03-21-2013, 07:16 PM
Oort,
I've been having fun with the YafaRay plug-in too! I'm going to get it to work for me if I bust!
Thanks to wegg's comments above, I fixed the Registry entry.
The plug-in installed correctly from the .tar file, but only when I copied the .tar file into the "..\wings-1.4.1\plugins" folder (- it did not like loading from a network folder for some reason).
It took a while to figure out the YafaRay problem with spaces in the path name - but this seems to have been be an old problem with Yafray too! (I read your old Handbook and there are hints there). So the answer is - install Yafaray & save Wings-Yafaray render files to folders with no spaces.
But the plug-in is hooked up now and it's rendering, and now I know what to watch out for.
Although these can be added to the options string at the moment, I think you could add some on the yafaray-xml.exe command options as separate boxes in the Plug-in preferences dialog. There seem to be some useful options, for example:
-op, --output-path for rendered images
-f, --format of image file (jpg, tga, png, etc)
-t, --threads ( note -1 is auto selection)
-a, --with-alpha saving the image with alpha channel.
-pp --plugin-path (I think this can be deduced anyway since it's part of the YafaRay folder)
(I appreciate that you were probably going to do this anyway)
BTW: I also looked at the performance of the thread count option - by adding -t 2 to the Options string (also -t 3 and -t 4). I stopped at 4 threads on my 4-core system.
Mmmmm? Interesting!
On my system it seems it's faster to render with the default single thread than with 2, 3, 4 threads! Which suggests that there's something iffy with the thread model on Yafaray and/or Python (I'm using 2.7) on my system. Using the same very simple scene and parameters:
Threads Render Time (s)
1 3.198
2 4.695
3 5.257
4 5.382
But this is early days yet and not under fully controlled conditions, so needs further investigation.
Looks like wegg's machine is a AMD(?) 6-core beast which is why he gets only 17% utilisation. My system an Intel i5 4-core - so I'm getting 25% out of my system on YafaRay's single thread. (Think about it :-) )
gerytery
I've been having fun with the YafaRay plug-in too! I'm going to get it to work for me if I bust!
Thanks to wegg's comments above, I fixed the Registry entry.
The plug-in installed correctly from the .tar file, but only when I copied the .tar file into the "..\wings-1.4.1\plugins" folder (- it did not like loading from a network folder for some reason).
It took a while to figure out the YafaRay problem with spaces in the path name - but this seems to have been be an old problem with Yafray too! (I read your old Handbook and there are hints there). So the answer is - install Yafaray & save Wings-Yafaray render files to folders with no spaces.
But the plug-in is hooked up now and it's rendering, and now I know what to watch out for.
Although these can be added to the options string at the moment, I think you could add some on the yafaray-xml.exe command options as separate boxes in the Plug-in preferences dialog. There seem to be some useful options, for example:
-op, --output-path for rendered images
-f, --format of image file (jpg, tga, png, etc)
-t, --threads ( note -1 is auto selection)
-a, --with-alpha saving the image with alpha channel.
-pp --plugin-path (I think this can be deduced anyway since it's part of the YafaRay folder)
(I appreciate that you were probably going to do this anyway)
BTW: I also looked at the performance of the thread count option - by adding -t 2 to the Options string (also -t 3 and -t 4). I stopped at 4 threads on my 4-core system.
Mmmmm? Interesting!
On my system it seems it's faster to render with the default single thread than with 2, 3, 4 threads! Which suggests that there's something iffy with the thread model on Yafaray and/or Python (I'm using 2.7) on my system. Using the same very simple scene and parameters:
Threads Render Time (s)
1 3.198
2 4.695
3 5.257
4 5.382
But this is early days yet and not under fully controlled conditions, so needs further investigation.
Looks like wegg's machine is a AMD(?) 6-core beast which is why he gets only 17% utilisation. My system an Intel i5 4-core - so I'm getting 25% out of my system on YafaRay's single thread. (Think about it :-) )
gerytery