Tweet
hi!
I'm reading documentation about "SolidRocks" plugin, and I've found a tool called "Share Light Cache".
This is what you can read in the documentation:
-------------
Share Light Cache tool is made for network rendering on slaves via the VraySpawner.
This tool speeds up the render slaves reactivity, and helps to reduce render times.
Very often the faster PC is the main one, and slower PCs are used as render slaves.
As you probably know, when you make a distributed render of a still image, each PC calculates exactly the same lightcache.
This lightcache calculation takes some time, and usually the main PC has finished before the slaves.
While the main PC is computing the next pass (irradiance map for example), slaves are still computing their own useless lightcache...
When they're done, they finally start the next pass.
Share Light Cache tool avoids this useless computation of the light cache by slaves, by doing the following :
first, light cache is computed locally on the main PC, and saved in a file
then, a distributed render is launched with previously computed lightcache (in "from file"mode)
the slaves simply load the lightcache file and start immediatly the second pass (irradiance map for example)
With this fully automated process, distributed rendering is now optimised and slaves are more efficient, leading to shorter render times.
--------------
But... I thought that Light Cache (during the DR) is calculated just on the Master Machine, not on the Slaves, only IrradianceMap is calculated from all nodes, isn't so?
Thank you.
I'm reading documentation about "SolidRocks" plugin, and I've found a tool called "Share Light Cache".
This is what you can read in the documentation:
-------------
Share Light Cache tool is made for network rendering on slaves via the VraySpawner.
This tool speeds up the render slaves reactivity, and helps to reduce render times.
Very often the faster PC is the main one, and slower PCs are used as render slaves.
As you probably know, when you make a distributed render of a still image, each PC calculates exactly the same lightcache.
This lightcache calculation takes some time, and usually the main PC has finished before the slaves.
While the main PC is computing the next pass (irradiance map for example), slaves are still computing their own useless lightcache...
When they're done, they finally start the next pass.
Share Light Cache tool avoids this useless computation of the light cache by slaves, by doing the following :
first, light cache is computed locally on the main PC, and saved in a file
then, a distributed render is launched with previously computed lightcache (in "from file"mode)
the slaves simply load the lightcache file and start immediatly the second pass (irradiance map for example)
With this fully automated process, distributed rendering is now optimised and slaves are more efficient, leading to shorter render times.
--------------
But... I thought that Light Cache (during the DR) is calculated just on the Master Machine, not on the Slaves, only IrradianceMap is calculated from all nodes, isn't so?
Thank you.
