floats versus doubles

When to use floats and when to use doubles is one of those classic topics of discussion and much less pointless than the code formatting religious fights.   A fun thing in the new edition of FCG is that Steve Marschner and I are in opposite camps.   I just checked what amazon previews to the reader with "look inside" and found this (note-- I think that discussion of debugging is something I wish I could send back in time to myself-- it's VERY basic but graphics debugging is different in some ways):

Screen capture from Amazon preview of Fundamentals of Computer Graphics 4th edition

See-- sometimes opposing camps can get along :)

Winner of the annual University of Utah teapot rendering contest

Once again Cem Yuksel hosted his annual University of Utah teapot rendering contest.    Last year's winner Laura Lediaev won again.   Here is one of her pictures (my favorite):

Laura's other images and some technical details (including Christmas trees(!) are here).

New edition of my graphics text is out.

The newest edition of my intro graphics book is done early.   It looks like the kindle edition is out and the hardcopy will at at amazon soon. I just got my copies in the mail today!   I am a big fan of this edition.   It is also now Marschner and Shirley et al., instead of Shirley and Marchner et al.   Any great improvements in the book I am pretending are correlation and not causation.   But being second author is like I hear being a grandparent is-- all the benefit and none of the responsibility.   Please report any problems to Steve.

This edition we have moved to color:

That retro image in the bottom right is there for nostalia's sake.  It was my first good image (1988 I think).   It was 1024x768 and took two weeks to do radiosity and light maps and 16 viewing rays per pixel on a VAX 11-780.   It was my first "real" C++ program (I think).     Modeled in vi (for real) and some hand rolled rotate a piece-wise linear curve code.   Here it is in all its low-res noisy glory.

Tone mapping colorful images

I recently noticed that a paper of mine from a few years ago had zero references.   This wasn't a paper with any deep research but instead was a "what is the simplest algorithm we can do that works" paper.   The idea was what to do with the practical problem of how most tone mapping pipelines kill bright blues.    Here is a figure from the paper on our inspiration for the user-controlled "dial" we added:

I think I figured out why nobody uses this paper-- we never put a copy online!   I just fixed that.  The paper is:

A copy is available near the bottom of my papers page.

iPad Pro with pencil stylus review

Summary:  Overall, the whiteboarding alone is reason enough for me to buy an iPad pro-- it has gone over that magic threshold of "good enough" that pads of paper will be a thing of the past for me.   Your excuse may be different (pencil drawing is a good one), so I think this device will be huge.


Dave Hart and Grue Debry got an iPad Pro with stylus for their company and they loaned it to me to try out tonight.   I used their limnu shared whiteboarding program to test it.

I tried it with the math I was messing with today (trying to meet Andrew Glassner's color space challenge to get uniformity into the prismatic color volume).

First, I **love** that I can rest my hand on the iPad while I draw (the iPad pro understands not to count that as a touch).

My hand is resting on the iPad as I draw and this is more important to my comfort than I would have thought

My biggest reaction was that this iPad is exactly the size I want.  It's about the size of 8.5 by 11 paper (actual working area size about 7 3/4" by 10 1/4"), so maybe that size evolved in paper to be the "right size" or maybe I am just so used to it that I like it.   Any bigger and it would be awkward to transport, and using this as a pad in a coffee shop is a great use case.   And of course you can pan so really it's a window into a much bigger sheet of paper.

The stylus is fantastic.   It feels good and has some features that has me not as eye-rolling about calling it a pencil.

As a white board marker I loved it.   Changing colors and nib sizes was more useful than I anticipated.   Using it as an eraser (which I had to do a lot as will be evident in some of the not very careful eases below-- I really do use limnu like a white board-- it's for blazing through ideas).   Here's my first screenful.  

A screenshot of my first session on limnu with the iPad Pro

 15 more 2x2 equations to solve (doing them as special cases to take advantage of zero dropouts) so I will definitely need the pan feature.   I used to use a big real white board or a giant artist pad for these situations, but I will most definitely use a tablet from now on.   Even without the saving and collaboration features I think it would be a win just because of physical portability and fluidity.  

Overall, the whiteboarding alone is reason enough for me to buy an iPad pro-- it has gone over that magic threshold of "good enough" that pads of paper will be a thing of the past for me.    I don't think it will make my laptop obsolete due to OS issues (Microsoft is making a better play for that now).   But the hardware of the iPad pro is in the laptop power zone.    John Gruber has a really interesting discussion of this hardware/software issue.

Cool 2D rendering project

Benedikt Bitterli has posted a really neat 2D rendering site that includes a javascript demo that he also has put on github.   (Via Dave Hart)

Uniform color models

I had an interesting discussion with Andrew Glassner (since this is often a ray tracing blog I'll tell the youngsters that Andrew is known for many things but he's also the inventor of the first sub-linear ray intersection algorithm!)  about the prismatic color model I have been touting.    Andrew points out that while it retains the good properties of HSV it also retains its worst property: terrible color uniformity.    The color scientists have long rightly held that it would be nice if a color space had intuitively similar changes in color for similar changes in distance in the space.   The so-called MacAdam ellipses on the CIE diagram (where each ellipse holds a collection of colors that are nearby) can be used to warp the CIE space to a more uniform one.

Each MacAdam Ellipse has a set of colors you can barely tell are different.   A "uniform" color space would have disks of the same size for these.   Source wiikipedia.

Andrew rightly points out that for a color space to be kick-ass (my term) it should be at least somewhat uniform.   So a challenge to all of you out there: create a uniform space that is RGB-centric.   Or if you know of one, tell me!