Do you think there is a place for a new professional raytracer in the current Game/Film/Architecture market ? Mental image seems to have completly locked the market by integrating their products in current 3D modelers. PS : By raytracer i mean all high quality rendering systems.

Sure there's a place for high-quality rendering tools. The problem is, someone else already holds it.

You'd be extremely hard-pressed to be able to compete with industrial-grade rendering systems. Speed is only a minor issue, as most industrial teams have a lot of computing resources available. Realism can already be acheived by existing solutions, and there is only a tiny window of remaining opportunity to set yourself apart in that category. Usability is painfully subjective, and up against the inertia of the Big Names, it's extremely unlikely that a mass market migration could be sparked on usability alone. Market growth isn't nearly rapid enough to support a product purely on newcomers, and the products tend to tie up large amounts of cash, so existing professionals are not likely to replace their costly investment without an extremely good reason.

It would take a lot of luck and some incredibly good development resources to be able to even start competing in that realm, let alone stay alive as a product line.

Also, many of the algorithms are protected by patent law.

Erff, i wasn't thinking about patent law... How can we know if MLT or MC raytracing are patented ?

ApochPIQ, i love your Freon 2/7 project !!!