From .flm to .exr or .png16bit

[vza]

Thank you!

[Dade]

pyluxcoretools merge now supports saving film AOVs via new --aov-output option, for instance:

Code: Select all

david@Desktop-3900x:~/projects/luxcorerender/LuxCore$ python3 lib/pyluxcoretools.zip merge --aov-output ALBEDO albedo.png film.flm 
[MainThread][2020-05-07 12:15:11,142] LuxCore 2.4alpha0
[MainThread][2020-05-07 12:15:11,143] Processing the base Film: film.flm
[MainThread][2020-05-07 12:15:11,341] Resulting film statistics:
[MainThread][2020-05-07 12:15:11,341]   Samples per pixel: 42.588431249999999
[MainThread][2020-05-07 12:15:11,341]   Radiance group count: 1
[MainThread][2020-05-07 12:15:11,341] [LuxCore][0.200] Outputting film: albedo.png type: 31
[MainThread][2020-05-07 12:15:11,366] Done.

The AOVs names can bee seen here: https://github.com/LuxCoreRender/LuxCor ... md.py#L147

[vza]

Wow!!!!! Thank you. I’ll give you soon a feedback!

[vza]

I have just tried to export ALBEDO in two ways like in the example you posted here but no output for me. Am I wrong? Can you help me?

1)
python3 /Applications/LuxCoreRender/pyluxcore/pyluxcoretools.zip merge -a ALBEDO ParzialeAlbedo.png -f /Users/vza/Desktop/Render3/Fiordo7_Dettaglio_LuxCore/00001/film.flm
[MainThread][2020-05-07 21:18:30,347] LuxCore 2.4alpha0
usage: pyluxcoretools.zip [-o FILE_NAME] [-f FILE_NAME] [-h]
[-a AOV_NAME FILE_NAME]

PyLuxCoreMerge

optional arguments:
-o FILE_NAME, --image-output FILE_NAME
Save the RGB_IMAGEPIPELINE film output to a file
-f FILE_NAME, --film-output FILE_NAME
Save the merge film to a file
-h, --help Show this help message and exit
-a AOV_NAME FILE_NAME, --aov-output AOV_NAME FILE_NAME
Save the merge film AOV to a file
usage: pyluxcoretools.zip [-p] [-s]
[-r SRC_OFFSET_X SRC_OFFSET_Y SRC_WIDTH SRC_HEIGHT DST_OFFSET_X DST_OFFSET_Y]
fileFilm

Film Options

positional arguments:
fileFilm .cfg, .flm or .rsm files with a film

optional arguments:
-p, --pixel-normalized-channel
The film will have
CHANNEL_RADIANCE_PER_PIXEL_NORMALIZED (required by all
render engines)
-s, --screen-normalized-channel
The film will have
CHANNEL_RADIANCE_PER_SCREEN_NORMALIZED (required by
BIDIRCPU and LIGHTCPU render engines)
-r SRC_OFFSET_X SRC_OFFSET_Y SRC_WIDTH SRC_HEIGHT DST_OFFSET_X DST_OFFSET_Y, --region SRC_OFFSET_X SRC_OFFSET_Y SRC_WIDTH SRC_HEIGHT DST_OFFSET_X DST_OFFSET_Y
Define the origin and the size of the region in the
source film and the placement in the destination film
where the it will be merged

!!-------------------------------------------------------------------------!!
!!-------------------------------------------------------------------------!!

2)

python3 /Applications/LuxCoreRender/pyluxcore/pyluxcoretools.zip merge --aov-output ALBEDO ParzialeAlbedo.png /Users/vza/Desktop/Render3/Fiordo7_Dettaglio_LuxCore/00001/film.flm
[MainThread][2020-05-07 21:20:28,976] LuxCore 2.4alpha0
[MainThread][2020-05-07 21:20:28,978] Processing the base Film: /Users/vza/Desktop/Render3/Fiordo7_Dettaglio_LuxCore/00001/film.flm
[MainThread][2020-05-07 21:20:29,889] Resulting film statistics:
[MainThread][2020-05-07 21:20:29,889] Samples per pixel: 41.952010030864194
[MainThread][2020-05-07 21:20:29,889] Radiance group count: 4
************* no output here************************
[MainThread][2020-05-07 21:20:29,889] Done.

[Dade]

The second looks correct to me but I have the feeling your .flm doesn't have an ALBEDO AOV inside. Tomorrow I will add some check so you get an warning message in that case. Can you do a test with a small .flm and post the file ?

[vza]

I think the problem is in 2.40 standalone because I can open the app but it can't load .cfg file and then I have to kill the process and restart my mac because some other application (Blender, Safari) don't run anymore.

[Dade]

I think the problem is in 2.40 standalone because I can open the app but it can't load .cfg file and then I have to kill the process and restart my mac because some other application (Blender, Safari) don't run anymore.

This is likely to be the MacOS OpenCL broken drivers taking 1 hour to compile the kernel. It shouldn't happen if you render with the CPU.

[vza]

You are right! With CPU it works fine, so my solution is to use BlendLuxCore v2.3 with GPU for normal scenes and v.2.4 (CPU) for super-complex scenes in network render and extract passes from merged .flm.
However I think I won't be able to use BlendLuxCore with openCL from version 2.4 and later.
Thank you very much!