Like the rest of WP 1, task 1.3 concerns itself with the interface between the software (algorithms) under development, and the final users of any systems that will use the software. Unlike the other two WP 1 tasks, which focus on the input requirements of the software, task 1.3 focuses on the output requirements. Task 1.1 ensures that users are able to process the geometry they wish to see illuminated, and task 1.2 ensures that users are able to specify the light sources that will illuminate the scene. Task 1.3, in contrast, focuses further down the pipeline - the radiosity solution is now available to display/query - the question is; how do we wish to convey the solution to the user of the system? A number of solutions present themselves:

LightWork Design have made an extensive study of typical user requirements, over the course of the ARCADE project. LightWork Design technical staff have browsed pertinent newsgroups and user groups, and have examined the functionality of existing lighting simulation software, to try and establish what a user's display requirements might be. LightWork Design marketing staff have made similar enquires of potential customers, and LightWork Design support staff have been consulted, to establish what their (very accurate) perception is, of the needs of existing customers. LightWork Design marketing have recently carried out a survey of existing customers, seeking to establish what functionality they would wish to utilise, that is not currently available within the LightWorks ADS.

The result of this work is that users in different market sectors have differing requirements. Whilst LightWork Design software could already cope with most of the output requirements listed above, we were finding (much as IGD did) that tone mapping functionality would be essential, if we were to produce realistic-looking renderings of the radiosity solution.

Now, LightWork Design's task 1.3 work effort will concentrate on developing perceptually-based code that will map large luminance ranges onto those of common display devices, and to alter the shading pipeline, within the LightWorks ADS, to utilise this new software. Not only does this meet the needs of LightWork Design users, but it also agrees with IGD's perception of where our common output model was falling short of the mark.

Some examples of LightWork Design's tone mapping work, can be found here.

Photometric consistency and calibration

To investigate the effects of various display devices, concerning the reliability of the qualitative and quantitative luminance values presented, a study on luminance on large-scale immersive projection devices has been performed by IGD [SKMU99]. It was shown, that the luminance range and the colour range of several displays (monitor, stereo screen, CAVE) are already sufficient to exactly reproduce certain interesting lighting scenarios (e.g., car interior lighting, by night) without the need for any tone mapping algorithms. As a consequence, displays of these simulated environments are perceived almost identically to the way the true environment is perceived. Some requirements for the calibration process and the consistency of the radiosity process have been identified. The result is a specification of a set of libraries and tools, which will be described in detail in deliverable 1.3-2:

In co-operation with BMW and Daimler-Chrysler, test scenarios will be established to compare radiosity results with measured data sets. First, radiance images from the real car interior will be generated with the approach from Debevec [DM97] and the camera position will be reconstructed. Then, a radiosity simulation will be performed based on a CAD model of the same car and presented for the same camera position. Finally, the results will be shown on large screen projections (CAVE and HOLOBENCH) and radiance pictures will be captured again with the Debevec approach. All results will be compared in order to optimise the radiosity approach.

In addition, tone-mapping algorithms will be evaluated for scenes with high dynamic luminance ranges that cannot be reproduced by the displays. This work will be performed in co-operation with Max-Planck-Institute of Computer Graphics in Saarbrücken and is not part of the ARCADE work plan. However, the results will be integrated into the final ARCADE deliverables, where they will help to create a truly usable radiosity system.

RedLine

Restricted information.

© LightWork Design Ltd. All information contained in
this document is the copyright of LightWork Design Ltd. (unless stated
otherwise) and may not be used or reproduced with prior written consent.

RedLine