|
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
THIMC,
If I run the attached POV file I get the following error.
Fatal error in renderer: Out of memory.
Render failed
If I comment out the cutaway_textures, or change the plane from y to -y
it works fine. I suspect a bug in the cutaway_textures code.
I have just tried it again after changing line 77 to read " #local
rx=rx+10;" and it now works... In fact any step value over 5.715
works, but anything less fails, but only with cutaway_textures!
It fails consistantly in both Version 3.7.0.beta.30.msvc8-sse2.win32 and
in Version 3.7.0.beta.30.msvc8.win32, but works fine in Version
3.6.1.icl8.win32
If you want any other info, just ask.
Rarius
Post a reply to this message
Attachments:
Download 'iso-8859-1' (2 KB)
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Are you sure you attached a POV scene file? IE just presents me with the same
text you wrote in your post...
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
On Fri, 23 Jan 2009 18:40:53 +0100, clipka <nomail@nomail> wrote:
> Are you sure you attached a POV scene file? IE just presents me with the
> same text you wrote in your post...
He did. Contents below. I tested the scene with
3.7.0.beta.30.msvc8-sse2.win32 and could reproduce the problem by changing
the number on line 77 to 5.
/////////////////////////////////////////////////////////////////////////////
// Include Files
#include "colors.inc"
#include "metals.inc"
/////////////////////////////////////////////////////////////////////////////
// Declared pigments, finishes, etc.
#declare sky_vect = z;
#declare lookat = <0, 0, 0>;
#declare cam_loc = lookat+(<.5, 1, .25>*600);
#declare cam_zoom = 10;
#declare light_loc = <1, .5, .25>*500000;
global_settings
{
max_trace_level 12
assumed_gamma 2.2
}
background { Black }
#default
{
pigment { rgb<.70, .70, .75> }
finish
{
ambient .25
diffuse .75
reflection 0
phong 0
phong_size 30
brilliance .5
}
}
/////////////////////////////////////////////////////////////////////////////
// Lights
light_source
{
light_loc
colour rgb<1, 1, 1>
}
/////////////////////////////////////////////////////////////////////////////
// Camera
camera
{
location cam_loc
direction <cam_zoom,0,0>
up sky_vect
right <0, -(image_width/image_height), 0>
sky sky_vect
look_at lookat
}
/////////////////////////////////////////////////////////////////////////////
// Objects
#local r=8.5852/2; //.338in
#local CaselessBullet=
intersection
{
cylinder{<0, 0, 0>, <39.5, 0, 0>, r translate<-12.5, 0, 0>}
#local rx=0;
#while(rx<360)
sphere
{
0, 1
scale<67, 15, 50>
translate<0, 0, -50+r>
rotate<rx, 0, 0>
}
#local rx=rx+5.715;
#end
translate<12.5, 0, 0>
}
difference
{
object{CaselessBullet texture{T_Chrome_5C}}
plane{-y, 0}
cutaway_textures
}
--
FE
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> "Rarius" <rar### [at] rariuscouk> wrote in message
> news:4979d845@news.povray.org...
> THIMC,
>
> If I run the attached POV file I get the following error.
>
> Fatal error in renderer: Out of memory.
> Render failed
>
> If I comment out the cutaway_textures, or change the plane from y to -y it
> works fine.
Guessing you wrote that wrong. I see error with a -y plane and okay if y.
Same thing otherwise, smaller number addition in the while loop shows error.
Curiously, if the camera is rotated by 180*z it either fixes or causes the
error when changing the differencing plane to +y or -y.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"Bob Hughes" <omniverse charter net> wrote:
> Curiously, if the camera is rotated by 180*z it either fixes or causes the
> error when changing the differencing plane to +y or -y.
Maybe it's a question of interior_texture?
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
"clipka" <nomail@nomail> wrote in message
news:web.497c644b42b9e70e3c6235530@news.povray.org...
> "Bob Hughes" <omniverse charter net> wrote:
>> Curiously, if the camera is rotated by 180*z it either fixes or causes
>> the
>> error when changing the differencing plane to +y or -y.
>
> Maybe it's a question of interior_texture?
Tried adding that in the sphere and box union just now, same result. If put:
texture{pigment{rgb 1}}
interior_texture{pigment{rgb 1}}
into the CSG after the differenced box it overrides the cutaway.
Something else I tried was adding 'inverse' ahead of cutaway_textures,
thinking it might move the blank cutaway surface outside or be non-blank,
and got an error about cutaway_textures needing to be used in intersection
or difference (3.6 errors on this too).
Putting inverse afterward doesn't error, but still only black cutaway.
Looked at the program code (3.6 lighting.cpp, 3.7 trace.cpp) just to see
what might have changed, just out of curiosity not knowing what's really
going on in there. Main reason why I admire the programmers, very little of
it makes sense to me. That stuff scatters around to so many places in the
code I could never decipher it.
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Rarius wrote:
> THIMC,
>
> If I run the attached POV file I get the following error.
>
> Fatal error in renderer: Out of memory.
> Render failed
>
> If I comment out the cutaway_textures, or change the plane from y to
> -y it works fine. I suspect a bug in the cutaway_textures code.
>
> I have just tried it again after changing line 77 to read " #local
> rx=rx+10;" and it now works... In fact any step value over 5.715 works,
> but anything less fails, but only with cutaway_textures!
>
> It fails consistantly in both Version 3.7.0.beta.30.msvc8-sse2.win32 and
> in Version 3.7.0.beta.30.msvc8.win32, but works fine in Version
> 3.6.1.icl8.win32
The problem here is that there are too many entries in the weighted texture
list. Making the step smaller increases the total number of textures.
Currently the weighted texture list is a fixed-size vector (max 64
elements). This is done to force the vector to be created on the stack,
rather than the heap (which forces a malloc/free, requires heap locking,
and in general badly damages performance).
Clearly this has to be fixed (this is not the first time we've run into
this), but it'll need to wait a bit.
-- Chris
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
|
|