Ambient Occlusion

Ambient Occlusion - the idea behind AO for me has always been "well, if I can include that in my renders, I will really be the cream of the crop."  However... I never have really looked into it in a hard core kinda way.  Go figure.

So, embolded by my foray into SEM textures, I did a little looking.  And by a little, I mean about 20 minutes worth.  So, really little.  I am working on this crazy iBook thing...  In any case.  AO is nifty.

Obviously, this isn't anything special.  My hand model, with just a few thumb tendons and some ligaments.  But... I love the way AO looks.  My fellow medical artist, Leslie White, has done some really nifty AO renders of the head and neck that I just adore.  They look fantastic.  I'll post some of those soon.

So, I have my settings saved now. An AO texture that I can apply and render out for all of my models, bake it on, etc. I really wanted to be able to show you a non-AO model vs. one with AO... but that's not happening today.  I have had much too much to do to tweak anything anymore.  So, here's a comparison between final gather & global illumination vs. none.

The Goldilocks senario of FG & GI:

No FG & GI

Too much FG & GI

Much closer to 'just right.'

I'm still tweaking the settings, but it's getting to a much better point that it was just a few hours ago.  And if I can bake that type of light map onto the model, I can export it out into Unity and iBooks, etc, and make the models look 4000x better.  In the mean time, I have a ... few renders that look a little less like crap in the iBook now.

Real or "fake"?

I had an interesting interaction with a doctor that I work with quite often.  It really brought up the question, "Is what we do as medical artists real or fake?"

(At this point, I should take a moment to apologize for the over-use of quotation marks in this post.) 

I don't have an good answer for that question, especially in light of the reactions from said doctor.  Before observing her reactions, I would have said with 100% confidence - what we do is "real."  So, yesterday we were working on the project we have been collaborating on, and I was updating many (most) of the images.  One of the things that Stanford (and myself, especially) is very aware of is copyright. 

i.e. You can't just use images from Google in things you are offering outside of the University, and especially not for profit.  

I thought that this was common sense, but it seems that many a professor disagrees.  One of the things I have had to do on this project was replace images pulled from Google with original work. On a personal level, I know that I wouldn't want people ripping off my hard work, so I tend to lean towards the copyright enforcer role sometimes...

In any case, we needed an X-ray that depicted a CMC fusion.  I pulled one that I thought portrayed that surgery from her library... but I was wrong.  It was a hemitrapeziectomy with a k-wire pin, not a CMC fusion with a k-wire pin.  Darn and blast!  It turned out that she didn't have a CMC fusion with a k-wire pin example in her library.

Now, if this book wasn't due the day before yesterday, she could go off to her case library, strip all identifying data from a patient case example, and use an x-ray from there.  Especially since she has a ton of x-rays from patients that have agreed to let their data be used in studies and papers.  But... needs must.

So I modified an x-ray that we had available.  It was ours, nearly showed what we needed it to, and was fairly straightforward to edit.  Which is better than what might have happened- yep, google was the first place to turn to.  Now, she made the argument that an x-ray with identifying data stripped from it is actually in the common domain.  I have no idea, truthfully.  She could have been 100% correct and I spent my time on nothing.  But I would rather err on the side of legality, especially since it's my name on the artwork in this book.

Looks like a CMC fusion x-ray to me!

As I was exporting the book for final edits tonight, her reaction to my editing this x-ray really struck me.  She was unhappy that we didn't have the real thing, or access to one via the internet.  On the other hand, I was happy that we had a drawing that looked accurate and was from source photos that we owned.  I was really wondering why editing an x-ray was bad, but editing photos is okay (I did a fair bit of that as well).  Or even teaching via 3D models and line drawings... why does a "false" x-ray ring so much more on the 'not good' side than the 'teaching' side? 

I don't have an answer for that, but I found the idea of it intriguing. 

Maya SEM texture

So, I've been busy working away at lots of really, really cool things... that I can't talk about.  I'll do another post on Amira soon.

In the mean time, one of the things that I was playing around with on my own time.  Nothing ground breaking, but I like the look of it.  A WIP SEM type texture for Maya.  It's not quite as hard as I would have feared, thank goodness, although not as easy as I remember Max being.

Shading network

This is the shading network that I worked out, with some google-fu and intuition.  I have two SampleInfo nodes connected to the V-coordinates and U-coordinates of a RampTexture that has a black to white ramp.  This is then fed into the transparency setting of a lambert shader.  I have another SampleInfo, connected similarly to another RampTexture.  This one has the color that I want reflected in my model (just a test here) as part of the ramp - as you might expect, this is fed into the color connection of the lambert.

To round it all off, and give a hint of style, I have a brownian bump map on the texture.

I also deleted all the lights in the scene, and amped the Ambient setting to the max, as SEM and Xrays don't have light sources...

This is what it looks like in the scene window... :(

But this is what it looks like once rendered!  Fun.

Started playing with the settings a bit more - adding in some iridescence to give a flatter look, etc.  Having fun!

So, nothing super spectacular, but at least I have the beginnings of how to simulate a neat looking render in Maya.  I'm going to save this Shader, and try out a few more permutations... and maybe see how it looks in Unity as well!