Color Temperature of Lights implementation of Vray?

[SBIrenderstation2]

Hi all,

So Im writing from an researcher lab using 3ds max + Vray for perception tests and recently we have been focusing on color temperatures of lighting in spaces. And so I discovered two things:

1. If I run two renders where my camera has color correction of 4000 and my lights are at 5000 Kelvin and a second one where my camera color correction is at 3000 and my lights are at 4000, the two produced images are really different in color. Is that the way it should be or am I just missing some physics of light knowledge?

2. There is a mismatch between input and output of color T. In my set up I had camera at 4000 K, lights at 4000 K and screen that normally displays white at around 6500. So in a set up light that I would expect to get a reading of my screen around 6500 and I kinda do. I get a value of 5975 K. However once I start changing the color temperature of my lights, I gain around 1400 K output for each 1000 K input I increase. So my question is: (if not confidential of course) how does Vray exactly work with color temperatures of lights and what process of validation and calibration did it undergo?

Im sorry for the technical questions, hope someone else is interested in this topic as well

Best,

Daniel/SBi

[vlado]

1. If I run two renders where my camera has color correction of 4000 and my lights are at 5000 Kelvin and a second one where my camera color correction is at 3000 and my lights are at 4000, the two produced images are really different in color. Is that the way it should be or am I just missing some physics of light knowledge?

I don't see any reason why they should be the same. The Kelvin temperature is not a linear "distance" between colors and it's wrong to assume that it is.

2. There is a mismatch between input and output of color T. In my set up I had camera at 4000 K, lights at 4000 K and screen that normally displays white at around 6500. So in a set up light that I would expect to get a reading of my screen around 6500 and I kinda do. I get a value of 5975 K. However once I start changing the color temperature of my lights, I gain around 1400 K output for each 1000 K input I increase.

How do get readings from your screen? Keep in mind that the screen is not a black body emitter; what you get on the screen is an approximation of the actual spectrum of a black body emitter.

Best regards,
Vlado

[SBIrenderstation2]

Hi Vlado and thanks for the super fast response,

Im using a recently calibrated Metrue SIM-2 plus to get readings from my screen. ( http://metrue.com/product_info.php?id=198 )

Here are some more readings I got recently, still same set up: Input/output = 1600/2000; 2300/3000; 3000/4000; 3700/4000. So as we discussed here there is some factor scaling happening somewhere along the way from Vray to my screen.

I think Ill try to run some tests on a different screen or may be, as you answered my first question, I should also try different camera white balance temperatures

Best,

Daniel

[vlado]

Is your monitor calibrated too?

Best regards,
Vlado

[SBIrenderstation2]

So, it seems that the monitor has been calibrated only for luminance and uniformity.

Anyways, I figured out it would be a better idea to have my camera correction in Vray at 6500 - the same as my screen's neutral white output. And now when I do my measurements I get much more predictive behavior - for each increase of 1000 in Vray I get increase of 1000 K output. Here is a sloppy plot if you are interested:



And it turns out that in order to calibrate our screens we need to get a 260$ device, so I think Ill just tweak the reds of my screen so I get more precise values. Thanks for helping me work it out Vlado

Best,

Daniel

[vlado]

Heh good to know that the math works out

Best regards,
Vlado