Setting up a material test scene for large surfaces

[Art48]

Hi,

For testing material setups we already have a scene with kind of standardized setup regarding lights, for example one Spot light in front which has its intensity set up so that a 128-gray material renders good mid-gray.
Stuff like that.

Now I want to set up a scene for testing large surfaces, floors in particular.

So I'm thinking a room with a few special things in it so it can be used for testing how materials react.

- One Window to let in sun light and sky reflections
- One light at the ceiling for general illumination
- One Self-Illuminated "ball"? Like a light bulb lying on the floor
- Maybe a red box, for testing reflections coming from a normal object instead of a light source


Now I don't quite know how to set up the scene completely physically correct, especially how to translate watts from a real light source to watts in the VRayLight since I have heard it's not the same (regarding energy input and resulting energy output etc). Are there any resources for real-world-references and counterparts in VRay regarding the lights?

[joconnell]

What works really nicely is if you either get an app for your phone or buy a lux meter. Vray has a light meter helper that'll measure light values in your scene, it'll use the same values as a lux meter will for measuring brightness. They're about 15 quid on amazon or ebay and you can measure any light source you want from then on. If you want to go for other values in vray you're right on the watt bit since watts is what the bulb uses but then a lot of them are inefficient in the light they produce from it. Lumens are the units of light that actually get produced from the bulb or the energy that's released so if you try using those values, online charts will likely be more useful for you.

[Art48]

I basically really do not care what unit I use since VRay automatically converts between them.
So you say I can put in the lm value I got on a chart for a specific light (let's ignore any light spread like found in IES files etc) and make a spherelight in vray with the appropriate bulb size and I will get the same amount of light as in reality?

[joconnell]

I did a check a while back with a macbeth chart, the luxi attachment and app for the iphone and an slr. The macbeth chart was only to have a measured colour value where I know the exact rgb values of each swatch so I could use a texture in vray that should match. I put the chart and my iphone on a table in my house with the luxi app measuring the light coming on - it gives you aperture / iso / shutter measurement but also gives you lux values. I took a photo with an slr of the iphone and the chart. With this lot, I had a reading of how much light was falling on the table, a chart I could recreate and also the shutter, iso and aperture values that my camera chose to take a photo of the lot. In 3dsmax, I made a plane for my macbeth chart and put a png of the chart values as it's diffuse. I made a vray camera using the same settings as my real camera shot my reference pic with so that's two things of the three you need set. Lastly I made a vray light at roughly the same distance as the real world bulb was hanging over the chart and iphone. I made a vray light meter helper sitting in the same position as the iphone lux meter was and then changed the values of my vray lights intensity so that the vray light meter helper read the same amount of lumens in that place as the iphone did. Then I did a render and the result was very very close to the original photo in terms of light levels. Note that luxmeters only measure brightness and won't take into account the hue / colour of your light.

[AlexP]

I did a check a while back with a macbeth chart, the luxi attachment and app for the iphone and an slr. The macbeth chart was only to have a measured colour value where I know the exact rgb values of each swatch so I could use a texture in vray that should match. I put the chart and my iphone on a table in my house with the luxi app measuring the light coming on - it gives you aperture / iso / shutter measurement but also gives you lux values. I took a photo with an slr of the iphone and the chart. With this lot, I had a reading of how much light was falling on the table, a chart I could recreate and also the shutter, iso and aperture values that my camera chose to take a photo of the lot. In 3dsmax, I made a plane for my macbeth chart and put a png of the chart values as it's diffuse. I made a vray camera using the same settings as my real camera shot my reference pic with so that's two things of the three you need set. Lastly I made a vray light at roughly the same distance as the real world bulb was hanging over the chart and iphone. I made a vray light meter helper sitting in the same position as the iphone lux meter was and then changed the values of my vray lights intensity so that the vray light meter helper read the same amount of lumens in that place as the iphone did. Then I did a render and the result was very very close to the original photo in terms of light levels. Note that luxmeters only measure brightness and won't take into account the hue / colour of your light.

Interesting.

Is it possible then to measure the hue/colour? I suppose this could be known based on the real world light source in use, if you know the Kelvin value?

[joconnell]

Hmm, you could definitely infer it from the macbeth chart and your camera's white balance by seeing what difference there is between your white light render versus the tinted light in the photo. I think you'd need a more complex scientific instrument to get a proper value, I'm not sure what format that data would come out in either!