Motion blur on procedural movements
-
THE POST BELOW IS MORE THAN 5 YEARS OLD. RELATED SUPPORT INFORMATION MIGHT BE OUTDATED OR DEPRECATED
On 01/03/2012 at 09:51, xxxxxxxx wrote:
Hi all,
I've a problem rendering an object whose movements are non-keyframed, but generated by a script instead. It seems the rendering engine cannot calculate the motion blur.
Following is a simple example, you can just paste the code in a Python tag on whatever object you want.
-----------------------------------------------------------------
import c4d, math
from c4d import documents, BaseTimedef main() :
doc = documents.GetActiveDocument()
tm = doc.GetTime()
fps = doc.GetFps()
fr = tm.GetFrame(fps)
obj = op.GetObject()
r = 500.0
ang = -0.3*frobj[c4d.ID_BASEOBJECT_REL_POSITION,c4d.VECTOR_X] = r*math.cos(ang);
obj[c4d.ID_BASEOBJECT_REL_POSITION,c4d.VECTOR_Y] = r*math.sin(ang);-----------------------------------------------------------------
It creates a procedural orbit on XY plane. If you try to render it with motion blur active (Physical or Standard engine) you'll notice the problem. I guess the problem is the engine is not able to calculate the subframe positions for procedural movements. That happens just for Coffee/Python expressions, because with a Vibrate Tag, just for example, it works well.
Any suggestion?
Riccardo
-
THE POST BELOW IS MORE THAN 5 YEARS OLD. RELATED SUPPORT INFORMATION MIGHT BE OUTDATED OR DEPRECATED
On 01/03/2012 at 11:02, xxxxxxxx wrote:
The problem is this line here: doc = documents.GetActiveDocument()
'doc' here refers to the document in the editor, not in the rendering.
The document is cloned for the renderer. Just delete line 6, 'doc' is
predefined in a Python tag and is always a reference to the document
where the tag is part of.Cheers, Sebastian
-
THE POST BELOW IS MORE THAN 5 YEARS OLD. RELATED SUPPORT INFORMATION MIGHT BE OUTDATED OR DEPRECATED
On 02/03/2012 at 01:39, xxxxxxxx wrote:
Thank you for your answer Sebastian.
That's a useful info, but unfortunately it doesn't solve the problem. If you try to render the scene you'll notice the motion blur doesn't work well. Apparently it's not able to calculate the interframe positions, so the effect is a simple overlap between a couple of frames.
In fact I've the same problem using the coffee version.
----------------------------------------------------------------
main(doc,op)
{
var fr = doc->GetTime()->GetFrame(doc->GetFps());
var r = 500.0;
var ang = -0.3*fr;
op#ID_BASEOBJECT_REL_POSITION:VECTOR_X = r*cos(ang);
op#ID_BASEOBJECT_REL_POSITION:VECTOR_Y = r*sin(ang);
}----------------------------------------------------------------
-
THE POST BELOW IS MORE THAN 5 YEARS OLD. RELATED SUPPORT INFORMATION MIGHT BE OUTDATED OR DEPRECATED
On 02/03/2012 at 03:27, xxxxxxxx wrote:
Hi Sebastian,
I didn't know that, too.
Is it missing in the docs? I think so, that should be added please. -
THE POST BELOW IS MORE THAN 5 YEARS OLD. RELATED SUPPORT INFORMATION MIGHT BE OUTDATED OR DEPRECATED
On 02/03/2012 at 04:00, xxxxxxxx wrote:
Thank you for your answer Sebastian.
That's a useful info, but unfortunately it doesn't solve the problem. If you try to render the scene you'll notice the motion blur doesn't work well. Apparently it's not able to calculate the interframe positions, so the effect is a simple overlap between a couple of frames.
In fact I've the same problem using the coffee version.
I can now reproduce it. The problem is related to GetFrame() which just returns the current frame (as integer). So instead of GetFrame() in Coffee I would use doc->GetTime()->GetSecond() and doc.GetTime().Get() in Python which returns a float value of the current time. To calculate the motion blur effect the renderer modifies the scene time in smaller steps ( < 1 frame) so these functions are more useful here.
Is it missing in the docs? I think so, that should be added please.
You can also call op.GetDocument() but I will mark it to be added. Thx.
Cheers, Sebastian
-
THE POST BELOW IS MORE THAN 5 YEARS OLD. RELATED SUPPORT INFORMATION MIGHT BE OUTDATED OR DEPRECATED
On 02/03/2012 at 09:10, xxxxxxxx wrote:
That definitely solve the problem, thanks again!
Riccardo