harkyman.com

I’m completely out of time, so I’ll call it done. Below you have HD embeds of Snowmen Will Melt Your Heart on both Vimeo and youtube. Pick your favorite.

I also made a 1920 x 1080 render of the coda shot for my desktop backgrounds. Gets me in the holiday spirit.

Cheers!

Vimeo:

youtube:

And finally, the 1920 x 1080 still:

I saw today that The Essential Blender, the Foundation’s official getting started book that I was fortunate enough to spearhead and edit a long time ago, is still selling on Amazon. In fact, it was #38 in the 3D Graphics category. I also noticed several reviews from the past year in which the people who got the book just trashed it and Blender.

Now, I’m a free market person. I believe that consumers should take responsibility for their purchases. How anyone could go on Amazon, with its copious publication information, recommendation system and reviews, and buy The Essential Blender today is completely beyond me. It was published almost five years ago, which is an eternity in the world of software. Recent reviews indicate clearly that the book is out of date. And yet, people continue to buy it.

A simple search on Amazon (or even teh Googles) would have shown them that there are more recent and almost certainly better alternatives (and not just my Blender Foundations). In that sense, maybe the people who are still buying it are a self-selecting low functioning group that are probably doomed to fail at 3D animation anyway. I mean, if you can’t do some basic research on the web before dropping your $25-$40, how well are you really going to be able to handle a full featured 3D animation system?

On the other hand, I think that the Foundation should work with No Starch Press (the publisher) to yank the book. I believe that it is irresponsible to continue to sell a product that is ridiculously out of date. As much love as I have in my heart for that book, it should be in the remainders bin at a discount/left overs type store, selling for $4. Furthermore, by reading the most recent reviews it appears that they are shipping the version of the book that had printing problems. The Foundation never publicly acknowledged it as such (to my knowledge), but one of the printings of the book had serious imaging issues and should never have been shipped. From the reviews, it appears that the publisher is shipping them now. If you ordered one of these books recently and the printing is bad (images unreadable/smudged), I urge you to attempt to return it under the grounds of defective product. As someone who worked in commercial printing for almost a decade, that kind of product should be considered unacceptable. To knowingly sell it as such is… less than upstanding in my opinion.

Of course, this may not even be the Foundation’s call. No Starch Press may have the right to list that book into perpetuity and sell it down to the last shredded cover, fifteen years from now. I certainly hope not, because it pains me to know that when someone buys that book and has a bad experience with it, they now associate myself, the Foundation and Blender with that bad experience.

In any case, maybe this post will show up on someone’s Google search if they are thinking of buying the book and will steer them in the right direction.

In my last post, I put up a link to a rough render of the last shot of Snowmen and said that it needed effects. Particularly, I wanted to include multiple particle systems for laser blasts, dynamic paint on the ground where the lasers hit the snow, and smoke simulations coming from the impact locations. I ran into some major challenges with rendering and compositing.

The first challenge was that halo particles always render behind faces with z transparency. So, for example, when the snowman blasts the camera directly, it simply doesn’t work with the zTransp sky in the background. Breaking the lasers out into a separate render pass and carefully recompositing fixed that.

The next gripe was that I could not get dynamic paint to render. The object it was supposed to paint on had previously been vertex painted, and the two were not playing nicely together. I had to do some serious workarounds to finally make it happen.

Finally, there were issues with the smoke. The renders were looking okay until I added some defocusing into the mix. The cube object that was the domain for the smoke was screwing up the z values within its volume, which obviously blows up defocus results. Even putting it on its own render layer and disabling it in the main one didn’t solve the problem. In the end, I had to selectively bring objects into a new scene and remove the smoke domain from the main scene, then put renders from multiple scenes together in comp.

While this was frustrating (cost me something like two days of work), I’m actually glad that it happened. The details of the process by which I discovered the problems and experimented to find a solution will make a great case study for the rendering and compositing chapter of Blender Productions, as well as providing some good instructional material for the simulations chapter. As always, my goal in things like this won’t be to necessarily teach about how to solve any specific problem, but how to approach complex problems in general.

Did a straight to video render last night of the last shot. Still some tweaking to do, and I have to add all of the effects like snow, lasers, smoke etc. I can see the end of the tunnel though!

Run Away! from Roland Hess on Vimeo.