LuxRender will keep on improving your image until you tell it to stop, either by pressing the stop button or by specifying how many samples should be calculated.
LuxRender will save the output image at regular intervals in the same folder as where you exported the .lxs file to. It is also possible to copy the rendered image to the clipboard.
By its nature, unbiased rendering tends to longer than other rendering approaches. We do our best to make the program faster, but we will not cut corners on image quality.
First, please note that even on powerful machines, many scenes might take over 12 hours to get satisfactory results at usable resolutions. However, there are several things you can try in order to get satisfactory results quicker:
-use the "Final MLT/Bidir path tracer" rendering preset
-don't use materials with a brightness higher than 80 or 90% as even white objects will always absorb a bit of light.
-for interiors, use portals where you have windows or other openings where light can enter
-if your scene contains glass but you don't need any refractive effect, turn on the "Architectural" option in the glass material.
As a last resort, once you've got a very smooth sampled image with still a bit of grain in it, you can try some denoising filter in image-editing software like Photoshop or Gimp. LuxRender itself already ships with 2 such options under the Noise Reduction tab: Greyscale Restoration and Chiu.
By default, LuxRender uses the Reinhard tonemapping operator. This tonemapper will adjust itself to the brightness of the scene, similar to what automatic photo cameras do. The linear tonemapping operator does not exhibit this behaviour.
There are two ways to use a fake background:
- use an environment map lighting
- render with an alpha channel and add the background in post production. In this case, it's best to use the “premultiply alpha” setting.
Yes, you can. The best way is to select "write/resume FLM" in the exporter, but you can also manually select manually "Save FLM" from LuxRender's File menu. In the latter case, you'll have to choose “Load FLM” and "Resume FLM" next time you start LuxRender.
Efficiency (indicated with “eff” in LuxRender) measures how succesful LuxRender is in finding light in your scene. Technically speaking, it measures the average number of light contributions per sample. With some settings a single sample can result in multiple light contributions, making it possible to reach percentages over 100%.
If your scene gets low efficiency, it means that LuxRender has trouble seeking out the light. There are a number of things you can try to improve the situation:
-use the "Final MLT/Bidir path tracer" rendering preset
-for interiors, use portals where you have windows or other openings where light can enter.
-turn on the "Architectural" option if you're using any glass material
If these settings do not resolve the problem, you might want to consider simplifying the scene or modifying the light setup (for example by introducing additional light sources).
Exposure Value is a photographic measurement. It should allow people with a photographic background to accurately set up linear tonemapping controls.
Please make sure you downloaded a version which is compatible with your processor. If in doubt, try a SSE (not SSE2) build.
If the log contains a message stating "Inconsistent vertex winding, aborting subdivision" it means that two triangles that are connected by a shared edge have normals pointing in different directions, preventing LuxRender to subdivide properly. To resolve this, make sure the normals of connected triangles face a consistent direction.
This effect may happen when using a high sharpness strength in the Mitchell filter. Either use a lower sharpness or use a Gaussian filter instead.
Most likely, the LuxRender folder path in LuxBlend's System tab has not been set up properly. Please click on the “...” button on the right of the “lux binary dir” field and browse to the folder in which LuxRender is located.
LuXSI doesn't perform a check yet if there is no light source or camera exported. Check you scene in XSI and see if the camera is hidden or the light source. You can switch on the "Export hidden objects/lights" box in LuXSI or unhide the things.
Select the object in XSI and change the subdivision level by clicking + on the num-pad or in the GeoApproximation menu
Commercial and freeware products already exist, but there is no unbiased, production-oriented free software renderer around (note: freeware and free software don't mean the same thing).
Choosing the free software way, has, in our opinion, several advantages:
We hope to centralize contributions with an open-source development model: anyone can join the project and add improvements, find bugs, port the project to an unsupported platform or contribute with exporters development and testing. This could be much faster than a "one-man-band". If a lot of open source renderer developers would join forces, there could be a free rendering package with professional features, usability and support (exporters, presets, forum) very soon that could rival commercial ones. In the current state of affairs, there are a lot of small freeware and open source renderers that are used by narrow groups of enthusiasts, but don't find widespread use in the 3d mainstream world.
It can't disappear: freeware developers could get bored and stop development, commercial ones may get bought off by other companies following economical goals or go bankrupt, stopping maintenance or even retiring the software. There are numerous examples of such situations. With LuxRender, GPL code always remains, even if all the original developers leave it, chances are high that when it is used and needed, there will be someone to take care of the code.
Anyone interested can use it at no cost, without any restrictions, presently and in future versions. Free renderers are often used by Blender users. Perhaps because of the internal Blender renderer (no GI), because they love free software or just like experiments. Providing an unbiased GPL renderer allows tight integration between such tools (we hope for a full Blender integration) and guarantees that users will always have a free software choice.
PBRT is a very robust foundation, so we don't have to reinvent the wheel. It has a clean, extensible design, exceptional documentation (the book) and it's used in academic environments, so a lot of students know it.
The goals of both project differ: PBRT is an academical research and educational tool, LuxRender is production and artist oriented: which creates differences in a lot of areas. In LuxRender speed optimization, removal of unnecessary features and adding practical ones (exporters, GUI, multithreading, file formats, MLT...) are the main goal. Design choices are made with the final user and artistic efficiency in mind.
Our fork is authorized by the PBRT authors.