Tech Artists ORG exclusive - a simple method of area lights approximation.:P

Coding
1

Here is me, trying to contribute a bit of basic shading “research” I did to the awesome Tech Artist ORG Community (is not a cry for help this time :P).
It is free for anyone to read/use/comment on :slight_smile:

http://freepdfhosting.com/81f3493456.pdf

I hope you find it useful; I appreciate any form of feedback, be it friendly or caustic :stuck_out_tongue: :slight_smile:
If what I am presenting has been already been deduced by someone else using this method, please me know asap so I don’t make a total fool of myself by thinking that I have something worth showing around :slight_smile:

If the other hand if you do find it interesting/new end up using it, or improving on it, if at all possible please mention me or at least shoot an e-mail my way :P.

Edit: Hmm…found a slip-up … I have overlooked a division by 2 when bringing the result from 2d space to 3d space (did not consistently apply the formula that links a 2d angle to its solid angle counterpart).
Corrected the formulas, updated the LUT’s too, less artist-requested tweaks were needed in this scenario.

Best regards,
Dragos

2

Update!
Made a pass over the whole document to account for the missing division by 2 implementation slip-up ( the theory was sound though :P).
With the new LUT’s specular looks better and does not need as much special care as before :slight_smile:

Updated link the the initial post.:smiley:

3

that is pretty clever! nice work and interesting read.

4

Thank you for reading it and the nice words :). Glad you liked it.

On a side note: updated the original based on a first wave of feedback (Corrected some typos, rephrased some paragraphs, some slide title changes, and also reversed the start and end values on some interval integrals (didn’t really matter because the values were absolute, but it was wrong how they were written before, even though the result was the same :P).