Hi all
I have a scene, which consist of only one polygon. This polygon is an illum and it renders visible. It renders unvisible as expected if I set either:
- ps = 6
- change the height of the illum face
- change the view a little bit.
The effect also depends on the Radiance version. I first found it in the old Windows version, but it is also in Radiance 3.5 and 3.6a. The latter renders the polygon outlined, with dashed lines, but it is visible.
This might be a problem in the Radiance kernel. The project files can be downloaded from
ftp://ftp.relux.biz/radiance/radProblem.zip
Siegbert
···
--
Dr.-Ing. Siegbert Debatin
Softwareentwicklung
Relux Informatik AG Dornacherstr. 377 CH-4053 Basel
Tel. ++41 61 333 07 70 Fax: ++41 61 333 07 72
no problem on 3.6a, macosX.
used objview *.rad to preview
and nothing appears,
I added a red polygon and changed between illum/light several times to test,
also tried rad -o x11 relux2_1.rif, all ok.
everything appears to be ok...
cheers,
giulio
PS may be you are using an old .oct that is frozen?
···
-----Original Message-----
From: [email protected]
[mailto:[email protected]]On Behalf Of Siegbert
Debatin
Sent: 20 October 2004 08:09
To: [email protected]
Subject: [Radiance-dev] Illum problem
Hi all
I have a scene, which consist of only one polygon. This polygon is an
illum and it renders visible. It renders unvisible as expected if I set
either:
- ps = 6
- change the height of the illum face
- change the view a little bit.
The effect also depends on the Radiance version. I first found it in the
old Windows version, but it is also in Radiance 3.5 and 3.6a. The latter
renders the polygon outlined, with dashed lines, but it is visible.
This might be a problem in the Radiance kernel. The project files can be
downloaded from
ftp://ftp.relux.biz/radiance/radProblem.zip
Siegbert
--
Dr.-Ing. Siegbert Debatin
Softwareentwicklung
Relux Informatik AG Dornacherstr. 377 CH-4053 Basel
Tel. ++41 61 333 07 70 Fax: ++41 61 333 07 72
_______________________________________________
Radiance-dev mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-dev
___________________________________________________________________
Electronic mail messages entering and leaving Arup business
systems are scanned for acceptability of content and viruses.
no problem on 3.6a, macosX.
used objview *.rad to preview
and nothing appears,
I added a red polygon and changed between illum/light several times to test,
also tried rad -o x11 relux2_1.rif, all ok.
This way, it also works on my system, probably because of a different parameter set. The problem only appears if I invoke Radiance via rad.
Regards
Siegbert
···
--
Dr.-Ing. Siegbert Debatin
Softwareentwicklung
Relux Informatik AG Dornacherstr. 377 CH-4053 Basel
Tel. ++41 61 333 07 70 Fax: ++41 61 333 07 72
try to delete the octree.
ad run again rad -o x11
hope it works...
giulio
···
-----Original Message-----
From: Siegbert Debatin [mailto:[email protected]]
Sent: 20 October 2004 12:32
To: code development
Subject: Re: [Radiance-dev] Illum problem
no problem on 3.6a, macosX.
used objview *.rad to preview
and nothing appears,
I added a red polygon and changed between illum/light several times to
test,
also tried rad -o x11 relux2_1.rif, all ok.
This way, it also works on my system, probably because of a different
parameter set. The problem only appears if I invoke Radiance via rad.
Regards
Siegbert
--
Dr.-Ing. Siegbert Debatin
Softwareentwicklung
Relux Informatik AG Dornacherstr. 377 CH-4053 Basel
Tel. ++41 61 333 07 70 Fax: ++41 61 333 07 72
_______________________________________________
Radiance-dev mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-dev
___________________________________________________________________
Electronic mail messages entering and leaving Arup business
systems are scanned for acceptability of content and viruses.
ok,
if I use rad and I write to a file I have the same problem:
you are right.
and you are right also when you say that it is due to the -ps setting....
no idea now!
just use -ps 1..., but you did already 
cheers
giulio
···
-----Original Message-----
From: Giulio Antonutto [mailto:[email protected]]
Sent: 20 October 2004 13:30
To: 'code development'
Subject: RE: [Radiance-dev] Illum problem
try to delete the octree.
ad run again rad -o x11
hope it works...
giulio
-----Original Message-----
From: Siegbert Debatin [mailto:[email protected]]
Sent: 20 October 2004 12:32
To: code development
Subject: Re: [Radiance-dev] Illum problem
no problem on 3.6a, macosX.
used objview *.rad to preview
and nothing appears,
I added a red polygon and changed between illum/light several times to
test,
also tried rad -o x11 relux2_1.rif, all ok.
This way, it also works on my system, probably because of a different
parameter set. The problem only appears if I invoke Radiance via rad.
Regards
Siegbert
--
Dr.-Ing. Siegbert Debatin
Softwareentwicklung
Relux Informatik AG Dornacherstr. 377 CH-4053 Basel
Tel. ++41 61 333 07 70 Fax: ++41 61 333 07 72
_______________________________________________
Radiance-dev mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-dev
___________________________________________________________________
Electronic mail messages entering and leaving Arup business
systems are scanned for acceptability of content and viruses.
_______________________________________________
Radiance-dev mailing list
[email protected]
http://www.radiance-online.org/mailman/listinfo/radiance-dev
___________________________________________________________________
Electronic mail messages entering and leaving Arup business
systems are scanned for acceptability of content and viruses.
ok,
if I use rad and I write to a file I have the same problem:
you are right.
and you are right also when you say that it is due to the -ps setting....
no idea now!
just use -ps 1..., but you did already
So I am not the only one. Thank you for testing, Giulio.
Regards
Siegbert
···
--
Dr.-Ing. Siegbert Debatin
Softwareentwicklung
Relux Informatik AG Dornacherstr. 377 CH-4053 Basel
Tel. ++41 61 333 07 70 Fax: ++41 61 333 07 72
Hi Siegbert, Giulio
unfortunately I don't have acess to my windows machine right now, so I can't unpack the zip file (if you could probide a normal linux/unix tar or tar.gz I'd have a look..)
There is one 'problem' with illums and the additional source-drawing routine I'm aware of, but this only occurs if the illum is viewed from a distant point of view
-cb
unfortunately I don't have acess to my windows machine right now, so I can't unpack the zip file (if you could probide a normal linux/unix tar or tar.gz I'd have a look..)
Hi Carsten
No problem:
ftp://ftp.relux.biz/radiance/radProblem.tar.gz
regards
Siegbert
···
--
Dr.-Ing. Siegbert Debatin
Softwareentwicklung
Relux Informatik AG Dornacherstr. 377 CH-4053 Basel
Tel. ++41 61 333 07 70 Fax: ++41 61 333 07 72
Hi Siegbert, Giulio
yepp, thanx. Now I see it right before my eyes.. it's looking very strange indeed..
I came so far by now with my first tests:
You look at the source at a very flat angle, and I suppose it's something with the additional source-drawing routine. The pattern doesn't appear in rvu (where this add. src-drawing is not performed).
I made checks with radzilla, too, where I've switched off source-drawing for illums by default. The pattern doesn't appear.
But if I set -dd 1 individually for the illum polygon (i.e. switch on add. source-drawing), I get the rim pattern, too. (but only for -ps 8 and higher)
The connection to the ps setting (which controls the sampling density for the picture generation) is not yet obvious to me. And probably the boxcorr function plays a role,too...
so there are still some mysteries to unravel 
-cb
Look in Srcsmp.c--if you dare--and you'll see either a. the holy grail or b. Armageddon depending upon your point of view.
-Chas
···
Carsten Bauer <[email protected]> wrote:
Hi Siegbert, Giulio
yepp, thanx. Now I see it right before my eyes.. it's looking very
strange indeed..
I came so far by now with my first tests:
You look at the source at a very flat angle, and I suppose it's
something with the additional source-drawing routine. The pattern
doesn't appear in rvu (where this add. src-drawing is not performed).
I made checks with radzilla, too, where I've switched off source-drawing
for illums by default. The pattern doesn't appear.
But if I set -dd 1 individually for the illum polygon (i.e. switch on
add. source-drawing), I get the rim pattern, too. (but only for -ps 8
and higher)
The connection to the ps setting (which controls the sampling density
for the picture generation) is not yet obvious to me. And probably the
boxcorr function plays a role,too...
so there are still some mysteries to unravel
-cb
Hi Siegbert, Guilio, Carsten,
There was indeed a problem with the source drawing routines used by rpict, which are not used in rvu. I thought I had fixed this awhile ago, but apparently not well enough. I just checked in a new version of srcdraw.c, which I'm hoping will be included in tomorrow's HEAD tarball, or you can get it individually from CVS if you like. The change was minor, affecting the edges of illum surfaces. Let me know if there are still problems.
And thanks for pointing this out!
-Greg
···
From: [email protected] (Carsten Bauer)
Date: October 20, 2004 9:22:15 AM PDT
Hi Siegbert, Giulio
yepp, thanx. Now I see it right before my eyes.. it's looking very strange indeed..
I came so far by now with my first tests:
You look at the source at a very flat angle, and I suppose it's something with the additional source-drawing routine. The pattern doesn't appear in rvu (where this add. src-drawing is not performed).
I made checks with radzilla, too, where I've switched off source-drawing for illums by default. The pattern doesn't appear.
But if I set -dd 1 individually for the illum polygon (i.e. switch on add. source-drawing), I get the rim pattern, too. (but only for -ps 8 and higher)
The connection to the ps setting (which controls the sampling density for the picture generation) is not yet obvious to me. And probably the boxcorr function plays a role,too...
so there are still some mysteries to unravel
-cb
Charles Ehrlich wrote:
Look in Srcsmp.c--if you dare--and you'll see either a. the holy grail or b. Armageddon depending upon your point of view.
its probably both at the same time
In fact, that special detail is really fantastic stuff, and after looking into srcdraw.c I was impressed by the amount of work which is necessary to implement the - otherwise - straightforward sounding idea.
And my approach of switching it off for illums in general simply serves to allow undisturbed cheating
-cb
From: [email protected] (Carsten Bauer)
In fact, that special detail is really fantastic stuff, and after looking into srcdraw.c I was impressed by the amount of work which is necessary to implement the - otherwise - straightforward sounding idea.
I thought so as well, but Rob Shakespeare really needed it for his neon lights, so I spent the time to get it working. It was a major headache.
And my approach of switching it off for illums in general simply serves to allow undisturbed cheating
The new change effectively switches source drawing off for illum's -- at least, I think it does!
-Greg
We checked out the head revision and compiled the Windows version, and it works. Thank you, Greg.
Siegbert
···
Greg Ward <[email protected]> wrote:
The new change effectively switches source drawing off for illum's -- at least, I think it does!
--
Dr.-Ing. Siegbert Debatin
Softwareentwicklung
Relux Informatik AG Dornacherstr. 377 CH-4053 Basel
Tel. ++41 61 333 07 70 Fax: ++41 61 333 07 72
