Hi,
I'm always find a way to have a real "invisble" area lights to fake lighting for interiors rendering.
I've tested the opacity option and it should be perfect but I've seen that the mesh, with full opacity, seems to spread less ligh than an area light (or the same mesh with opacity at 0 ).
If you open the scene and measure the RGB values, the lights reflected on the ground are different: mesh with emission and opacity is 1/6 of the area light.
These two images for comparison with a glossy metal material.
Is it impossible to have, at the moment, the same result?
This is the scene if someone would play with it.
Emission and Opacity: lower power?
Forum rules
Please upload a testscene that allows developers to reproduce the problem, and attach some images.
Please upload a testscene that allows developers to reproduce the problem, and attach some images.
Emission and Opacity: lower power?
- Attachments
-
- FakeLights.zip
- (82.96 KiB) Downloaded 131 times
Re: Emission and Opacity: lower power?
This is due to the (lack of) MIS and it is expected but you are not using the tools in the way they are meant to be used. In this case, you should use the camera invisible object option to obtain the correct result:
If you don't want the are light shadow, use a point light source with radius > 0.0 (i.e. a not intersectable light source).
Back transparency is intended for objects enclosing a light source (but not emitting light their-self): viewtopic.php?f=5&t=984
P.S. I could fix the (lack of) MIS but it may be not worth the effort because, like I wrote, it is not the way intended to be used.
Re: Emission and Opacity: lower power?
About the light shadow, yes I don't want it cause as I'm faking lighting using "big" area lights and obviously I'd like to have them completely hidden and a shadow will reveal the area light.
(My "wet" dream is to have the option to filter both specular and reflection visibility...additionally to direct visibility and shadows )
I've tried your suggestion to use a point light but this has many disadvantages like the shadow blurriness depends by the radius instead of the light distances from the objects and to have an area light shape I should, everytime, add a IES file to emit light from just one direction.
In my opinion, add MIS to an "invisbile" light it worths. I use everyday this kind of setup cause many time I don't have full lighting design and designers often can judge more easily a project with a more general lighting instead of full lighting setup. And.. last but not least, it's faster to add 20 area lights to illuminate 90square meters than take a full 3d model lamp, duplicate them at least 80 times and rotate everyone to make a diffused lighting.
In any case, the last word it's yours... I mean that as developer you're surely more entitled to estimate the effort to develop this and if this is something you want to develop. From your previous post it seems that you'd like to avoid but as you left open a door for this I've expressed my whis
(My "wet" dream is to have the option to filter both specular and reflection visibility...additionally to direct visibility and shadows )
I've tried your suggestion to use a point light but this has many disadvantages like the shadow blurriness depends by the radius instead of the light distances from the objects and to have an area light shape I should, everytime, add a IES file to emit light from just one direction.
In my opinion, add MIS to an "invisbile" light it worths. I use everyday this kind of setup cause many time I don't have full lighting design and designers often can judge more easily a project with a more general lighting instead of full lighting setup. And.. last but not least, it's faster to add 20 area lights to illuminate 90square meters than take a full 3d model lamp, duplicate them at least 80 times and rotate everyone to make a diffused lighting.
In any case, the last word it's yours... I mean that as developer you're surely more entitled to estimate the effort to develop this and if this is something you want to develop. From your previous post it seems that you'd like to avoid but as you left open a door for this I've expressed my whis
Re: Emission and Opacity: lower power?
I may add the MIS support for a special case (and probably the only useful): when opacity is set to 0 (i.e. fully transparent like in your case). Adding full support for partially transparent objects is likely to be a pain.
P.S. I'm out of the town, I will check next week.
P.S. I'm out of the town, I will check next week.
Re: Emission and Opacity: lower power?
It should be perfect!Dade wrote: ↑Wed Jul 10, 2019 3:08 pm I may add the MIS support for a special case (and probably the only useful): when opacity is set to 0 (i.e. fully transparent like in your case). Adding full support for partially transparent objects is likely to be a pain.
P.S. I'm out of the town, I will check next week.
Thanks in advance.
Re: Emission and Opacity: lower power?
On daily builds page I can see "Dade916 released this 9 days ago · 3 commits to master since this release".
So, if I'm not wrong, it seems that the last build older than this new feature..or not?
So, if I'm not wrong, it seems that the last build older than this new feature..or not?