Ok
will try all of that soon and keep you posted with renders and render times.
Thanks.
Ok cool.
Standing by 8)
Daforce
one last quastion ![]()
i’m gonna try later on with the store with irr map unticked and u said i should lower my vray light subdv .
8 is good for that case?
If its unticked i would still use between 10-20
If its ticked i would use 20-50 depending on scene.
3 Subdivisions is way too low on a material IMO. That setting speeds things up but causes unwanted noise (unless you really want it or have an object at a considerable distance.) 8 is the default for Vray - try raising it to a little above 10 and see what happens.
‘Check sample visibility’ option is good for the points like your beam touches to wall.. I think, turning it on would solve the lighting problem at that place.
Thanks.
what is this 1/2 size IRR map business?
Isn’t that making a lot of grain at full-res?
Can you show one with full-size IRR?
And for goddsakes take the cds out of the lighting tests.
what is this 1/2 size IRR map business
it means that you can calculate the irr map at 1/2 the size of the final render , as far as i know this is a recommended methhood by chaos.
here are what i came up with after all your comments ,keep in mind that this is still wip.
all of them share DaForce lc settings.
pic 1: vray light - store with irr map = ticked, no pre calculated ir, render time = 4.52 hrs.
pic 2 :vray light - store with irr map = ticked, pre calculated ir at 1/2 size, render time = 31 min for ir calc , 32 min for render.
pic 3 : vray light - store with irr map = ticked, pre calculated ir at 1/2 size, render time = 31 min for ir calc , 32 min for render.
pic 4 : vray light - store with irr map = unticked, pre calculated ir at 1/2 size, render time = 28 min for ir calc , 5.00 hrs for render.
pic 5 : vray light - store with irr map = ticked, pre calculated ir at 1/2 size, render time = 34 min for ir calc , 1.10 min for render.
on this one i raised the light subdv to 40 and glossy to 10.
tell me what you think and which way is the best you think.
Gili
The last one is by far the best.
As you can see by raising up the wall glossy subdivs it makes it look alot better. I would suggest (as a test) try turning of the glossy on the wall and lower the vray light subdivs to 25 and then render again. Ideally it should turn out a nice smooth wall. Im just curious if it was the vraylight subdivs or the glossy subdivs that fixed it. (and im pretty certain it was the glossy subdivs, as 3 is way to low)
oo and that girl at the front is way to bright
(and im not talking about her intelligence)
I think that as far as shadows the one with store with irr map = unticked is the best.( look at the door handle)
as far as 1/2 size irr map , i don’t see it making any changes..and its faster.
Yeah i was looking mainly at the wall.
Yeah having the store with IR map UNTICKED will result in much better shadow details. But longer render times… so yeah. I reckon leave it unticked ![]()
Where is this recommendation?
I saw that you CAN do it - and not necc 1/2 - but never recommended.
All I know is when I do that - the images are much more grainy. Just try it.
Those people are horrible… can you adjust the levels on them so they aren’t blown out?
meanadam
You are right it is possible and if you want to save some render times then you will use it.
the people will be fixed later on , right now i have some changes being made in the design ,its still a wip.
anyway thank u for your reply.
Lowering the min. max irrad map samples is the same as lowering the output resolution when saving the irrad map calculation. I believe that it’s actually recommended that you lower the min/max samples.
--Jon
What about the settings on the QMC sampler… Do you guys ever mess with that? The min sample and the global subdiv multiplier and the noise thresholds? Just curious cause I usually mess with those quite a bit. It seems to work differently than just simply turning up the subdivs on the light cache and the Hsph subdivs.
I find that I can change the multipliers in the QMC sampler and keep my Hsph subdivs down under 20 on a single pass like -3, -3 and keep my light Cache samples around 500 and still get some decent speed and quality for my test renders… probably wouldn’t keep it that low for my final render though. If you bring the noise thresh down to .002 also seems to get rid of a lot of the grainyness.
But hey… i could be doing it all wrong… in which case don’t mind me. ![]()