So I spent a lot of time programming interrupt-driven PWM (Pulse Width Modulation Code) This also allowed me to control the effective brightness of the lines but flickering the various EL Strands on and off very quickly.

As usual, I took no photos of the finished product when I was at Burning Man and… It takes a bit of work to assemble. So I’ll show the prototype here in a couple of pieces.

First, the bike canopy itself The white center stripe is stitched with open overlapping ‘scales’ so that wind coming from the front will pass through them with less resistance and wind coming from behind closes them and works like a sail.

A demo of the PWM software is here:

In this demo, the code is just passing a greyscale fractal image through the 8 channels. Because I spent so much time working around the hardware limitations, I didn’t get to do the animation I was planning. Now that I understand the hardware better, I’m going to try designing a new board that can control more strands and will either not require or will offload the PWM handling so it will be a little while before I come back to the animation.

The actual animation for the EL Array will ultimately be controlled by manipulating symbols in a Flash timeline. These then export the data to a compressed file format which will be stored on an SD card and loaded into the bicycle’s onboard computer. Lots of neat stuff left to do on this project but I’m going to detour from it for a short time to work on another project in order to gain the experience required to finish the hardware redesign for this one.

After decades of a split personality on the internet, I’m trying to unify my identities under a common name.

However, it also left me with questions. Skeletons gave me an idea of some of the movement but I kept wondering about the range of motion for certain limb designs so I came away with as many questions as answers.

When I returned home, I began looking for some good books to really help me nail down convincing anatomical movement. I’ve got a few books.

Gray’s Anatomy

, the ubiquitous Burne Hoggarth book,

Dynamic Anatomy

, and a few others. The only one that I might recommend is the

Weatherly Guide to Drawing Animals

which is pretty decent, but I found I was still wanting something more in depth.

A little searching at the local bookstore lead me to a pair of books:

Anatomy Drawing School

by András Szunyoghy and György Fehér. One on humans and a separate volume on animals. These books I feel really go into a lot of depth on all the details I was searching for. The renderings of the structures are quite good and go into a good deal of depth about the behaviors of the different types of bones, joints, ligaments, and muscles. It also includes descriptions of the ranges of motion. The text information is quite terse but somehow manages to convey a lot more detail than I felt I gleaned from Gray’s (at least for my field of interest) And most text descriptions are accompanied by graphical depictions of the motions and ranges in question.

I have a few very minor nits with the books. In particular, some of the drawings when showing the various views of a bone will also switch from right to left, which can be a little confusing. Also, I with in the animals book they had included a plantigrade quadruped (such as a bear) but on the whole, these are far and away the best art reference anatomy books I’ve come across and was so moved as to write a review.

During the move, my scanner (which had been making funny noises for a while) finally conked out, and my printer was both near-dead and no longer supported by the manufacturer so I took the opportunity to replace both with an Epson Artisan 810 combo unit. Since it comes with a sheet feeder, I thought it might be a good opportunity to check out the state of inexpensive solutions to optical registration of bulk scans.

All of these work by requiring you to black out part of the scan bed where the peg bar sits so that you get solid black peg holes to make it easy for the software to identify the location of the punches. Finding information about automatic optical registration proved difficult. There are of course many high-end solutions in the several hundred dollar range but at the moment, these solutions are out of my range.

Two solutions I did find within my price range are DigiCel’s Flipbook Lite (just shy of $80) and a free beta of a java application called ScanFix written by Duane M. Palyka.

Both applications performed reasonably well. Flipbook’s import is easier to use and it also has some built-in smarts about adjusting the contrast of the drawings to make them show up well. The only minor confusion I had with it was in configuration of the registration offset but this was quickly fixed. It also supports TWAIN libraries and was able to import directly from my scanner with no interim files. I did not invest the time to tinker around with the rest of flipbook’s features for actually creating animation since I’m already familiar with and invested in another tool however, it seemed fairly straightforward and intuitive and may well be worth looking at.

Scanfix is a little less user friendly. I had to do a bit of tweaking to my scan configuration to get files where it could reliably detect the holes. It is also in beta test and I encountered a few problems in using the optional rotations but, once I resolved these issues, it performed adequately. In addition, Mr. Palyka has been quite pleasant to converse with.

On the whole, scanning at 150DPI, the scanner imported equivalent to the fastest I could possibly photograph work on my animation stand (meaning that on average, the scanner was far faster). I even got so lazy as to write a short script to reverse the order of scanned pages so I didn’t have to resort my cels after taking them off the animation desk. Finally, the scanner does not require me to operate the camera. Throw in a stack of cels, hit a button, and off it goes. So in good situations, it’s a bit of a time saver. However, I did run into issues a couple of times where the scanner grabbed more than one page and caused me some problems with a sequence and bogged down my progress. Since this is an invisible fail until you run the pencil test, it’s a bit of a sticking point.

Still. If you’re a student or otherwise on a tight budget, both tools are worth checking out, depending on your personal needs. Though I must confess that I enjoy drawing 12 field and since my scanner does not support paper of that size combined with the scanner feed issues means that I shall not be tossing out my animation stand just yet.

Here’s a bad webcam photo of some beadwork I did recently. (Just the necklace for the ocarina, I’m afraid, not the other necklace) Though speaking of photos, I met Michael Olsen of ZorkMagazine last night at an event for the Timothy Leary archive and he snapped a couple photos of me. His photos of the event can be found here. Met a lot of really amazing and fascinating people there. For the past couple of weeks, I’ve been downloading tons of video off archive.org to put together a retrospective of the late Dr. Leary. I’ll post some updates on this blog when there’s more to show.

Clay’s most recent guest, Eric Goldberg has a new book out titled, Character Animation Crash Course

Since I’d just come off a long animation hiatus and my character animation has always been stiff, I bought a copy and the hardest thing has not been just ripping through it cover to cover.

This is the absolute best book I’ve ever read on character animation.  His style of writing is easy to read, his examples clear, and it just makes everything click for me in a way that it really hadn’t before.

Curiously, before getting his book, I’d decided to take a stripped down character and re-animate him doing the same ‘take’ over and over in different ways to convey character, emotion, etc.  In his book, Goldberg had taken this exercise one step further, starting and ending with the same pose and doing different in-betweens.  Very good stuff.

I’m also really appreciating his information on timing.  I’ve long understood timing in a strictly mathematical and mechanical sense.  I can graph parabolas like nobody’s business, and I understand how they relate to good motion, but it hadn’t clicked.  I had all the right steps in there but it was still stiff and mechanical.

Somewhere between reading Goldberg’s book and freeze-framing through Clampett animation, it all just suddenly clicked and, like Victor Frankenstein, I now feel as if my corpses have been brought to life!

Two thumbs up!  (But four frames apart to avoid twinning!)

After a long interlude, I’m finally returning to doing some animation. In the interim, I’ve largely been working on still art, writing, and sewing.

One of the many projects I created during this hiatus was the creation of a fantasy blog. Rather than post art and stories with the continual disclaimer of ‘This is just make believe.’, I created a blog where faeries are real and the writer lives in a strange world. The writing still takes the form of short stories but for some reason, the framework of living in a world of fantasy helps me bypass a form of writer’s block that would otherwise impair the work.

So. Hang on! Animation is coming soon and in the meanwhile, how about a story? Please check out my fantasy worlds with The deviant spirit, wanderer of the twilight worlds at http://dvnt-spirit.livejournal.com

Happily, Premiere has returned for the Mac. I purchased the CS3 production suite and paid a visit to my old friend.

The interface of Premiere CS3 was instantly familiar. On some level, I was rather surprised by this. I’d been nursing along an aging copy of Premiere 6 and I expected it to look more changed. The big improvements in UI are the dockable windows and speed. The way I’d been running 6 was heinously slow. Stability also seems to be much improved. I didn’t experience a single crash during the entire project.

It took a few minutes to find the audio and video effects for clips as they are not available from the menu bar or right-click in the timeline as with 6 but once they were located, it was easy to adapt (although right-click from the timeline was very convenient and will be somewhat missed) I was also unable to find an A/B track editing mode for transitions however it wasn’t needed for this project so I didn’t search very hard and it may yet live. Addition of entire-track sound level setting versus for an individual clip was also extremely useful.

On the other hand, there are several new bugs which are rather frustrating. First and foremost: Cut and paste of clips. Paste would paste at the current time marker but which track it would paste in was seemingly random and often impacted other clips. This was extremely frustrating. The only way I really found to deal with it was by locking all other tracks. This was time-consuming and often not a fantastic solution.

Audio clips also had a particularly perplexing habit of refusing to live in some audio tracks, but I could never determine which one a clip would refuse to live in until I dragged the clip to it and it simply refused to go there. What gives?

For mono audio clips, I also could not seem to get the track to play on both channels or crossfade from one channel to another. I suspect this functionality has been shuffled off to SoundBooth but honestly, it’s such a basic and integral feature, it should have been preserved in Premiere.

Parceling features out to other applications seems to be the name of the game for the CS3 version of Premiere. Many things that had worked in Premiere since time immemorial were missing or did not function correctly. Track mattes are another example. While Track mattes still exist in Premiere, they do not do well with scaled or moved clips. Scaling the clip to be track-matted resulted in unpredictable matting areas and if it were scaled down too much, my track matte would often end up being just a rectangle in the center of the screen that did not scale. It was very finicky. Of course, the solution here is simply to use After Effects and render that clip then import it into Premiere but as I was limited both on time and disk space, I tried to do it directly in Premiere and met with a lot of frustration. Junk mattes were similarly broken when dealing with HD 120i clips scaled down to DV pan-and-scan resolutions.

On the other hand, one feature that’s drifted from AE to Premiere is the ability to set multiple key-out colors on a track. This was fantastic and gave me a lot of flexibility in removing a very wrinkled and unevenly lit greenscreen from the scene.

All in all, I rate my recent experience with Premiere CS3 as rather positive. Despite the few bugs it is overall very solid and very usable. iMovie is far too limited and I’ve never been able to get into Final Cut so it was good to welcome an old friend back to my system.

Doing NaNoWriMo yourself? I’d love a peer group and support network! Friend me!