Loading...
> > >

Using DzBox3::getIntersection

OmnifluxOmniflux Posts: 427
in Daz Script Developer Discussion

Not sure if it is the method or my understanding that is incorrect here...

The result appears to always be [0, 0, 0] for the first intersection point with both DS4 and DS6, and the DS4 SDK.

 

In the following example, I expect [[0, 0, -50], [0, 0, 50]]. Am I wrong in expecting this?

var oSettings = DzSettings();
Geometry.getDefaultPrimitiveOptions(Geometry.Cube, oSettings);
var oNode = Geometry.createPrimitive(oSettings);

var oLine = DzLine3([[0, 0, -4000], [0, 0, 4000]]);

var oIntersection = oNode.getWSBoundingBox().getIntersection(oLine);
print (oIntersection.intersects);
print (oIntersection.firstIntersection);
print (oIntersection.lastIntersection);

Executing Script...
true
[0,0,0]
[0,0,50]
Result: 
Script executed in 0 secs 7 msecs.

Comments

  • sidcarton1587sidcarton1587 Posts: 54

    Assuming that the cube is resting on the XY plane at Z=0, centered on the Z axis, that could actually make sense? What's the bounding box for oNode? 

  • OmnifluxOmniflux Posts: 427

    The bounding box in that example is [[-50,0,-50],[50,100,50]].

    Here is another example with the bounding box centered around the world origin, with the same result

    var oSettings = DzSettings();
Geometry.getDefaultPrimitiveOptions(Geometry.Cube, oSettings);
var oNode = Geometry.createPrimitive(oSettings);
oNode.getYPosControl().setValue(-50);

var oLine = DzLine3([[0, 0, -4000], [0, 0, 4000]]);
var oIntersection = oNode.getWSBoundingBox().getIntersection(oLine);

print (oNode.getWSBoundingBox())
print (oIntersection.intersects);
print (oIntersection.firstIntersection);
print (oIntersection.lastIntersection);

    Executing Script...
[[-50,-50,-50],[50,50,50]]
true
[0,0,0]
[0,0,50]
Result: 
Script executed in 0 secs 6 msecs.

    And here is an example with the object bounding box not containing the world origin at all

    var oSettings = DzSettings();
Geometry.getDefaultPrimitiveOptions(Geometry.Cube, oSettings);
var oNode = Geometry.createPrimitive(oSettings);
oNode.setWSPos(DzVec3(150,100,150));

var oLine = DzLine3([[100, 100, -4000], [100, 100, 4000]]);

var oIntersection = oNode.getWSBoundingBox().getIntersection(oLine);

print (oNode.getWSBoundingBox())
print (oIntersection.intersects);
print (oIntersection.firstIntersection);
print (oIntersection.lastIntersection);

    Executing Script...
[[100,100,100],[200,200,200]]
true
[0,0,0]
[100,100,200]
Result: 
Script executed in 0 secs 6 msecs.

    The firstIntersection property is still set to the world origin.

  • sidcarton1587sidcarton1587 Posts: 54

    Sure looks like a bug. 

  • OmnifluxOmniflux Posts: 427

    DzBox3::getIntersection(...) seems like such a useful method I find it hard to believe I am the first one to have an issue with it...

     

    As no one has responded stating that I am trying to use it incorrectly, I have submitted ticket #480447 for the issue.

     

    I do wish there was a similar method for testing a nodes geometry instead of just the bounding box.

  • PraxisPraxis Posts: 274
    edited October 12

    Omniflux said:

    DzBox3::getIntersection(...) seems like such a useful method I find it hard to believe I am the first one to have an issue with it...

    Not the first: Some background info...

    There were 2 old threads on the topic, which seem to no longer exist.

    In DS v4.9.3.33, DzBox3::getIntersection() did not work, because it returned no DzVec3 data - you can see why:

       Boolean : getIntersection( DzLine3 ray, DzVec3 firstIntersection, DzVec3 lastIntersection )

    I see now that since DS 4.14.1.36 it has been improved to this, as you tested:

       Object : getIntersection( DzLine3 ray )

    As no one has responded stating that I am trying to use it incorrectly, I have submitted ticket #480447 for the issue.

    I believe you are using it correctly - I agree a ticket is in order.

    I do wish there was a similar method for testing a nodes geometry instead of just the bounding box.

    Me too.

    I never put in a ticket for the bug in DS v4.9.3.33 because I needed a solution a.s.a.p. - so I wrote RayIntersectTriangle/Facet/AABB/etc. functions in DAZ Script: See attached RayIntersect.txt extracts from my library. The reason I have posted this code is because I've found it surprisingly fast, and has it been good enough for my purposes so far.   I'm sure you can improve the code a lot.

     

    mCasual (of course) has both scripted and plug-in solutions relating to this:

    https://sites.google.com/site/mcasualsdazscripts/mcjcollider-plugin

    https://sites.google.com/site/mcasualsdazscripts9/mcjscriptedcollider

     

    txt
    txt
    RayIntersection.txt
    25K
    Post edited by Praxis on
  • OmnifluxOmniflux Posts: 427

    Praxis said:

    I'm sure you can improve the code a lot.

    This is the funniest thing I've heard in the last decade.

  • PraxisPraxis Posts: 274

    Omniflux said:

    Praxis said:

    I'm sure you can improve the code a lot.

    This is the funniest thing I've heard in the last decade.

    Well...  I wasn't thinking of the algorithms (which aren't  mine), or the vector maths (also not mine), but I thought you may want to replace the for() loops with that funky JavaScript forEach() stuff, etc.

    Then I had a look and found that the code I posted contains no for() loops, and no while() loops.

    Sorry.

    Perhaps you could standardize the capitalization of the symbol names, and/or weed and improve the excessive commentary?  :)

  • surrealsurreal Posts: 198
    edited October 18

    They still exist however they offer no help. I don't recall if I submitted a ticket.

    https://www.daz3d.com/forums/discussion/66054/dzbox3-getintersection-question/p1

    https://www.daz3d.com/forums/discussion/comment/7980511/#Comment_7980511

    I don't have access to all my notes or code at this time. Did you try it with the line passing through the DzBox3 rather than starting/finishing inside (or on boundary) ?

    Post edited by surreal on
  • OmnifluxOmniflux Posts: 427

    Haha, I even commented in the second thread there that it was a known issue.

    I cannot find either of those threads through google search though.

    Yes, I did test passing though the box at a point not on the boundary. Not sure why I didn't post that as my example case though...

Sign In or Register to comment.