I have been experiencing strange and erratic problems with isosurfaces
in beta 8, in both the Windows and Unix versions. The symptoms include
POV-Ray crasing, going into an infinite loop or getting just black images
(when there should be a visible isosurface). The symptoms are really strange
and erratic because different people seem to get different problems with the
same code (or no problems at all!).
  Please test isosurfaces with beta8 and forward a report to this article
if you find any kind of problem.
  For starters, test with this simple scene at different resolutions (eg.
320x240, 512x384 and so on), with and without antialiasing, with and without
display, etc. If you get a problem, also test uncommenting the first line
to see if the problem disappears.

//#declare foo=function{0}  // probable workaround for problems

camera { location -z*5 look_at 0 angle 35 }
light_source { <100,200,-300>, 1 }

isosurface
{ function { x^2+y^2+z^2-1 }
  contained_by { sphere { 0,1 } }
  pigment { rgb x }
}


-- 
#macro N(D,I)#if(I<6)cylinder{M()#local D[I]=div(D[I],104);M().5,2pigment{
rgb M()}}N(D,(D[I]>99?I:I+1))#end#end#macro M()<mod(D[I],13)-6,mod(div(D[I
],13),8)-3,10>#end blob{N(array[6]{11117333955,
7382340,3358,3900569407,970,4254934330},0)}//                     - Warp -

Post a reply to this message

320x240: POV-Ray traces a few lines then it crashes.
400x300 and higher: POV-Ray stops to trace when it reaches the iso, but
doesn't crash. The only thing I can do is to close POV.

Antialiasing did not have any effect in my tests.

Declaring any function makes it work, but only if it comes before the iso.

POV-Ray v3.5 b8 Windows Me Athlon

--
Jonathan.

Post a reply to this message

800x600 and 1024x768 make it crash too.

"JRG" <jrg### [at] hotmailcom> ha scritto nel messaggio
news:3c0958d3$1@news.povray.org...
> 320x240: POV-Ray traces a few lines then it crashes.
> 400x300 and higher: POV-Ray stops to trace when it reaches the iso, but
> doesn't crash. The only thing I can do is to close POV.
>
> Antialiasing did not have any effect in my tests.
>
> Declaring any function makes it work, but only if it comes before the iso.
>
> POV-Ray v3.5 b8 Windows Me Athlon
>
> --
> Jonathan.
>
>

Post a reply to this message

Oops, forgot to mention that when it crashes this error message is
displayed: "Rendering Error: Maximum function evaluation recursion level
reached."

"JRG" <jrg### [at] hotmailcom> ha scritto nel messaggio
news:3c0958d3$1@news.povray.org...
> 320x240: POV-Ray traces a few lines then it crashes.
> 400x300 and higher: POV-Ray stops to trace when it reaches the iso, but
> doesn't crash. The only thing I can do is to close POV.
>
> Antialiasing did not have any effect in my tests.
>
> Declaring any function makes it work, but only if it comes before the iso.
>
> POV-Ray v3.5 b8 Windows Me Athlon
>
> --
> Jonathan.
>
>

Post a reply to this message

Warp wrote:
> 
>   I have been experiencing strange and erratic problems with isosurfaces
> in beta 8, in both the Windows and Unix versions. The symptoms include
> POV-Ray crasing, going into an infinite loop or getting just black images
> (when there should be a visible isosurface). The symptoms are really strange
> and erratic because different people seem to get different problems with the
> same code (or no problems at all!).
>   Please test isosurfaces with beta8 and forward a report to this article
> if you find any kind of problem.
> [...]

Your sample scene does not crash, neither on K6 nor Athlon here, several
resolutions tried. But i experienced some quite strange things today, like
occasional crashes and wrong render results.  They were not reproducible
and therefore very difficult to track.  

Christoph 

-- 
Christoph Hormann <chr### [at] gmxde>
IsoWood include, radiosity tutorial, TransSkin and other 
things on: http://www.schunter.etc.tu-bs.de/~chris/

Post a reply to this message

JRG <jrg### [at] hotmailcom> wrote:
: Declaring any function makes it work, but only if it comes before the iso.

  Really? Now that's strange (in my experiments declaring the dummy function
at the beginning of the scene seemed to get rid of all problems, but if what
you say is true it's even more erratic than I thought).

-- 
#macro N(D,I)#if(I<6)cylinder{M()#local D[I]=div(D[I],104);M().5,2pigment{
rgb M()}}N(D,(D[I]>99?I:I+1))#end#end#macro M()<mod(D[I],13)-6,mod(div(D[I
],13),8)-3,10>#end blob{N(array[6]{11117333955,
7382340,3358,3900569407,970,4254934330},0)}//                     - Warp -

Post a reply to this message

Warp <war### [at] tagpovrayorg> wrote in news:3c0956db@news.povray.org:

>   Please test isosurfaces with beta8 and forward a report to this
>   article if you find any kind of problem.

(Use monospaced font)

Tests on a P3-866 running WinME, without the workarond function 
declaration:

Res      AA  Display Result
160x120  off on      Try #1: Pov _disappears_ with no error message or trap
                             report.
                     Try #2: pvengine trap on 9th line of the image.
160x120  off off     pvengine trap on 9th line of the image.
320x240  off on      pvengine trap on the 17th line of the image.
320x240  off off     pvengine trap on the 17th line of the image.
400x300  off on      pvengine trap on 21st line of the image
400x300  off off     pvengine trap on 21st line of the image
640x480  0.3 on      Render thread appears to enter an infinite loop and
                     generates no more content after line 34 as indicated
                     on (lower) status bar.  Title bar says 'line 1 of
                     480'.  Alt-G/Stop icon has no effect.  Pov must be
                     shut down in order to reset the renderer.
640x480  0.3 off     As above.
800x600  0.3 on      As above except render halts at line 43.
800x600  0.3 off     As above.

The traps which occurred during low-resolution renders were at EIP=
0x0010CE10, caused because the processor attempted to execute move 
[esi],edx while [esi] pointed into la la land.  There was no regular 
pattern to [esi]; null and other values were observed.

Tests on a Celeron 400 running Win2K, without the workaround function:

Res      AA  Display Result
160x180  off on      Trapped in 2/2 attempts.
320x240  off on      Renders in 8/10 attempts.  Failed with function
                     division by zero error on fifth attempt.  Error
                     message did not give either the correct file name or a
                     function name.  Pvengine trapped on the tenth attempt.
400x300  off on      Renders 8/10 times.  Attempt 3 rendered very slowly (~
                     600 pps).  Attempt six trapped.
400x300  off off     Trapped 4/4 times at line 21.
640x480  0.3 on      Renders 5/10 times.  Function division by zero on
                     attempts 1 and 9; incorrect filename given on attempt
                     1.  Rendered very slowly on attempts 5 and 6.
                     Rendered a completely black image on attempt 8.
640x480  0.3 off     Renders 4/10 times. Trapped on attempt 4; fell into an
                     infinite loop @ line 34 on attempts 5 and 7-10.
800x600  0.3 on      Renders 6/10 times.  Division by zero on attempts 4, 5
                     and 9; correct filename only given on attempt 4.
                     Trapped on attempt 7.

The scene renders without incident at all resolutions with all combinations 
of aa and preview display on both PCs when the workarond function 
declaration is included.

Post a reply to this message

The below is my result.   I have experienced such bug once (or twice) 
during the isosurface patch development.   I think there are several 
other bugs relating isosurface core code.  Now I can take some time
for POV-Ray for a while (maybe a few months) and can help POV-Team
if the source code is available.

 CPU       OS         Compiler
Celeron  Windows Me   : Intel : It crashes.
Celeron  Windows Me   : VC    : It crashes.
Athron  Windows 2000  : Intel : No plobrem.
Athron  Windows 2000  : VC    : No plobrem.
Pentium III Windows 2000: Intel : It crashes.
Pentium III Windows 2000: VC   : It does not crash but Fatal error.

R. Suzuki

Post a reply to this message

As discussed in p.p-b I know there are problems, but there is no pattern for
those is available because they are memory/pointer related.  So everybody
will see different problema nd at this time there is no fix possible.

    Thorsten

____________________________________________________
Thorsten Froehlich, Duisburg, Germany
e-mail: tho### [at] trfde

Visit POV-Ray on the web: http://mac.povray.org

Post a reply to this message

Thanks for your exhaustive testing.
  Now I can be sure that it's not just me, but something weird is really
happening.
  The seemingly random nature of the problem may indicate that some illegal
memory section or an already freed memory section is being read, or something
similar. Perhaps also an unitialized variable or table might cause this kind
of erratic behaviour.

-- 
#macro N(D,I)#if(I<6)cylinder{M()#local D[I]=div(D[I],104);M().5,2pigment{
rgb M()}}N(D,(D[I]>99?I:I+1))#end#end#macro M()<mod(D[I],13)-6,mod(div(D[I
],13),8)-3,10>#end blob{N(array[6]{11117333955,
7382340,3358,3900569407,970,4254934330},0)}//                     - Warp -

Post a reply to this message

Thorsten Froehlich <tho### [at] trfde> wrote:
: As discussed in p.p-b I know there are problems, but there is no pattern for
: those is available because they are memory/pointer related.  So everybody
: will see different problema nd at this time there is no fix possible.

  Since the problems were so erratic I just wanted a confirmation that I am not
crazy and it doesn't happen just to me. :)

-- 
#macro N(D,I)#if(I<6)cylinder{M()#local D[I]=div(D[I],104);M().5,2pigment{
rgb M()}}N(D,(D[I]>99?I:I+1))#end#end#macro M()<mod(D[I],13)-6,mod(div(D[I
],13),8)-3,10>#end blob{N(array[6]{11117333955,
7382340,3358,3900569407,970,4254934330},0)}//                     - Warp -

Post a reply to this message

In article <3c0a22e8@news.povray.org> , Warp <war### [at] tagpovrayorg>  wrote:

>   Thanks for your exhaustive testing.
>   Now I can be sure that it's not just me, but something weird is really
> happening.
>   The seemingly random nature of the problem may indicate that some illegal
> memory section or an already freed memory section is being read, or something
> similar. Perhaps also an unitialized variable or table might cause this kind
> of erratic behaviour.

Where my problem is that the code didn't change since beta 6 :-(

    Thorsten

____________________________________________________
Thorsten Froehlich, Duisburg, Germany
e-mail: tho### [at] trfde

Visit POV-Ray on the web: http://mac.povray.org

Post a reply to this message

"R. Suzuki" <r-s### [at] aistgojp> wrote in message
news:3c0a0b8b@news.povray.org...
>
>  CPU       OS         Compiler   <<< Compirel
> Celeron  Windows Me   : Intel : It crashes.  <<< Cerelon, clashes
> Celeron  Windows Me   : VC    : It crashes.
> Athron  Windows 2000  : Intel : No plobrem.  <<< Inter
> Athron  Windows 2000  : VC    : No plobrem.
> Pentium III Windows 2000: Intel : It crashes.
> Pentium III Windows 2000: VC   : It does not crash but Fatal error.  <<<
ellol

:-D  Sorry, I saw a great opportunity to have some fun and couldn't pass it
up.  :-D

bob h

Post a reply to this message

bob h <omn### [at] charternet> wrote:
: :-D  Sorry, I saw a great opportunity to have some fun and couldn't pass it
: up.  :-D

  Having studied some Japanese it's not strange to me anymore that they
don't distinguish very well between r and l.
  In fact, the syllables ra, ri, ru, re and ro are not pronounced neither as
with an 'r' nor with an 'l' as in most western languages, but a kind-of
in-between sound.
  When translitterating foreign words to katakana, both l and r are written
with ra, ri, ru, re or ro.

  By the way, Japanese is a very interesting language. If someone wants to
learn an exotic foreign language which is not too difficult, I recommend
Japanese. It's surprisingly easy to learn (specially compared to its
neighbour language, Chinese, which I think is extremely difficult).

-- 
#macro N(D,I)#if(I<6)cylinder{M()#local D[I]=div(D[I],104);M().5,2pigment{
rgb M()}}N(D,(D[I]>99?I:I+1))#end#end#macro M()<mod(D[I],13)-6,mod(div(D[I
],13),8)-3,10>#end blob{N(array[6]{11117333955,
7382340,3358,3900569407,970,4254934330},0)}//                     - Warp -

Post a reply to this message

Warp wrote:
>   By the way, Japanese is a very interesting language. If someone wants to
> learn an exotic foreign language which is not too difficult, I recommend
> Japanese. It's surprisingly easy to learn (specially compared to its
> neighbour language, Chinese, which I think is extremely difficult).

That's funny: some days ago I had a look through my mother's English books,
since I wanted to find one to improve my 'poor' English, and, neglecting
Shakespeare, Pirsig and Stone, I chose 'A guide to reading and writing
Japanese' by Charles E. Tuttle Co.
So, now I'm learning two languages and paying for one... :)

--
Jonathan.

Post a reply to this message

"JRG" <jrg### [at] hotmailcom> wrote in message
news:3c0aae91@news.povray.org...
> Warp wrote:
> >   By the way, Japanese is a very interesting language. If someone wants
to
> > learn an exotic foreign language which is not too difficult, I recommend
> > Japanese. It's surprisingly easy to learn (specially compared to its
> > neighbour language, Chinese, which I think is extremely difficult).
>
> That's funny: some days ago I had a look through my mother's English
books,
> since I wanted to find one to improve my 'poor' English, and, neglecting
> Shakespeare, Pirsig and Stone, I chose 'A guide to reading and writing
> Japanese' by Charles E. Tuttle Co.
> So, now I'm learning two languages and paying for one... :)


My Dad was over in Japan around the end of the Korean War and loved the
place, brought back some Japanese words/phrases and still says something on
a rare occasion.  Because of that one of my neices took a Japanese language
class.  Something I never considered would be easy despite them, and now
you, saying so.
Actually, it's not so much the speaking it that deters me it's the writing.
I dare not stray from the Roman alphabet and it's varied forms  :-)
--
text{ttf"arial","bob h",.1,0pigment{rgb 9}translate<-1,-.2,3>}

Post a reply to this message

Ha-ha-ha!

If you search those words using google, etc., you will find 
 thauthands of websites (including computer journals).

As Warp described, we usually don't care 'r' and 'l', also
's' and 'th', 'f' and 'h', 'b' and 'v', ... in pronounciation.

POV-Ray in japanese : povulei, poburei, pobulei, povulei, poblei,... 

Sorry for off-topic.

R. Suzuki

Post a reply to this message

Well, the confirmation does not confirm that you are not crazy.  It does
confirm that others share your problems ...

:>)

Jim


"Warp" <war### [at] tagpovrayorg> wrote in message
news:3c0a2e9c@news.povray.org...
> Thorsten Froehlich <tho### [at] trfde> wrote:
> : As discussed in p.p-b I know there are problems, but there is no pattern
for
> : those is available because they are memory/pointer related.  So
everybody
> : will see different problema nd at this time there is no fix possible.
>
>   Since the problems were so erratic I just wanted a confirmation that I
am not
> crazy and it doesn't happen just to me. :)
>
> --
> #macro N(D,I)#if(I<6)cylinder{M()#local D[I]=div(D[I],104);M().5,2pigment{
> rgb M()}}N(D,(D[I]>99?I:I+1))#end#end#macro M()<mod(D[I],13)-6,mod(div(D[I
> ],13),8)-3,10>#end blob{N(array[6]{11117333955,
> 7382340,3358,3900569407,970,4254934330},0)}//                     - Warp -

Post a reply to this message

On 1 Dec 2001 17:16:59 -0500, Warp <war### [at] tagpovrayorg> wrote:
>  I have been experiencing strange and erratic problems with isosurfaces
> in beta 8, in both the Windows and Unix versions. The symptoms include
> POV-Ray crasing, going into an infinite loop or getting just black images
> (when there should be a visible isosurface).

POV 3.5 b 8 on PII 233 128 MB with NT 4 Sp 6
I can confirm. I have started sample animation i_internal.pov. It works fine but
I forgot to set png output. I stoped this and restarted with +FN and is stoped
at first frame with message

Scene contains 2 frame level objects; 0 infinite.
00:00:01 so far, rendering frame 1, going to 77.

File: C:\Program Files\POV-Ray for Windows v3.5\INCLUDE\functions.inc  Line: 16
Fatal Error: Floating-point exception detected in function 'f_algbr_cyl1'. Your
function either attempted a division by zero, used a function outside its domain
or called an internal function with invalid parameters.

ABX
--
#declare _=function(a,b,x){((a^2)+(b^2))^.5-x}#default {pigment{color rgb 1}}
union{plane{y,-3}plane{-x,-3}finish{reflection 1 ambient 0}}isosurface{ //ABX
function{_(x-2,y,1)|_((x+y)*.7,z,.1)|_((x+y+2)*.7,z,.1)|_(x/2+y*.8+1.5,z,.1)}
contained_by{box{<0,-3,-.1>,<3,0,.1>}}translate z*15finish{ambient 1}}//POV35

Post a reply to this message