meshlight and scene scale

General questions about Indigo, the scene format, rendering etc...
Post Reply
8 posts • Page 1 of 1
User avatar
suvakas
3rd Place Winner
Posts: 2613
Joined: Mon Sep 04, 2006 11:08 pm
Location: Estonia
Contact:

meshlight and scene scale

Post by suvakas » Wed Feb 14, 2007 11:21 pm

Can someone explain to me how is the scene scale related to light intensity and vice versa?
I did a small test with cornell box. It seems, that the light intensity goes weaker while the scene scale increases. I think this is a bit strange, cause i think light power should be relative to its size. I assume, that 50cmx50cm light source should emit 5 times more light than 10x10cm light. Or isn't it so?

Here are 2 images. CornellBox side on the left one is 0.5m x 0.5m and on the right it's 2m x 2m. Light settings, f-stop etc etc remain the same on both of the images. Only the scale changes.

Image

Why this is happening? What am i missing?

Emitter and camera settings for both images:

Code: Select all

	<camera>
		<pos>-0.285 -0.648048 0.275</pos>
		<up>0.0 0.0 1.0</up>
		<forwards>0.0 1.0 0.0</forwards>

		<focus_distance>0.921015</focus_distance>
		<aspect_ratio>1.0</aspect_ratio>
		<aperture_radius>0.010864</aperture_radius>
		<sensor_width>0.036</sensor_width>
		<lens_sensor_dist>0.0434558</lens_sensor_dist>
		<white_balance>D65</white_balance>
		<autofocus/>
	</camera>

	<meshlight>
		<pos>0.0 0.0 -0.00114896</pos>
		<scale>0.001</scale>
		<rotation>
			<matrix>
			1.0 0.0 0.0 0.0 1.0 0.0 0.0 0.0 1.0
			</matrix>
		</rotation>
		<mesh_name>light_box</mesh_name>

		<spectrum>
		<blackbody>
			<temperature>2800</temperature>
			<gain>0.001</gain>
		</blackbody>
		</spectrum>

		<efficacy_scale>
			<power_drawn>100</power_drawn>
			<overall_luminous_efficacy>17.5</overall_luminous_efficacy>
		</efficacy_scale>
	</meshlight>
Suvakas

User avatar
OnoSendai
Developer
Posts: 6243
Joined: Sat May 20, 2006 6:16 pm
Location: Wellington, NZ
Contact:

Post by OnoSendai » Wed Feb 14, 2007 11:23 pm

The power drawn is over the whole emitter surface, not per unit area.
So bigger area -> smaller emitted radiance

User avatar
suvakas
3rd Place Winner
Posts: 2613
Joined: Mon Sep 04, 2006 11:08 pm
Location: Estonia
Contact:

Post by suvakas » Thu Feb 15, 2007 12:00 am

hmm..still abit confused.
100w efficacy_scale is not the same thing as 100w light source in real life?
What would be the correct way to light that 2mx2m box? Just by increasing efficacy scale or tone maping until satisfied? So far I've thought, that efficacy scale is about the same thing as W in real life. (an 100W light bulb would sure illuminate that 2x2 meter room nicely)

suvakas

User avatar
OnoSendai
Developer
Posts: 6243
Joined: Sat May 20, 2006 6:16 pm
Location: Wellington, NZ
Contact:

Post by OnoSendai » Thu Feb 15, 2007 12:09 am

Well..
At this point it depends on the tonemapping, and in general how sensor irradiances (or energies) are mapped to output pixel values. Reinhard always maps to normalised pixel value ranges (eg. the light src will be RGB=(1,1,1)).

Linear is different, but the mapping is still not calibrated against real camera sensors. Also exposure time is not simulated now.

This would be something cool to do in the future for Indigo :)
However i'm not sure where to get this calibration data from.

User avatar
suvakas
3rd Place Winner
Posts: 2613
Joined: Mon Sep 04, 2006 11:08 pm
Location: Estonia
Contact:

Post by suvakas » Thu Feb 15, 2007 12:23 am

Gotcha! Ok. Thanks!
I spent a lot of time today figuring out why the scene with a larger scale comes out so dark.
Good to know, that it's supposed to be like that. 8)

Suvakas

User avatar
suvakas
3rd Place Winner
Posts: 2613
Joined: Mon Sep 04, 2006 11:08 pm
Location: Estonia
Contact:

Post by suvakas » Thu Feb 15, 2007 2:59 am

i didn't want to do a new thread about this so i'm posting it here.

There are 3 spheres:
a) diffuse mat + bump
b) specular transparent + bump
c) glossy transparent + bump

Now while diffuse and transparent materials are ok, the bump on glossy sphere seems to be somewhat wrong to me. Like the pole of the UV has been rotated 90 degrees or something.

Image

...and the mat for glossy (just in case)

Code: Select all

	<medium>
		<name>glass</name>

		<precedence>10</precedence>
			<basic>
			<ior>1.33</ior>
			<cauchy_b_coeff>0.0</cauchy_b_coeff>

			<absorption_coefficient_spectrum>
            		<rgb>
              			 <rgb>0.001 0.001 0.001</rgb>
           		 </rgb> 
			</absorption_coefficient_spectrum>
			
			</basic>
	</medium>





		<material>
			<name>02 - Default</name>
		
			<glossy_transparent>
				<internal_medium_name>glass</internal_medium_name>
				<exponent>10000</exponent>


				<bump_map>
				<uv_set>02 - Default-bump_bitmap</uv_set>
				<path>C:\3D\textures\Bump\Weave2_t.jpg</path>
				<b>0.5</b>
				<exponent>2.2</exponent>
				</bump_map>
			</glossy_transparent>

		</material>
Well..i'm probably missing something again :lol: ..but would be good to get some comments on this.

Thanks
Suvakas

User avatar
OnoSendai
Developer
Posts: 6243
Joined: Sat May 20, 2006 6:16 pm
Location: Wellington, NZ
Contact:

Post by OnoSendai » Thu Feb 15, 2007 11:38 am

Oops.. forgot to bind the UV coords on the glossy transparent map.. will fix next release. thx!

User avatar
suvakas
3rd Place Winner
Posts: 2613
Joined: Mon Sep 04, 2006 11:08 pm
Location: Estonia
Contact:

Post by suvakas » Thu Feb 15, 2007 9:31 pm

Great! Thanks!

Suvakas

Post Reply
8 posts • Page 1 of 1

Who is online

Users browsing this forum: No registered users and 23 guests