Skip to main content

iray 1: iray in standalone mental ray

* As far as I know, we can only run mental images' iray in NVidia's RealityServer. I hope Autodesk 3dsmax/maya/xsi supports iray soon.

mental images (a subsidiary company of NVidia) demonstrated iray renderer in GDC's keynote and SIGGRAPH ASIA 2009.

GDC NVidia iray demo (YouTube)
mental images iray page

mental ray or some other renderers can simulate indirect illumination in some extent, but the methods final gather or photon map are quite rough approximation. They have no guarantee to converge to the correct answer. Also there are many renderer parameters for performance optimisation. These parameters heavily depends on the scene. It is difficult to achieve high quality and faster rendering.

These methods are anyway time consuming, therefore, we sometimes put invisible light sources by hand, that simulates indirect lights. This iray is a brute force path tracer and there is no renderer parameter, but nowadays computer has not enough power to compute the light transport simulation in straightforward way. Then they use GPU. The demo shown in YouTube uses a GPU cluster. GDC demo shows interactive rendering with such a high quality. It is impressive. But you need 15 Tesla cards. (How much is it?) I heard Tokyo Institute of Technology has Tesla or Fermi cluster, does anybody know do they run iray on that? I am interested in that. SIGGRAPH ASIA 2009's demo showed five Quadro 5800 FX cards machine. (Still I can not afford it. But demo was nice I think.)

iray is a physically correct light transport simulator. At least it is based on more physics than mental ray. SIGGRAPH ASIA demo people explained that iray has only build-in shaders, actually only one build-in shader. Someone asked if he needs to use a custom shader, what iray can do. The answer was:

  1. iray only support this world physics. Unfortunately iray doesn't have a plan to support other physics.

  2. If you really want not non-physical stuffs, please use mental ray or ReallyServer's software renderers.

That's true, they said this is a physical light transport simulator and he asked how can he customize the physics. They also explain that the target customers are architect, car designer, industrial designer, and so on. So not oriented to the people in film production. Since artists rather cares good pictures than physically correctness. mental ray will continue and that's for the artists. After all, it seems

   1. to the people who wants physical simulation: iray
   2. to the people who wants customize the image: software renderer(e.g., mental ray) + MetaSL.

First time people saw iray, they were a bit confused. Why mental images will have two renderes? Because there are different customers and needs.

Comments

Popular posts from this blog

What is shader?

Shader is one of the important component of mental ray. Let me tell about it a bit. A shader is a program code fragment that computes how light interacts with objects. This was the original concept of shader in early computer graphics time and nowadays the role of shader and its meaning are expanded. In one of the common scenario in computer graphics, objects are usually represented as polygons. First, the intensity of each vertex is computed, then inside the polygon is filled by an interpolation method. This becomes more complex as the computer graphics technology developed. There were two shader development directions. One is a hardware rendering that concentrates rendering speed and interactivity. The other is a software rendering that concentrates the image quality. Nowadays graphics hardware is called GPU (Graphics Processing Unit). The early stage GPU has fixed functionalities. Because fixed functionality can be achieved more performance in practice. The implementation was based ...

A note of mental ray/iray camera terminology.

The original mental ray camera is a pinhole camera. Therefore, some of the terminology differs with a normal camera. I will write them down here. aperture: This is the film width size. This has no relationship with lens aperture/opening. mental ray's camera was a pinhole camera, there is no lens aperture. focal: This is the distance from lens to the film (sensor). Pinhole camera always focused, so there is no focal length meaning. But using parameters aperture and focal length, we can compute the horizontal field of view. I think these two parameters exist for this purpose. aspect ratio: This is film(sensor)'s aspect ratio. There is no relationship with pixel resolution and camera geometry.  aspect ratio = film width/film height. Note: If we use aperture, focal, and aspect ratio, we can compute the vertical field of view. mental ray/iray camera model (pinhole camera) without lens shader Current mental ray/iray can have depth of field effect, etc., this termino...

mental ray license bug is fixed on an Intel Core i7 machine

There was a critical license related bug in mental ray 3.7 on Intel Core i7 and the fixed version seems to be distributed now. http://forum.mentalimages.com/showthread.php?t=5511 The fixed mental ray (SPM) version (might depends on OEM) is the following. mental ray 3.7: 3.7.1.31 (2009 September 16) or higher mental ray 3.7+: 3.7.53.6 (2009 October 22) or higher The detail of this problem is: When you want to run mental ray on Intel Core i7 machine, you will get the following error message: Sorry, no license available and can not run mental ray even you have enough licenses left. It seems recent Intel's CPU becomes complex to count the number of cores. This bug seems related with this. http://software.intel.com/en-us/articles/intel-64-architecture-processor-topology-enumeration/ Personally, I don't like the license that limits the number of cores. Because the new machines has more cores, but we can not use these cores because of license shortage. If I buy a new machine...