 |
|
|
|
|
|
| |
| |
|
|
|
|
| |
| |
|
|
Hello..
it seems that it is always myself complaining about the same thing?
anyway I made my first scene in pov 3.5 I just got today: checker,
sphere, area light.
render. dog slow. 4 minutes andsome seconds
I take the scene and render it in megapov latest version (I can't
compare to pov3.1 since that aerealight wasn't supppoerted) and i get 6
seconds
i then render a trivial sample scene and i get 16 minutes... my G3 slow
as a quadra??? the same scene on pov 3 took less on my quadra!
now it is sooo lae and it's better to go to bed... but I read that
povray should be faster and more stable than megapov while using the
patches and I hoped the rest would remain unchanged!
-r-
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <20020711014929711+0200@news.povray.org> , WhiteGandalf
<zus### [at] libero it> wrote:
> anyway I made my first scene in pov 3.5 I just got today: checker,
> sphere, area light.
> render. dog slow. 4 minutes andsome seconds
Turn off the freaking preview as the documentation tells you! The Mac 3.5
version is significantly faster than any version ever out there unless you
of course screw up your settings so much that you end up comparing Apples
and Bananas...
Thorsten
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
> Turn off the freaking preview as the documentation tells you! The Mac
> 3.5 version is significantly faster than any version ever out there
> unless you of course screw up your settings so much that you end up
> comparing Apples and Bananas...
yes, ideed. Sorry I didn't read the docs, but I was just eager to try
out the new version
So these are the timings. I made a little scene using area lights with
some new features and a sphere:
PoV 3.5
preview: 4 minutes
without: 18 sec
(dramatic increase)
MegaPov latest rev with with patches turned on (so to use the bnew
arealight as in PoV)
preview: 6 sec
without: 5 sec
so definitely preview is not something for the final render. But I quite
like the progress bat, it adds a "mac" look.
-ric
PS: in the preferences my version says it is a beta built 8 Jul with
timeout july 15 (which is tomorrow..) is this just an old thing or was
the official mirror from where I did download not updated???? maybe I'm
using a beta!
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <20020714195121389+0200@news.povray.org> , WhiteGandalf
<zus### [at] liberoit> wrote:
> PoV 3.5
> preview: 4 minutes
> without: 18 sec
> (dramatic increase)
>
> MegaPov latest rev with with patches turned on (so to use the bnew
> arealight as in PoV)
> preview: 6 sec
> without: 5 sec
There is still something wrong with your measurement. First of all, are you
running on Mac OS Xor Mac OS 9? You should really do benchmarking on Mac oS
9. Anyway, that said, you also can only benchmark if you actually run for
more than a few seconds. Something like 10 minutes or so. Right now you
are measuring everything, but not render speed...
> PS: in the preferences my version says it is a beta built 8 Jul with
> timeout july 15 (which is tomorrow..) is this just an old thing or was
> the official mirror from where I did download not updated???? maybe I'm
> using a beta!
No, it is just that one string taht didn't get changed.
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
|
|
| |
| |
|
|
|
|
| |
| |
|
|
To be more precise, try a more complex scene that simply takes longer. One
that will run on much longer. That would be the typical classical benchmark
scenes like skyvase.pov, chess2.pov or of course (but 3.5 specific)
benchmark.pov, all of which may be found in the:scenes:advanced: folder.
Oh, and of course, don't forget when under Mac OS 9 to set the application
friendliness in the POV-Ray prefs dialog to "Tyrannical". And while you are
there, turn of the "Use Temporary Memory" feature (you need to restart
POV-Ray for the memory changes to take effect).
Then render at a resolution of at least 640*480, and make sure nothing
secondary is running. In particular turn off file-sharing and virtual
memory!
Oh, and in MegaPOV, set the Cooperation to 11 and turn Verbose on, which
then approximately equals the POV-Ray 3.5 settings.
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
|
|
| |
| |
|
|
|
|
| |
| |
|
|
>
> There is still something wrong with your measurement. First of all,
> are you running on Mac OS Xor Mac OS 9? You should really do
> benchmarking on Mac oS 9. Anyway, that said, you also can only
> benchmark if you actually run for more than a few seconds. Something
> like 10 minutes or so. Right now you are measuring everything, but
> not render speed...
I am running X (8.6 is on another computer). Both programs were executed
under the most similar conditions, one after the other. Same settings,
same file.
I know it isn't a definitve benchmark, it was just a quick test!
As for benchmarks, I ried to compile pov from the spources (from 3.1)
under apple cc, I made several compiles using every possible combination
of optimizations, the result was that megapov still was an edge faster (
on skyvase). This is somehow strange, since pov from command line should
be advantaged.
I will render skyvase as soon as I have time.
bye
-ric-
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <20020715222950264+0200@news.povray.org> , WhiteGandalf
<zus### [at] liberoit> wrote:
> I am running X (8.6 is on another computer). Both programs were executed
> under the most similar conditions, one after the other. Same settings,
> same file.
> I know it isn't a definitve benchmark, it was just a quick test!
>
> As for benchmarks, I ried to compile pov from the spources (from 3.1)
> under apple cc, I made several compiles using every possible combination
> of optimizations, the result was that megapov still was an edge faster (
> on skyvase). This is somehow strange, since pov from command line should
> be advantaged.
Out of curiosity, do you see a spinning beachball when rendering with
MegaPOV and switching between applications?
____________________________________________________
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
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <20020715222950264+0200@news.povray.org>,
WhiteGandalf <zus### [at] liberoit> wrote:
> As for benchmarks, I ried to compile pov from the spources (from 3.1)
> under apple cc, I made several compiles using every possible combination
> of optimizations, the result was that megapov still was an edge faster (
> on skyvase). This is somehow strange, since pov from command line should
> be advantaged.
You forgot to take the compiler into account...the version of gcc in the
development tools is nowhere near as good at optimization as Metrowerks
CodeWarrior (which is used for the official version). The next update of
the development tools will use a more up to date version of gcc which is
apparently much better at optimization of PPC code, I don't know how it
compares to CodeWarrior. If you want to compare 3.1 and MegaPOV, compile
MegaPOV with gcc as well.
And anyway, being a GUI program won't really affect the render
speed...you probably ran the command line version from Terminal.app,
which (obviously) is a GUI program itself. The only way there could be a
significant difference is if you booted up without any GUI at all, which
I don't consider worth the few % of CPU time gained. Just being a CLI
program doesn't give any special advantages.
There may be something specific to how POV-Ray Mac works though...maybe
the way it handles events uses up CPU time or prevents it from using
100% of available cycles.
--
Christopher James Huff <chr### [at] maccom>
POV-Ray TAG e-mail: chr### [at] tagpovrayorg
TAG web site: http://tag.povray.org/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
some benchs:
skyvase, with original antialiasing rendered. I set also the bounding
options in the same manner.
megapov: 1 min 46 sec
pov 3.5: 6 min 54
> Out of curiosity, do you see a spinning beachball when rendering with
> MegaPOV and switching between applications?
yes and no. Not while I render. But if the preference pane is open when
it needs to be refreshed (ie. becoming oncovered from a window or such)
it takes a long time and the spinning ball appears.) This independently
from the CPU occupation. I wanted to report this to smellenbergh, but i
really forgot!
-ric-
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <20020716014246106+0200@news.povray.org> , WhiteGandalf
<zus### [at] liberoit> wrote:
> some benchs:
> skyvase, with original antialiasing rendered. I set also the bounding
> options in the same manner.
>
> megapov: 1 min 46 sec
> pov 3.5: 6 min 54
You had the preview on again, right?
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
|
|
| |
| |
|
|
|
|
| |
| |
|
|
I do not see the point in using the program without the preview for i use
it as well as partial rendering to finish most all of my projects even the
most simple ones... Once again i will stick to Mega Pov... =]
WhiteGandalf wrote:
> Hello..
>
> it seems that it is always myself complaining about the same thing?
>
> anyway I made my first scene in pov 3.5 I just got today: checker,
> sphere, area light.
> render. dog slow. 4 minutes andsome seconds
>
> I take the scene and render it in megapov latest version (I can't
> compare to pov3.1 since that aerealight wasn't supppoerted) and i get 6
> seconds
>
> i then render a trivial sample scene and i get 16 minutes... my G3 slow
> as a quadra??? the same scene on pov 3 took less on my quadra!
>
> now it is sooo lae and it's better to go to bed... but I read that
> povray should be faster and more stable than megapov while using the
> patches and I hoped the rest would remain unchanged!
>
> -r-
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In <3d33b973@news.povray.org> Thorsten Froehlich wrote:
>> megapov: 1 min 46 sec
>> pov 3.5: 6 min 54
>
> You had the preview on again, right?
no it was off in both programs, I saved the file directy to disk. Ypou
advised me to test it such way.
-r-
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In <3d33b973@news.povray.org> Thorsten Froehlich wrote:
>> megapov: 1 min 46 sec
>> pov 3.5: 6 min 54
>
> You had the preview on again, right?
no it was off in both programs, I saved the file directy to disk. Ypou
advised me to test it such way.
-r-
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In <chr### [at] netplexaussieorg> Christopher
James Huff wrote:
> You forgot to take the compiler into account...the version of gcc in
> the development tools is nowhere near as good at optimization as
> Metrowerks CodeWarrior (which is used for the official version). The
> next update of the development tools will use a more up to date
> version of gcc which is apparently much better at optimization of PPC
> code, I don't know how it compares to CodeWarrior. If you want to
> compare 3.1 and MegaPOV, compile MegaPOV with gcc as well.
This exaclty was the point. I used gcc and complained about it!
> And anyway, being a GUI program won't really affect the render
> speed...you probably ran the command line version from Terminal.app,
> which (obviously) is a GUI program itself. The only way there could be
> a significant difference is if you booted up without any GUI at all,
> which I don't consider worth the few % of CPU time gained. Just being
> a CLI program doesn't give any special advantages.
Well, that's not completely true, after all without preview povray isn't
doing much.
Specifcally I run my CLI compile both in terminal.app and then i booted
console only. Appple did a fine work, no noticeable difference (under 1%).
This of course if you leave the GUI in peace, if you move windows around
it isn't fair.
Being a CLI program shouldn't give any significant advantages during
render, but it was one way to test gcc from orignal sources.
And I must add, CodeWarrior, at least the older versions, isn't such a
speedy compiler. Under os classic where MPW was available it was
consistently outperformed as T. Fr?hlich will remind. He sent me an
unofficial MPW compile and it was quite faster.
> There may be something specific to how POV-Ray Mac works though...
> maybe the way it handles events uses up CPU time or prevents it from
> using 100% of available cycles.
strictly speaking "cycles" doesn't make much sense in a unix environment.
It is not a single program that decides to acquire or release the CPU
but the kernel.
-ric-
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <chr### [at] netplexaussieorg> ,
Christopher James Huff <chr### [at] maccom> wrote:
> There may be something specific to how POV-Ray Mac works though...maybe
> the way it handles events uses up CPU time or prevents it from using
> 100% of available cycles.
No, but under Mac OS X one has to give control to the system as frequently
as possible to prevent the spinning beachball cursor which effectively locks
the user out of the program. So under mac OS X POV-ray 3.5 for Mac OS is
forced to be more friendly than under Mac OS 9.
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
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <3D347959.649CC718@earthlink.net> , Ryan Mooney
<rdm### [at] earthlinknet> wrote:
> I do not see the point in using the program without the preview for i use
> it as well as partial rendering to finish most all of my projects even the
> most simple ones... Once again i will stick to Mega Pov... =]
Nobody forces you to use POV-Ray 3.5.
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
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <20020716223638440+0200@news.povray.org> , WhiteGandalf
<zus### [at] liberoit> wrote:
>>> megapov: 1 min 46 sec
>>> pov 3.5: 6 min 54
>>
>> You had the preview on again, right?
>
> no it was off in both programs, I saved the file directy to disk. Ypou
> advised me to test it such way.
Well, then you obviously tested with completely different settings.
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
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <3d348cd6@news.povray.org>,
"Thorsten Froehlich" <tho### [at] trfde> wrote:
> No, but under Mac OS X one has to give control to the system as frequently
> as possible to prevent the spinning beachball cursor which effectively locks
> the user out of the program. So under mac OS X POV-ray 3.5 for Mac OS is
> forced to be more friendly than under Mac OS 9.
And I guess you probably can't fix this without making it Mac OS X
only...
--
Christopher James Huff <chr### [at] maccom>
POV-Ray TAG e-mail: chr### [at] tagpovrayorg
TAG web site: http://tag.povray.org/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <20020716230156187+0200@news.povray.org>,
WhiteGandalf <zus### [at] liberoit> wrote:
> In <chr### [at] netplexaussieorg> Christopher
> James Huff wrote:
> > You forgot to take the compiler into account...the version of gcc in
> > the development tools is nowhere near as good at optimization as
> > Metrowerks CodeWarrior (which is used for the official version). The
> > next update of the development tools will use a more up to date
> > version of gcc which is apparently much better at optimization of PPC
> > code, I don't know how it compares to CodeWarrior. If you want to
> > compare 3.1 and MegaPOV, compile MegaPOV with gcc as well.
>
> This exaclty was the point. I used gcc and complained about it!
You said "pov from command line should be advantaged.", which seemed to
say you were comparing gcc-compiled POV 3.1 against the GUI MegaPOV,
which is compiled with CodeWarrior. What exactly were you doing?
> Well, that's not completely true, after all without preview povray isn't
> doing much.
Uh, what? You aren't supposed to have the preview on for benchmarks
anyway. Of course a preview will eat up CPU cycles...but without preview
it is still a GUI program.
> And I must add, CodeWarrior, at least the older versions, isn't such a
> speedy compiler. Under os classic where MPW was available it was
> consistently outperformed as T. Fr?hlich will remind. He sent me an
> unofficial MPW compile and it was quite faster.
True, but it has gotten much better over time...I'm not sure which is
faster now, but they are close, and both are much better than gcc 2.95.2.
> strictly speaking "cycles" doesn't make much sense in a unix environment.
> It is not a single program that decides to acquire or release the CPU
> but the kernel.
I am aware of that, but POV-Ray Mac works in the Carbon environment, and
is written to be compatible with pre-OS X versions of Mac OS. This
probably makes it less efficient (having to constantly stop rendering
and check for events to stay responsive) and might prevent it from
getting all the available time. The command-line version shouldn't have
this problem...at least my command line programs have no trouble pinning
the processor use at 100%.
--
Christopher James Huff <chr### [at] maccom>
POV-Ray TAG e-mail: chr### [at] tagpovrayorg
TAG web site: http://tag.povray.org/
Post a reply to this message
|
|
| |
| |
|
|
|
|
| |
| |
|
|
In article <chr### [at] netplexaussieorg> ,
Christopher James Huff <chr### [at] maccom> wrote:
>> No, but under Mac OS X one has to give control to the system as frequently
>> as possible to prevent the spinning beachball cursor which effectively locks
>> the user out of the program. So under mac OS X POV-ray 3.5 for Mac OS is
>> forced to be more friendly than under Mac OS 9.
>
> And I guess you probably can't fix this without making it Mac OS X
> only...
Yes and no. With Mac OS 9.1 and later it is possible to create MP-services
threads (which are preemptive) that do file-io. So it is not impossible to
make the render engine run in a preemptive thread on those Mac OS versions.
Then, under Mac OS X while all the function calling restrictions still
apply, it would also run preemptively just like a Cocoa preemptive thread.
That way it should get absolutely all available CPU resources (of one CPU,
of course).
In fact, I have such an experimental version and it will run just fine on
one processor of the dual G4 and leave the GUI on the other processor. So
the GUI is as responsive as without rendering, and the Mac (under Mac OS
that is) is fully usable. Even better, this way one can run two instances
of POV-Ray to use both processors just like one can currently in the Windows
version.
However, this version is not ready for prime-time yet, and won't for another
few month. It may (note that I say may *not* will) be ready when there will
be a POV-Ray 3.5.1 or whatever the maintenance releases will be called.
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
|
|
| |
| |
|
|
|
|
| |
