romullus wips

[Ludvik Koutny]

Aside from rescaling parameters, not... There was some change maybe about a month ago... and from my quick tests, it looks great ever since. It's just fun to play with, but I don't have much time lately. You on the other had, have a WIP thread... ;)

[romullus]

Following discussion in daily builds topic. Never again to badly floating things in CG water :]

[TomG]

Looking good! I can see a movie in this one - "Increasingly abandoned as technology advances, three standard CGI objects set off in search of a new life...."

[Juraj]

The symmetrical flare looks rather fake. We could use mask for more natural shape.

[romullus]

Not sure i understood you correctly. Do you mean radial symmetry? Apart from the fact that every other streak should be slightly smaller and of slightly different shape when flare forms from odd numbered blades, i don't see nothing too wrong with our flare. But maybe you have something else in your mind?

[Juraj]

Well, everything about it is symmetric and even.

I expect soft flare ( +/- F4 in wide angle lens ) to look like this:


And sharp flare ( F11+ in wide angle lens ) to look like this (both are very high-end lens)



We need shape mask support ( .hdr ) for that.

[romullus]

Well, we already have mask - DOF shape custom. Ideally its shape should govern glare. Unfortunatelly those two aren't tied in any way in Corona... :[

Also i'd like to see lens flare too. As well as other optical abberations. maybe someday :]

[Juraj]

Well, we already have mask - DOF shape custom. Ideally its shape should govern glare. Unfortunatelly those two aren't tied in any way in Corona... :[

Also i'd like to see lens flare too. As well as other optical abberations. maybe someday :]

What is the relationship of DOF (you mean BOKEH ? ) mask and glare ? Now I am confused.

You can get the example I posted above with VrayLens effects (or VFB+) is using glare mask. It doesn't affect bokeh shape.

[romullus]

AFAIK starburst shape depends on aperture shape and aperture in Corona is called DOF mask. The more rounded aperture shape, the less prominent is glare streaks and vice versa. Obviously in Corona aperture blades are considered perfectly poligonal, so we have that very symetrical somewhat less real looking glare. Although in real life sun flare is getting very symetrical and regular when aperture is closed to maximum and thus its shape becomes polygonal.

[Juraj]

Aperture does affect the size/shape (hence I wrote the f-stops ), but lot other shape factors are due to lensing (diffraction, irregularities, obstacles, etc..). Different lenses provide wildly different glare streaks on same aperture shape (which is quite identical across modern high-end DSLRs, hence why people put in custom shape cutouts for "artistic" effect in real-life).

It does not substitute glare mask, which is simple solution to complicated lens simulation.

[romullus]

Just a bunch assets from free megascan samples thrown together.

[antanas]

Well that's what I keep telling all my cg artist friends I've got for at least 4-5 years - photogrammetry Rules and it is a 3d modelling's future, well for some things at least and at least a decent part of that future, and now when megascans finally "hit the shelves" and they finally begin to realize that, it feels so good to be right for a change ))

[romullus]

Recently did some sci fi gates for client and decided to render it for myself. Render straight out of VFB. Who needs photoshop when Corona does everything you need.

[maru]

Render straight out of VFB.

Ok I am officially starting counting these sentences. :D

[romullus]

I think after 1.5 release you will have to count instances where usage of PP software is mentioned - it will be easier that way :]