:054-4962254
>>> [email protected]
>>> [email protected]
>>>
>>>
>>>
>>
>>
>> --
>> Michael Ivanov ,Programmer
>> Neurotech Solutions Ltd.
>> Flex|Air |3D|Unity|
>> www.neurotechresearch.com
>> http://blog.alladvanced.net
>> http://www.meetup.com/GO3D-Games-Opensource-3D/
>> Tel:054-4962254
>> [email protected]
>> [email protected]
>>
>>
>
>
> --
> Michael Ivanov ,Programmer
> Neurotech Solutions Ltd.
> Flex|Air |3D|Unity|
> www.neurotechresearch.com
> http://blog.alladvanced.net
> http://www.meetup.com/GO3D-Games-Opensource-3D/
> Tel:054-4962254
> [email protected]
> [email protected]
>
>
>


-- 
Michael Ivanov ,Programmer
Neurotech Solutions Ltd.
Flex|Air |3D|Unity|
www.neurotechresearch.com
http://blog.alladvanced.net
http://www.meetup.com/GO3D-Games-Opensource-3D/
Tel:054-4962254
[email protected]
[email protected]

Reply via email to