PDA

View Full Version : Basic HDRI for IBL question

Kinematics
10-09-2009, 09:04 AM
Hello guys,

I've been busy discussing and doing research on using mirrorball images to light environments. Further information here Mirrorball Discussion (http://forums.cgsociety.org/showthread.php?f=2&t=813475)

But while working it has come to my attention that my knowledge of HDRI is rather limited. What is the purpose of using HDRI images to light environments? Technically, we are only using one exposure isn't it? So is Tone mapping the images the key to HDRI for image based lighting?

And Maya(mental ray) specific - How do you turn the exposure for the IBL up and down in maya? I tried adjusting the exposure but it makes no difference in the result as far as I can tell. I am working in a linear workflow though but I never ran the HDR image I made through a gamma correction node (the 0.454) thing.

Please advise. Thank you all =) :applause:
playmesumch00ns
10-09-2009, 12:23 PM
Yep I think you're misunderstanding what an HDRI is. It's not some magical constructrion that encodes multiple images in one file. All it is an image format that's capable of representing values greater than one. This is necessary for lighting since light sources have values much higher than one. The values can be in the thousands for the sun, for example.

The reason we take multiple exposures when building the HDRI is that a normal camera CCD can only capture values in a very limited range (0-1, basically). So if you take a photo of a lightsource, it's values will be clipped. But if we take multiple photos with different exposures then with a bit of maths we can figure out the light source's original values.

I don't know about Maya's IBL controls but when you create the HDRI it *should* already be linear, so you don't need to gamma-corrrect it.
phix314
10-09-2009, 04:49 PM
Yep. Gamma is a pretty deep subject, but basically your "run-of-the-mill" 8 bit image is 2.2ish, but in a 32 bit float image, there is no gamma (i.e. "linear"). Once you reduce it to 8 or 16 bits is when the gamma curve is applied.

This book:

http://www.hdrlabs.com/book/

Is fan-fu**ing-tastic. Goes over absolutely everything one would need to know about using HDR imagery and why. Some of it quite deep.

As for the exposure, it would be the color gain, pushing the value slider past 1. Maya can't show a correct thumbnail, so it may look white at some point, but if the dynamic range of the image is high enough, it'll be fine in the render.
playmesumch00ns
10-09-2009, 05:25 PM
Yep. Gamma is a pretty deep subject, but basically your "run-of-the-mill" 8 bit image is 2.2ish, but in a 32 bit float image, there is no gamma (i.e. "linear"). Once you reduce it to 8 or 16 bits is when the gamma curve is applied.

The bit-depth of an image file has absolutely nothing to do with the colour space it is in. By convention floating-point EXR/TIFF are normally linear, 10-bit DPX/CIN are normally log and 8- or 16-bit images are normally sRGB, but there is no reason they have to be and no guarantee that they are. Simply assuming that a particular file format is indicative of a particular colour space is likely to lead you into trouble. For example I have seen HDRI floating about on the internet that are clearly sRGB instead of linear (god knows why).
CGTalk Moderation
10-09-2009, 05:25 PM
This thread has been automatically closed as it remained inactive for 12 months. If you wish to continue the discussion, please create a new thread in the appropriate forum.