|
It appears that InstallShield, on some configurations, cannot deal with the large install size of modo 302. If you experience this error, please click here to contact us and we help you through this install.
(Update posted 4/4/08 12:00 p.m.) |
|
First, you need to create user account and make sure you express that preference in the sign-up form.
If you are already a Luxology community member and do not get the Newsletter,
please go into My Account » Preferences and make sure you have
"Subscribe to Luxology Newsletter" checked ON.
Also, make sure your spam filter allows email from luxology.com to arrive in your inbox.
|
|
modo‘s built in distributed rendering is based on Bonjour. Bonjour comes with every Mac computer and is bundled with the Windows version of modo. You can use a combination of Windows and Mac machines to render across a network. Once a Bonjour connection is working between your master computer and other workstations that have modo running in “Slave” mode, modo will show the buckets being computed by other computers as different colors in the render window on the Master system.
modo can also be run in a headless or command line fashion, which can be handy for rendering.
The built-in network rendering in modo is best suited to fewer large frames that take a long time to render, and is especially useful when rendering large stills. The biggest gains with modo's built-in network rendering will be for high resolution images that are dominated by bucket rendering time (as opposed to pre-pass time), such as scenes using non-cached indirect illumination, blurry reflections, depth of field, subsurface scattering, complex refraction, etc.
If you have lots of small, fast frames to render the built-in network rendering is not ideally suited for this. These are best rendered with a conventional “one frame per processor” render controller, from a 3rd party such as the ones from GwynneR, Paul Lord and others. These control modo in a more conventional manner, where each instance of modo/modo_cl is operating independently of the others, but in unison with a controller that divvies out specific frames for each individual machine.
Some history should be noted: In modo 301, every modo “slave” had to compute all the irradiance cache pre-passes, which meant that slaves sometimes never had time to contribute any finished buckets if they were slower than the master. In modo 302, the master computes the pre-passes and then sends the cache to the slaves, so slower slaves can get started on the buckets sooner. This is less optimal than the 301 method if the slaves machines are faster than the master, but we found that most people use their fastest machine as the master. More importantly, the new method allows network rendering to take advantage of Walkthrough Mode.
|