chalk it up to whatever you want (new compiler?, cleaned code?, ...?) but
the official 3.6 binary shaved 19 minutes off of benchmark.pov using the
benchmark.ini settings, compared to the official 3.5 binary with the same
official benchmark.pov/benchmark.ini. i never benchmarked pov before, so i
thought i'd do it before I installed 3.6.

computer info:
AMD Duron 800MHz running Slackware 9.1
X was running, as was Enlightenment, as would be my normal situation while
using povray.

on my slowish laptop, these were the results (handtyped, forgive any changes
in formatting):

3.6:
Parse Time: 0 hours 0 minutes 4 seconds
Photon Time: 0 hours 1 minutes 54 seconds
Render Time: 1 hours 30 minutes 34 seconds
Total Time: 1 hours 32 minutes 32 seconds

3.5:
Time for Parse: 0 hours 0 minutes 7 seconds
Time for Photon: 0 hours 2 minutes 3 seconds
Time for Trace: 1 hours 49 minutes 34 seconds
Total Time: 1 hours 51 minutes 44 seconds


please refrain from making jokes about my slow computer :) Regardless of
what caused the improvement i'm glad to see it.

-r

Post a reply to this message

In article <40ca2d06@news.povray.org> , "Ross" <rli### [at] everestkcnet> 
wrote:

> chalk it up to whatever you want (new compiler?, cleaned code?, ...?) but
> the official 3.6 binary shaved 19 minutes off of benchmark.pov using the
> benchmark.ini settings, compared to the official 3.5 binary with the same
> official benchmark.pov/benchmark.ini. i never benchmarked pov before, so i
> thought i'd do it before I installed 3.6.

This was mentioned bfeore: You *cannot* compare the 3.5 and the 3.6
becnhmark results due to anti-aliasing changes in POV-Ray 3.6!

    Thorsten

____________________________________________________
Thorsten Froehlich
e-mail: mac### [at] povrayorg

I am a member of the POV-Ray Team.
Visit POV-Ray on the web: http://mac.povray.org

Post a reply to this message

Thorsten Froehlich wrote:
> 
>>chalk it up to whatever you want (new compiler?, cleaned code?, ...?) but
>>the official 3.6 binary shaved 19 minutes off of benchmark.pov using the
>>benchmark.ini settings, compared to the official 3.5 binary with the same
>>official benchmark.pov/benchmark.ini. i never benchmarked pov before, so i
>>thought i'd do it before I installed 3.6.
> 
> 
> This was mentioned bfeore: You *cannot* compare the 3.5 and the 3.6
> becnhmark results due to anti-aliasing changes in POV-Ray 3.6!

For this speed difference the changes in photons are probably more 
important.

Christoph

-- 
POV-Ray tutorials, include files, Sim-POV,
HCR-Edit and more: http://www.tu-bs.de/~y0013390/
Last updated 01 May. 2004 _____./\/^>_*_<^\/\.______

Post a reply to this message

Thorsten Froehlich wrote:
> In article <40ca2d06@news.povray.org> , "Ross" <rli### [at] everestkcnet> 
> wrote:
> 
> 
>>chalk it up to whatever you want (new compiler?, cleaned code?, ...?) but
>>the official 3.6 binary shaved 19 minutes off of benchmark.pov using the
>>benchmark.ini settings, compared to the official 3.5 binary with the same
>>official benchmark.pov/benchmark.ini. i never benchmarked pov before, so i
>>thought i'd do it before I installed 3.6.
> 
> 
> This was mentioned bfeore: You *cannot* compare the 3.5 and the 3.6
> becnhmark results due to anti-aliasing changes in POV-Ray 3.6!
> 
>     Thorsten
> 
> ____________________________________________________
> Thorsten Froehlich
> e-mail: mac### [at] povrayorg
> 
> I am a member of the POV-Ray Team.
> Visit POV-Ray on the web: http://mac.povray.org

Where was mentionned changes in AA with pov 3.6, I didn't saw them in 
"Changes between pov 3.5 & 3.6" on the povray download page !
I'm prbably too idiot to see them... can you explain the changes and why 
this AA 'trace' quicker ?

Post a reply to this message

I hardly use AA nowadays. I use focal blur instead! And photons are MUCH
faster :D

Post a reply to this message

"Thorsten Froehlich" <tho### [at] trfde> wrote in message
news:40caa57d$1@news.povray.org...
> In article <40ca2d06@news.povray.org> , "Ross" <rli### [at] everestkcnet>
> wrote:
>
> > chalk it up to whatever you want (new compiler?, cleaned code?, ...?)
but
> > the official 3.6 binary shaved 19 minutes off of benchmark.pov using the
> > benchmark.ini settings, compared to the official 3.5 binary with the
same
> > official benchmark.pov/benchmark.ini. i never benchmarked pov before, so
i
> > thought i'd do it before I installed 3.6.
>
> This was mentioned bfeore: You *cannot* compare the 3.5 and the 3.6
> becnhmark results due to anti-aliasing changes in POV-Ray 3.6!
>
>     Thorsten

I can compare the time it used to take for me to render a scene to the time
it takes now to render a scene. i'm not trying to dig too deeply, i'm just
saying I experience a rather noticeable speed gain.

what the hell are benchmarks for if not to compare and old something to a
new something?

Post a reply to this message

"Ross" <rli### [at] everestkcnet> wrote in message
news:40cdacbb$1@news.povray.org...
>
> I can compare the time it used to take for me to render a scene to the
time
> it takes now to render a scene. i'm not trying to dig too deeply, i'm just
> saying I experience a rather noticeable speed gain.
>
> what the hell are benchmarks for if not to compare and old something to a
> new something?
>
>
>

bah. it's early.

ignore my grumbling, thorsten.

Post a reply to this message

In article <40cdacbb$1@news.povray.org> , "Ross" <rli### [at] everestkcnet>
wrote:

> what the hell are benchmarks for if not to compare and old something to a
> new something?

You are supposed to use the benchmark to compare computer systems, not to
compare POV-Ray with itself! :-)

    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

Thorsten Froehlich <tho### [at] trfde> wrote:
> You are supposed to use the benchmark to compare computer systems, not to
> compare POV-Ray with itself! :-)

  Comparing the newer version of POV-Ray with the older version is feasible.
If both give approximately the same image but the newer version does so
noticeably faster, that's something positive to notice.
  The newer version being faster than the older version *is* something
to brag about.

  In this case the speedup is mainly due to photons, isn't it? It's
interesting to see in practice how much they have been improved... :)

-- 
plane{-x+y,-1pigment{bozo color_map{[0rgb x][1rgb x+y]}turbulence 1}}
sphere{0,2pigment{rgbt 1}interior{media{emission 1density{spherical
density_map{[0rgb 0][.5rgb<1,.5>][1rgb 1]}turbulence.9}}}scale
<1,1,3>hollow}text{ttf"timrom""Warp".1,0translate<-1,-.1,2>}//  - Warp -

Post a reply to this message

Thorsten Froehlich wrote:
> In article <40cdacbb$1@news.povray.org> , "Ross" <rli### [at] everestkcnet>
> wrote:
> 
> 
>>what the hell are benchmarks for if not to compare and old something to a
>>new something?
> 
> 
> You are supposed to use the benchmark to compare computer systems, not to
> compare POV-Ray with itself! :-)
> 
>     Thorsten
> 
> ____________________________________________________
> Thorsten Froehlich, Duisburg, Germany
> e-mail: tho### [at] trfde
> 
> Visit POV-Ray on the web: http://mac.povray.org

this reminds me of my pure maths teacher at school

if you have something like

prove sin^3 (x) + 3 cos^2 (x) = 5sin(x)cos(x)

i used reduce lhs to 1 and then reduce rhs to 1 or some other term

and then i'd have to write lhs derivation followed by the rhs derivation 
in reverse to make him happy

the system and scene source are constant, therefore running to me at 
least, povray 3.5 versus povray 3.6 with i assume same arguments, is a 
valid performance indication

unless someone can explain and convince me differently ?

stephen

Post a reply to this message