What are the system requirements for Kurtis ?
Can I use standard and Lite versions without Kurtis Creative ?
Yes, we provide more than 120 example nodes under “modified BSD License”.
But you must have at least 1 Kurtis Creative License to create you own nodes in order to use them with other versions.
Is Kurtis Lite available ?
Not yet. Kurtis Lite (with a simplified graphic user interface) will be available in Q1-2018. The price of this version will remain the same. At the moment, you can use Kurtis API. It allows an R&D team to create plugins for third-party software and communicate with Kurtis.
How to buy Kurtis ?
Is there a trial software version?
Because Kurtis is a pipeline software and requires advanced knowledge, there is no trial downloadable version.
Please contact us for testing Kurtis. For now, we provide an evaluate version only for studios and senior TDs. We are working on an online test version.
What are the available licenses ?
Kurtis is available only as a floating license (not nodelock) for renting and buying.
There are 4 license types :
¹5 render licenses are provided with Kurtis Creative.
3 render licenses are provided with Kurtis Standard.
1 render license is provided with Kurtis Lite.
Please contact us about :
- additional render licenses
- educational licenses for schools and universities
- site license for big studios
- multi-site licenses for international studios
How to install the software ?
We provide client and server installers.
2 options are possible :
- install the client software on your network for all users.
- install the client software on each computer.
You don’t need to install Kurtis on your render farm.
How Kurtis sends processing graphs to a dispatcher ?
Definition of a Graph
By graph, we mean a series of interconnected nodes arranged in a directed acyclic graph (DAG). Each node contains a Python code which evaluates inputs and produces outputs, which can be numeric values, text, files, etc… The graph can be executed to perform complex processing of data.
Graph Execution Patterns
The three patterns below can, of course, be used together in the same graph.
Execution on a Remote Machine
All nodes in the graph are sent to a render farm and executed on a single machine corresponding to the user’s needs (RAM, CPU, etc…). The graph is executed on a single machine, but nodes in the graph can contain multi-threaded code.
Execution of a Node on Several Machines
A Maya rendering node, for example, might be identified by the dispatcher as a scene whose images will be computed on a render farm by packets (each machine might render 10 images, for example).
Execution of Several Graph Branches
The branches can be evaluated sequentially (each branch is sent to the render farm in turn) or in parallel (branches are sent to the render farm at the same time, and evaluation of the graph proceeds when they have all finished).
Kurtis provides a simple API to send a processing graph to your own dispatcher. We provide a simple implementation example for Royal Render, Deadline or local execution in multi-processing.
To interface with a dispatcher, we need, at the very least, an API (in Python, ideally) allowing you to:
- execute a Python script with parameters
- retrieve a unique ID for a submitted job
- ask the dispatcher whether the job with the given ID has finished
The following, optional functions are also useful:
- retrieve the execution logs for a job
- query the status of a job
- define dependencies among jobs (i.e. wait for A and B to finish before running C)
Do you offer consulting services ?
Yes, you can contact us about consulting services.