So in the process of recreating this process realistically I should do lens flare after motion blur ? How about DoF ?
Don't know, depends on how your graphics pipeline is setup. In theory motion blur pretty much consists of averaging the render over some time duration, but in practice there are lots of hacks to make stuff run faster. I doubt most people would notice the lens flare getting blurred anyway, the displacement is generally not fast enough (remember objects far away where the diffraction effect is strongest do not actually move quickly on screen, whereas closer/larger objects do not have such distinctive lens flare features).
As for DoF, that's a tricky one, because it depends on the current focal length which can change during fast motion. But if the DoF is constant, it doesn't really matter either way because objects that are moving (and hence susceptible to a change in appearance with depth) are going to be blurred anyway, far more than what depth of field would normally do.
So, no, do stuff before motion blur if you can, but if it doesn't look good or looks unnatural you can always do it after.
Unless this isn't for real time but you're going for true realism..
Edited by Bacterius, 22 May 2013 - 06:59 AM.
The slowsort algorithm is a perfect illustration of the multiply and surrender paradigm, which is perhaps the single most important paradigm in the development of reluctant algorithms. The basic multiply and surrender strategy consists in replacing the problem at hand by two or more subproblems, each slightly simpler than the original, and continue multiplying subproblems and subsubproblems recursively in this fashion as long as possible. At some point the subproblems will all become so simple that their solution can no longer be postponed, and we will have to surrender. Experience shows that, in most cases, by the time this point is reached the total work will be substantially higher than what could have been wasted by a more direct approach.
- Pessimal Algorithms and Simplexity Analysis