fireflies generated from spesific kind of ies file.

Corona Corona for 3ds Max Issues
1

hi, I have an asymmetric ies file to work. but it generates fireflies when bounced to glossy object I think..
and I make a comparison to max photometric light with same ies file. and the result shows that with photometric light I got no fireflies at all.
I tried to changed the LSM to 8-16, and MSI to 10, fireflies is still there. and no fireflies when use another ies file, and the same ies file rendered with vray.

check attachments for ies file.
eWFusePowercore_2700K_30x60deg_071210_ITL65261.IES (2.36 KB)

2

I believe the difference is that the photometric light has no surface really, so it produces less reflective caustics. The suggested solution would be to use rayswitch material or map to decrease/disable reflective caustics for the metal material. You can do this by darkening or setting to black the GI rays of the metal material. (either the “wall” or the inner part of the light case as it looks like chrome)

3

I wonder what’s about that very dark region on the wall under light’s hotspot with Corona lights? Photometrics doesn’t show that.

4

About reflective caustics and corona lights.

There is one more way to get rid of the fireflys. The less sharp reflections on the reflector of the lamp (the part where the light bounces first time) something like .35-.25 will produce more soft caustics and therfor there will be no fireflys.

And of cours you can use rayswich if you need them to remain visibily with sharp reflections.

5

I suppose it is shadow of the light object. Disabling “occlude other lights” could fix this.

6

yes i’ll use rayswitch material, never think about that solution.. thanx :smiley: but I hope that’s just a temporary solution not the only solution. because I don’t experience this with other renderer.
check the attachments, its the scene file.
btw, why 1000 lm in photometric light isn’t show the same brightness like 1000 lm in coronalight?
and I saw that in photometric light 1lm is equal with 1.025cd, and in coronalight 1lm is equal with 0.318cd. will appreciate if there is non-science explanation about this.
noise from ies .7z (192 KB)

7

It is a temporary solution until a better way of rendering caustics is introduced. This happens because Corona tries to be as physically correct as possible only with some little bias and in other renderers the caustics would get clamped/blurred.

btw, you didn’t attach the textures and most importantly IES in the scene :wink:

8

I didn’t use any texture in the scene, i’m sure the texture is called from unused material slot and the IES is attached in the first post :smiley:

update: sorry for always late to reply, cuz I’m in Indonesia (UTC+07) which has a bit far different timezone from Czech (UTC+01)

9

fixed

10

huraayy! :smiley: