Strange color points

Sinister · Post by **Sinister** » Mon Dec 01, 2008 3:48 am

ok, points gone now, but there is incorrect stroke instead of points..
incandescenceR: 7000 (I know that it is too much

)
any value of ambient B and G canal gives nothing when gain is 0, image is black

I am still confused, how to change brightness of floor?
material is black already.
gain (incandescenceG canal) do almost nothing for me (alterate brightnes, but very little) .
MtI_use Adv Opt is on. Whithout it I get a nice sun-like nuclear reactoin on my glass

CTZn · Post by **CTZn** » Mon Dec 01, 2008 4:57 am

Please share the Maya scene, there is something wrong that I must check about. Or the schott material is hell slow to converge... replace it with a basic specular glass for testing purposes, then if the problem persists please share the scene. I will parse the options and tell what can be ehanced, if not related to the material.

any value of ambient B and G canal gives nothing when gain is 0, image is black

Thanks for reporting, use gain only for now.

Sinister · Post by **Sinister** » Mon Dec 01, 2008 9:24 am

http://downloadatt.nextmail.ru/w/x/35b1 ... f/Glass.mb it's scene
I'll test basic glass little bit later.

Post by **fused** » Mon Dec 01, 2008 9:36 am

Sinister wrote:ok, points gone now, but there is incorrect stroke instead of points..
incandescenceR: 7000 (I know that it is too much )
any value of ambient B and G canal gives nothing when gain is 0, image is black

I am still confused, how to change brightness of floor?
material is black already.
gain (incandescenceG canal) do almost nothing for me (alterate brightnes, but very little) .
MtI_use Adv Opt is on. Whithout it I get a nice sun-like nuclear reactoin on my glass

wtf? o__O

nice effect, tho

CTZn · Post by **CTZn** » Tue Dec 02, 2008 5:00 am

I downloaded the scene sinister, first thing is that I have a MtI version clash and second I must expand the schott materials library, I will tell more later.

Actually that's a shame but you learned me ho to use the phongE material, I wonder how you found out since bkircher never commented her creation, despite my repeated calls...

rgigante · Post by **rgigante** » Tue Dec 02, 2008 5:36 am

I really don't understand your point guys: those fu..in points are only "aperture diffracted" fireflies....i don't really understand why you're trying to hack indigo lights in order to get rid of them.

They are only generated by a need-to-be-fixed bug.

Ric.

CTZn · Post by **CTZn** » Tue Dec 02, 2008 5:49 am

Because I believe that by setting the right thresholds in sampling one can prevent those "bugs" to pop in. I'm refering to the first image of this page, the single colored splats. I believe too that with crazy settings they can converge to something correct, at the least I guess so. Or they just must be clamped with the right settings.

I would not call this a bug, but after all I could be wrong, I just thought that was worth a try.

That was my point rgigante

CTZn · Post by **CTZn** » Wed Dec 03, 2008 12:47 pm

First thing is that I fixed efficacy scale in MtI (available soon), second will have to wait for, cheers.

CTZn · Post by **CTZn** » Wed Dec 03, 2008 5:45 pm

Sinister, your glass is eleven meters tall, since you work in cm, mti Global Scale Factor should be 0.01. Also the strength of your light is atomic

PureSpider · Post by **PureSpider** » Thu Dec 04, 2008 1:08 am

Tri-tri-triple post! *UT voice*

Sinister · Post by **Sinister** » Thu Dec 04, 2008 5:50 am

Ooops

I'll remember

Also the strength of your light is atomic

What you mean? That 7000K is too much for light?

PureSpider · Post by **PureSpider** » Thu Dec 04, 2008 6:04 am

WAY too much

CTZn · Post by **CTZn** » Thu Dec 04, 2008 7:08 am

PureSpider wrote:Tri-tri-triple post! *UT voice*

Thanks for that

Sinister, no pun intended:

Code: Select all

	<spectrum>
		<blackbody>
			<temperature>7000</temperature>
			<gain>100</gain>
			<efficacy_scale>
				<power_drawn>4000</power_drawn>
				<overall_luminous_efficacy>4000</overall_luminous_efficacy>
			</efficacy_scale>
		</blackbody>
	</spectrum>

Did I say... atomic ?

Correct the global scale factor (avoid scaling the camera in all cases) and try this:

Code: Select all

	<spectrum>
		<blackbody>
			<temperature>7000</temperature>
			<gain>1</gain>
			<efficacy_scale>
				<power_drawn>400</power_drawn>
				<overall_luminous_efficacy>80</overall_luminous_efficacy>
			</efficacy_scale>
		</blackbody>
	</spectrum>

Really these values do not correspond to a real luminaire I know but they are by far most reasonable than the ones you used.

7000Â°K are allright, but if you refer to the Indigo documentation <power_drawn> is in watts and <overall_luminous_efficacy> basically multiplies it. The point was to put 4000 into both. Also, everytime you want to calibrate lights do not use Reinhard tonemapping. It's as fierce as it is versatile.

Sinister · Post by **Sinister** » Tue Dec 09, 2008 8:36 pm

I wrote it directly into exported xml scene, into mesh light section. But I see no difference between
power_drawn = 400
overall_luminous_efficacy = 80
and for exaple
power_drawn = 40
overall_luminous_efficacy = 8

even if one of this paraneters = 0 scene do not change..but temperature change color..

I'm using just Maya and Indigo, is there some option about Reinhard tonemapping? may be it turned on?

PureSpider · Post by **PureSpider** » Wed Dec 10, 2008 12:46 am

If there is no change in the brightness of the light (even if you actually do change it) then I assume you are using Reinhard already... change to linear or cam

Strange color points

Who is online