From: Henning Rose (roseh++at++cci.sema.de)
Date: 05/09/2001 02:50:21
Hi pfUsers,
I'm working fpr virtuel car driving simulator with Performer 2.4 and RedHat
LINUX with a Pentium 533Mhz Prozessor and a NVIDIA TNT 2 graphicboard.
I need to place the current Channelview into the coordinates of an object
(for instance inside a Car), which is depending on various DCS,s in the
scene path. So, I need the complete transformation of this object (root to
object). Actually I use inside a viewing_update_callback(pftraverser * trav,
void * data) of the Application Traversal the Performerfunction:
pfGetTravMat() to get the apparently current transformation matrix of this
node!
But it doesent work so fine.
I'm getting only the transformation matrix of a frame one step in the past.
So if my parent DCS node is triggerd with a new postion the viewing position
is staying on its old value, during the object changed its postion. Only in
the next frame step the viewing position is triggered by the parent DCS's.
But if the objekt is still moving with a velocity the viewing position will
be shiftet during the time the DCS is triggerd with a volicity. I also tried
out to place the callbackfunction as Pre or as Post Traversal function in of
the PFTRAV_APP. But I only got the same problem.
I solved this mismatch, that I take the transfromation matrix of the
traversed node directly from the parrent DCS nodes instad from
pfGetTravMat() with a own function. But due to that I found a missing of
performance.
My question is. Is there a bug in pfGetTravMat()? Or how can I get the
current transformation matrix of an traversed node o a beter way?
Thanks for help.
Henning
----------------------------------------------------------------------------
---
Dr. Henning Rose -Research Physicist-
Competence Center Informatik GmbH
Departmet of Simulation
Lohberg 10
49716 Meppen
-Germany-
Tel.: (+49) 5931 / 805 361
email: Henning.Rose++at++cci.de
----------------------------------------------------------------------------
---
This archive was generated by hypermail 2b29 : Wed May 09 2001 - 02:50:37 PDT