Profiler resolution

[Anonymous]

I couldn't find this covered previosly anywhere, my apologies if it has been.

Does anyone know how ArchiCAD's profiler decides what resolution it will profile along a curved path? A typical example is a curved handrail to a curved balcony, which comes out badly segmented. I wondered if the resolution relates to the magic wand settings, but it doesn't seem to. The GDL script when you open the created objects has plotted nodes along this path using tube, but how has profiler decided how much to segment along the curve? Is it completely internal within the profiler operation and uneditablejavascript:emoticon(':?:')

[Anonymous]

Peter wrote:
Ray,
I think that regardless of whether Profiler is using the TUBE command
or the REVOLVE command the resolution is 36.

But if this was the case how did Erika get the results she showed in her post of Fri 18 Feb.

Right now it's Thu 17th March 3:35PM (15 hours ahead of you). What are you doing tapping away after midnight

[Anonymous]

Ray,
You are right, there is still an issue.
I just looked at a Profiler script of a drum shaped object
that was generated from a profile (polyline) revolved through
360° but the number of facets is 48.
It is late for me to be tapping away so I'm not thinking clearly
I will have look at this again in the morning.

Thanks for giving me the time off day
Talk to you tomorrow,
Peter Devlin

[David Maudlin]

Peter wrote:
I checked in the 3D script of a Profiler object that used the REVOLVE command and again there are no additional status codes for the RESOL command to operate on.
The 2D script of this object does use additional status codes but there is no RESOL command in the 2D script.

The 3D script is as follows:

MUL a/0.704003, b/0.704003, zzyzx/0.914400
RESOL 36
MATERIAL mat
ROTY 270+offset
ROTX begang
issolid=1
IF angle<360 THEN mask=16+32 ELSE mask=0
IF angle=360 AND NOT(issolid) THEN angle=359.9
IF issolid THEN cmask=mask+4+8 ELSE cmask=mask
REVOLVE 12, angle, mask,
0.889000, ABS(0.000000*dir+rad), 2,
0.889000, ABS(0.254000*dir+rad), 2,
0.914400, ABS(0.254000*dir+rad), 2,
0.914400, ABS(0.266700*dir+rad), 2,
0.735901, ABS(0.323343*dir+rad), 1,
0.550836, ABS(0.352002*dir+rad), 1,
0.363564, ABS(0.352002*dir+rad), 1,
0.178499, ABS(0.323343*dir+rad), 1,
0.000000, ABS(0.266700*dir+rad), 2,
0.000000, ABS(0.254000*dir+rad), 2,
0.025400, ABS(0.254000*dir+rad), 2,
0.025400, ABS(0.000000*dir+rad), 2
DEL TOP

Thank you,
Peter Devlin

Peter and others:

While I'm not sure about the case of the TUBE statement in objects generated by Profiler, in the use of REVOLVE the RESOL statement will matter: it will control the resolution of the object defined by the angle parameter. So while it does not affect the resolution of the profile, it will control the number of segments in the arc defined by angle. So the RESOL statement might be there for only the REVOLVE type of objects, but someone might find an example of where this would affect a TUBE, I think it would affect only the profile, not the path, as the "additional status codes" are only available for the profile. Hope this helps.

David

[Anonymous]

David,
You are correct.
When Profiler uses the REVOLVE command it needs the
RESOL statement to control the number of facets of
the generated revolution.

When Profiler is using the TUBE command and it is
operating on arcs for either the polyline or the path
it defaults to a resolution of 36 and the only way
to change this is to force it to follow
a path made of straight line segments.

When Profiler is using the REVOLVE command the resolution
of the revolution can be controlled, as gkmethy said,
by changing the RESOL value in the script.

When Profiler is using the REVOLVE command and it is
operating on an arc that is defining the profile to be revolved
it uses a resolution of 36 to define the number of points.
Again the way to control this is to define the profile
to be revolved with straight line segments.

I think this is the way it works. I could be wrong, so I will
have to do more tests.
Peter Devlin

[Anonymous]

Peter and David,

I'm sure you're both right about this.