glossy rays question

for a camera flythrough (no moving objects) and using IR/LC precalced maps but scene has lots of glossy reflections that must be sharp and accurate, what is the best way to calcualate those? if ‘use LC for glossy rays’ is off when the LC map is calced, what part of vray is calculating the glossies then? would it be done during the IR map phase (which would mean high IR settings needed for good reflections) or part of the actual rendering phase (ie by direct comp) and therefore dependent more on high rQMC and AA settings?

i have really small geometry that the reflections are flickering a bit…
thanks for any help/advice!

it basically uses the lightcache algorythm for speeding up rendering of glossies, instead of the slower qmc method. That said, irrad map would most likely have no impact. You wont be able to store the lightcache with the irrad map. I did a test last week with alot of glossy reflections and a simple animation with both methods. I couldn’t really tell a difference in quality, though the lightcache method rendered faster.

so storing glossy in LC and not storing it (done by direct comp) there is no diff but it just faster when stored with LC? Does that also mean the quality of the reflections is based then on samples number and size in LC? there are some flickering reflections and i wondered if NOT storing them and letting direct comp do the work would be advised…

thanks

ya i couldn’t say for sure. maybe someone else has more experience with it.

If you use a saved LC the flickering should go away. No way to merge them though so they need to be calc’d in flythrough mode.

AFAIK it’s still using raytracing to calculate the glossy relfections, it’s just using the light cache rays to point the glossies in the right direction - almost like doing a really low quality solution first to guide the glossies to what surfaces they’ll actually hit.

Hey thanks, guys :slight_smile:

Rerender:You mean i cant merge them because the LC is being used for glossies right? yes, i have both IR and LC maps loaded at render time

Joconnell: that makes sense, nice explanation. i just wondered if the settings of the LC affect glossy accuracy the same way shadow accuracy would be improved if the LC samples/setting were raised.

these are the settings I am using

scene is just env skylight and one vray area light

LC in flythrough
1000 samples, 0.04, screen (since its open to sky and big ground plane), store glossies on
store direct light OFF
inter samples 10, nearest
saved out and loaded

then IR multiframe inc
-3/0, high animation settings, 128 Hsph, 20 int samples, every 5th frame.
saved that out and loaded

then for final
adaptive 0/3, normal checked (didnt use QMC since glossies already done by LC and large white surfaces)
rQMC 0.005, 16, 4 subs

its for a this scene

its the reflections and shadows near the seams and area around the power button that are giving me the most trouble (small geometry)…

maybe a better way to ask is. if store glossies in LC is the same qualtiy as not storing them but just faster, in what situation would you ever NOT want to store them?

Funny thing is that having QMC AA vs Adaptive AA makes a BIG difference in render time AND got rid of the flicker..it had nothing to do with my IR or LC settings. I thought since store glossy in LC was on, there would be no benefit, in terms of speed, by using QMC AA but there was. Dont quite get why though. QMC AA is noisier but the flicker is gone!

a reason you might choose to not store glossies is of course, you have very little glossies in the scene. Multiple files needing to be loaded across a network, is one too many if really not needed.

i hear that…good point :slight_smile:

For the record, for this scene QMC AA won hands down in terms of time and flicker reduction (4 min/frame NTSC). I used a cook-variable AA filter to also help with with the thin stobe-y lines I was getting when viewed on NTSC. ok..

I guess there is a difference between CALC glossies and RENDERING glossies (duh)

OK, I’ll shut up now..
here the final settings applied