I'm attaching hair/beard to the Cornelisz head. It's a bit awkward and nerve-wracking, as I know that every little hair is going to get more screen time than anything else. Hopefully it won't suck too hard.

Almost there - almost ready to make some actual shots. Been dicking around with various After Effects tricks - simulated camera movement, motion blur, parallax movement, etc. I'd like to be pretty well versed with that stuff before we start. Everytime you discover a new technique, there seems to be a brief 'honeymoon' period with it, where you want to really emphasize it and use it on everything. Then later on, you calm down and use it more objectively. That's where we want to be when we're actually shooting.

couple quick track tests, at different exposure times. A bit hard to see the results at such low res. But its something fun to look at.

We've been doing more photography test of the set. Some of this shit is pretty dry, but i'll try to lay out some of the technical stuff in case somebody ever reads this and it's beneficial.

We think we're going to be able to do HDR tracks (high dynamic range). That is to say, we'll be able to move the camera itself incrementally (taking a frame each time) to create the illusion of camera movement. This would let us make much more interesting shots. As well, the HDR shots have a graphic, 'storybook' look to them. When you see them with a still shot, they don't look real. But when the camera starts to move, something interesting happens - the objects register as 'real', but strange (good) looking.

The problem has been an extreme flicker that shows up when we do the HDR processing on a camera track. We thought it was just inherent with HDR, but maybe not. What it seems to do is emphasize any flicker dramatically. We also think that our lights are flickering in very subtle ways that don't show up until we process them in HDR.

So, the answer seems to be very long exposures, that average out the flicker. As well, we'll most likely be getting one of them 'power supply' boxes that give you an even, regulated supply.

I like this one. Most of them havent been too hot.

Couple more HDR pics. Still not right. Need to make the back wall lighter than the rest, with a nice gradient. Doesn't work compositionally otherwise.

Mr. Head, in HDR. With a bit of hair taped on. Though, if Dave can't solve the HDR flicker, we can't do it this way. Bummer. Head is pretty much done. Now it's beard and costume time.

Here's a quick track test with the current set, after strong HDR processing. This ain't supposed to be a good shot, or good HDR, or good lighting, but it tells us what we need to know. The original flics have a tiny bit of flicker, but the HDR processed shots flicker like a bastard. Dave thinks it's due to the basic nature of HDR. There's been other tests before and since; Dave thinks the problem might be solvable, somehow. I'll try to include technical details if we do solve the problem, for the benefit of others. If it's not fixable, that means no tracking with the sets. we can do imitations of pans via wide stiches and AE, which will let me do what needs to be done. But it'll be a bummer if we can't track.

Things are moving along now. Dave the Photographer is back, and helping out. Which is good, as i'm no pro with the lighting and whatnot. We gots a camera, software, and a place to shoot the big-ass greenscreened armatures.

We've been doing tests with tracking shots with the set. Multiple exposure HDR shots. It ain't working - the HDR processing is erratic, and flickers. It's looking like we're not going to be able to do camera movements with the set. Oh well. We're going to do some AE parallax shit instead with layers.

We gots a new camera and software. We're doing some HDR tests; HDR batch processing, in particular. Or, whether or not it's possible to take multiple HDR frames, same shot or otherwise, and make them all look the same without any weird HDR rendering glitches or flickering. Seems to be working so far.

A few knock-off HDR pics. They're a little over the top.

You can see how this works.

It's a bit awkward to use, but in retrospect, i'm pretty lucky to have made anything that actually works at all in any sense. It was a bit overambitious to attempt this kind of animatable head. So i'm going to count my blessings and move on.

Drilled all the peg holes, slapped the mask on. and did a quick test. A little crude, but not bad. The eyebrows didn't work - I omitted to hollow out the resin shell behind the attachments - i.e. they bulge out & get stuck. But it works.

Here's a quick little test using the aforementioned pin system. Not too shabby - this was done with only 3 holes. Should be smoother with double that at least. The whole skin of the face moves to an excessive degree, but that'll be corrected. Pretty sure this is gonna be my official wire-pulling system.

Here's the solution - i'm a little unsure about it. It seems a bit too simple, almost goofy. It's just a belaying pin. Tie the string to a pin, and stick it in a little hole. To pull the string more, take it out and stick it in the next hole. It seems crude, but it has its advantages - it's easy to tie the string at the appropriate tension, and the holes can be exactly located after tieing the string to the pin. No accidental looseness or leeway with the line.

The disadvantages - i'm a little worried that reaching in and relocating the pin will be hard to do without moving the puppet's head unintentionally. Secondly, even with carefully located tiny holes, i'm only going to have about 7 frames maximum to execute a facial movement.

But who gives a shit - it works and it's easy. I'd like to be more robotic and less nautical, but hey. What can you do.

Well, my plan of winding up the facial control lines with bolts (threaded through nuts in the top) didn't work at all. Not sure what the hell I was thinking. Of course, when you wind the string on the bolts, there's no tension to keep them wound. They just unwind. As well, the mess of hardware and strings and the trouble of tieing strings among a mess of hardware makes it unworkable.

I solved the 'attaching lines to the latex' problem. I tied the line to a tiny little washer, glues that to a little piece of latex, and then glued that on. It stays good enough, but they can be torn off (making them repositionable and test-able). Once established in their final spots, i'm going to coat them in latex to really bond them on there.

I did find a solution to the first problem though. Sort of.

The above picture is the top of the head with a pool of resin, to fill all the unneeded test holes.

There's a good face-yank. Still not maximum tugging, but getting there. The left cheek (your left) is being pulled, in case you weren't sure what was going on.

Here we go, some genuine facial expressions! Hoo-ray! top left is pretty sweet. Nice eyebrow lift. Top right (brow furrowing) aint so hot - can barely tell it moved. But it works. And i'll improve it. Bottom right works pretty good too - was expecting it to suck ass and look glitchy. It can be pulled further, but it was all I could do with my hands full with the camera.

Sweet.

It's looking pretty robotic and sciency back there now. Awesome.

I realized that I need to install more nuts in the top of the head, but still have access to it. So I chopped the back half of the resin shell in two, and resined it to the front half.

As i'm doing this, i'm keeping in mind future acces and repairs - i've got to be able to get back inside when shit breaks. Which i'm sure it will. I'm thinking that the 'skin' will be glued to the shell as little as possible, if at all. And the lower-back part of the head will bolt on, instead of being resined on.

Facial Controls

The skin-yankers are turning out to be easier than I thought. There's been glitches - the lines seemed to be firmly attached to the 'skin', but some of them popped off unexpectedly. Not because of the tension from pulling them, but because I bent them the wrong way, and the Krazy glue seemed to have cracked.

But the principle works. Here's how kids! The forehead lines run upwards (the direction they need to be pulled in), through little holes, and to bolts at the top. The bolts are inside nuts (expoxied to the shell). The lines will be pulled by winding them up on the bolts. They're allen-key bolts, and will be accessed through the top of the hat. Gonna be a trap door up there.

Did a little test for Cornelisz; hair, beard, hat, etc. I made his face dark to hide the eyes. They need a bit more work. It's really hard to cut and trim liquid latex. I'm going to try woodcarving tools.

I think he looks pretty ominous. He's not going to be so goth looking though; my aesthetic is going to be dark, greasy leather that blends with the studio.

I've found that the neck has to elongate to allow it to incline foward and back properly. This gets pretty complicated; it means another joint at the base of the neck, that I have access to when animating. And the head has to be able to move up and down on a rod of some sort. And, I have to be able to tighten the head to the rod very firmly. This makes things a little awkward for animation; I can hardly tighten and loosen the joint for every frame while animating. It will need to be tight, as the rubber mask/neck will be attached to the shoulders, which will create alot of stretchy tension when I extend the head along the theoretical rod.

There's one simple thing i'm going to try; there's a type of electrical wire covering you can buy at hardware stores that is pretty stiff, and bends like vertabrae. I'm going to go look at the stuff before I try to get too complicated.

Today was turning out pretty productive, but it turned into alot of head scratching and photoshopping.

I did a quick PS test to see how far above the neck the head should be. Then I started fiddling around with the layers, to see how well it would flex on the neck joint. It wouldn't have moved very much at all, had I done it the way I planned to; just a little gentle incline forward and back.

I didn't bring pictures, but I buffed the fixed mask (to remove the sheen), recut the eyes and mouth, and installed the teeth. You don't see much of them, but they're there. Also repainted and lacqered the eyeballs. Also did the resining for the back half of the head. Going to be joining the two halves together, next working day. After that, it's just a matter of attaching a line to pull the brow up and down. Going to try lines in the mouth too.

Then I get to install a beard.

After ruining my test mask, I got to do something i've wanted to try for a while; pouring a coat of lightly tinted liquid latex over the whole thing. Worked pretty good - it's nice and smooth now, and it fixed some of the little glitches from the casting process. Just needs to be buffed to remove the sheen.

I'm still making another mask, though. As a backup in case this one gets wrecked.

Not terribly thrilled with the job I did on the eyes. Liked him better when he had squinty Spaghetti Western eyeballs. As well, the rubber mask has been getting quite dirty and stained. I therefore decided to make another one (a pretty quick process) and use this one to play around on. I tried painting it with acrylics and liquid latex. Didn't work at all - the skin has to be be very smooth - any surface texture shows up to a great degree with high-angle lighting.

Eyeballs. After drilling the holes, I stuck my eyeballs in them, then formed a piece of sculpy behind them. After baking it solid, I trimmed it with a dremel tool. I'll epoxy it into place when i'm done fiddling.

He talks! The mouth opens more than this, but you get the idea. Eyes are coming out okay. Need to recut the shape of them - they look a little gimpy. The basics are going to be moving eyes and up & down mouth. After I have those finished, i'm gonna work a little more on some other facial expressions. If possible. The resin shell looks pretty beat up, but it works. There was quite a bit of breaking and fixing going on.

I had some pictures, but I can't upload them to the crappy old cafe computer i'm on.

I've got an animatable mouth now (used armature wire on the jaw). Also cut out the eye holes. Looks allright. Not fantastic, but allright. I'm expecting the head to look alot better when I paint it. I've been doind research on how it's done; think i'm going to try something different. I'm going to try either pastels (the powdery kind) or actual make up. Should be fun buying that. Then i'm going to spray it with drawing fixative. Should work good.

I'm almost done fiddling around with this head. After that, things should get alot easier - no more really tough technical problems. I'm feeling like this stage is the top of the roller coaster; after I get past this hard part, things should happen really fast. Hopefully.

Work continues on my 'robot head', but a little slowly during the Christmas season. It's not really 'work' per se. Alot of staring into space, rummaging through boxes of dodads and hardware, making little sketches, staring at the thing, scratching my head, and wondering what i'm going to do next.

In case anybody was wondering, the expressiveness of my animated heads is going to be pretty low. A bit of a weak spot in the film. But there ain't that much I can do about it; replacement heads isn't much of an option (probably using HDR photography, which shows up every little texture variation that isn't even visible to the naked eye). And what i'm doing already is ambitious enough for the time I have. Almost too ambitious; seems a little naive in retrospect that I thought I could whip up an animatable head in a short period of time. And none of it is turning out to be easy.

I'm going for an opening and closing mouth, at least, and moving eyeballs. I'm attempting to make animatable eyebrows, but I might have to go back to plan B, which is adhesive backed eyebrows that I can move around on the forehead. I'm also going to feel out the possibility of more movements around the mouth area, but not sure how successful that's going to be.

Mask turned out okay. No catastrophic suckage going on. The mask looks rough; partly due to the original sculpting job. It's also quite glossy, which shows up all the suckiness. Should look better when it's matte, and painted. This is also going to be an expirement to see how fixable it is. We'll see if i'll be able to fill in lines, trim off bits, etc.

Resin head is okay too. Going to need some reinforcement - it's very brittle and pronse to snapping as is.

I've finished latexin' up the impression, and just swished around coat #2 of fiberglass resin. The stuff smells. Like boats. Once it's thick enough to hold its shape, i'll be able to take the whole thing out, and see if it worked. I'm a bit concerned that I  missed parts with the lacquer, and the latex mask has bonded to the latex impression. We'll see.

I figgered it. After trying two different mold relsease materials (2 parts methyl hydrate to 1 part dish soap and 'mold lube') which failed completely, I finally figured out a way to make liquid latex not stick to itself; I sprayed the latex mold with lacquer. In my test sample, it didn't stick at all, not even the teeniest bit. The above picture shows the mold with lacquer, and with the first 'latex pour' after brushing the first coat. Just like it says; pour the latex in, brush out the edges, then drain the excess. Makes it nice and even-like.

If anyone reading this attempts the same thing, I recommend gloss lacquer, so it's easier to tell where you have and haven't lacqered, and thin-ass coats. If it drips, it's all over.

The liquid latex is giving me a hard time. It sticks to nothing at all; 2 part 5 minute epoxy glue won't bond it to itself. The only thing it sticks to is itself (even when applying over well-cured latex). And I need to make it not stick to itself. I'm trying version 2 of a mold release - the first one got beat. Once I get over this hurdle (hopefully), that'll be it. Shouldn't be anymore little problems preventing me from getting on with it.

The fiberglass resin is good; very strong and hard, and it doesn't stick to cured latex.

Well, the rubber impression worked. It's nice and thick and holds its shape well. And it's good that it's flexible. Wouldn't have been able to peel it off the clay head otherwise. Which is why I did it with rubber.

The pink thing is an alginate test cast of the inside of the rubber impression. It's full of bubbles (another indifferent casting job), but it tells me what it needs to, that the rubber impression came off without any glitches.

Aside from the more glamorous stuff, i've been doing inumerable little tests. The plan is to make two halves of a full-head liquid latex mask, which i'll then attach (in certain points) to a resin shell. I was thinking that I had to account for the thickness of the rubber mask when making the resin shell, and somehow cast the inside of the rubber mask. But a previous attempt with the mask told me that I might not have to; can just stretch the mask a little bit over a resin shell of the exact same size. It'll increase the size of the head slightly, which would be good.

But I have to figure out;

Will liquid latex bond to itself, and if so, what kind of mold release will work

Will epoxy glue bond the latex to cured fiberglass resin

Will fiberglass resin bond to the rubber impression; if so, mold release.

Can I make the shell out of pure resin, or do I need to embed reinforcement in it

Is the resin going to be brittle enough to saw and cut

Is the resin going to be so brittle that bolts will crack it

What colour and how much do I have to tint the rubber mask

Is the rubber mask going to be opaque enough so that the whole thing won't be see-through in strong light

There's lots more I can't think of of the top of my head. Point is, there's all sorts of ways for it to not work. And i'd like to forstall all those potential screw ups to make this thing work.

Made a plaster cradle for the rubber impression to sit in, to help it hold its shape when i'm layering latex in it.

A couple test casts. Lots of bubbles; it was an indifferent casting job. Just for fun. The one on the left is from my first failed rubber impression of the face. Not strong enough to hold its shape, and full of creases. I filled it with plaster and turned it upside down. Which is why its so squished. The one on the right is from back in the alginate days.

Sometimes I forget that you only get points for finishing things. And for what it looks like in the final film. If it's made of chicken wire and newspapers, it's just as good as titanium and fiberglass. If it looks the same.

Which is to say that I need to stop thinking in terms of just putting in hours. Like 'I had a good day because I worked for 10 hours'. Instead of finishing things and getting the ball rolling.

I keep putting coats of latex on the clay bust, but I swear it ain't getting any thicker. At least not to the point where it's going to be adequately rigid.

Here's the clay bust with the first half of the two-part latex impression, in progress. It feels like it's taking forever, but I only started with the latex yesterday afternoon. I tinted the stuff green because I was having a hard time seeing the edge with the latex drying clear. Plus, bright green is fun. Hooray!

Hands in progress. I'm making two pairs of hands; i'm always a bit iffy about making more than one of something before I know if the prototype works or not. But i'm feeling lucky with this one. The metal bits are:

5/32" hollow aluminum tubing, with a piece of 1/8" aluminum armature wire epoxied inside.

Designed me up some hands today. Needed something else productive to do while coats of latex were drying on the Cornelisz bust. I took a photo of my hand, made it into a sihouette in photoshop, then printed off a bunch of various sizes.

I did a couple photoshop tests to see how big I would like the head to be, ideally. Not much bigger than what I have, actually. I've found that i'd like a tiny bit more bulk on the sides, and another 1/8" on the chin. These things can be fudged with the hair and beard. And the hat. So that means that i'm using the one I have. Which is nice, as I really wasn't looking forward to going back to square one.

Here's that failed mask I was talking about. Top right is the mask in the alginate impression. Note white appearance of drying alginate. You can also see the mask wrinkling from the shrinkage. The mask looks pretty good, aside from the blotchy colour. Bottom picture is the mask next to the original clay head. It's too small.

Well, the liquid latex works. The 'whole head impression' thing ain't so great, however. To get it off, I just peeled it off inside out. After taking a test cast of the inside, I found that taking it off had creased it. It also isn't thick enough to hold its shape on its own. If I made it thick enough to do that, I doubt i'd be able to take it off.

So, i'm going to do liquid latex impression in two separate halves, and then try to marry the two resulting latex halves together by painting the seams with more liquid latex. Should work.

I'm also going to make the impression super-ass thick. I'm going to buy a big enough container of the stuff that I can just dip it in everytime I want to put a new coat on. Should be really fast to do.

I'm also fairly sure i'm going to redo the clay head. It's just not quite big enough, and it's bugging me. Should be fairly easy the third time around though. It's not too discouraging; it's always better once you've had a practice run. And I would have been bummed if the key piece of sculpure for the film, that's going to get the most screen time, wasn't as good as it could have been. After all, it's only about 10-16 hours of work in a project involving hundreds and hundreds of hours of work.

It seems that it was a good idea to do Cornelisz's head and armature first; he's got a beard, and a great deal of his head is covered by a hat. So it's comparatively easy to cover up suckage.

I've also solved another problem; up to now, i've been endeavoring to cast the heads in two separate halves - that means that i'll end up with a prominent line where the two halves meet on the finished thing. Not a problem for this head, as the beard covers it. But none of the other characters have beards.

I've found that I can cover the entire head with liquid latex (thereby taking the impression with one piece only), and then peel the whole thing off inside out.

I think i've solved the problem; what i've been looking for is a flexible material that I can use to take an impression of the clay head; flexible so that I can remove it even with the undercuts (if I cast it in plaster, it would get stuck). I just used the same liquid latex.

The stuff has different working properties than I thought; I applied very thick poured-on coats, and dried them partially with a hair dryer. Then I put more on. That is to say, I was able to build up a very thick layer in one evening. When I was making the rubber mask for my previous failed attempt, I was putting on very thin coats and letting it dry for half a day. You can really go hard with the stuff, and it will still cure & dry evenly.

So now i've got a complete rubber mould of the head. I'm going to take a test cast of the face area with alginate to make sure the details transferred correctly. Then i'm going to make the rubber mask, by pouring latex inside the other one, swishing it around, and draining out the excess.

Boy, i'm getting pretty sick of this crap. I'd like to begin making a film already.

FAILURE

It sort of worked. The liquid latex captured the detail very well, and it's suprisingly durable.

But the goddamn alginate shrunk over the week that I was applying coats of latex. The final rubber mask was about 2/3 the size of the original. Which means it didn't work.

Jesus, i'm just trying to take an impression of the front side of a clay head. It's taking up alot of time and money. Shit.

I'm doing some ambitious stuff with this film; it feels like the real challenge is to get all the stuff built, and working properly. Including the greenscreen composite stuff.

Doing the actual shooting doesn't seem intimidating at all by comparison. I feel like i'll be able to knock off shots very quickly.

Discovered something else about alginate; i've noticed that little thin pieces of the stuff lying around my table are turning white and becoming brittle; that means that the stuff has a limited life. Once it dries out, it's toast. Though I guess I could keep it in a tupperware container with some water. Like tofu.

I put the first coat of liquid latex on my alginate impression. The alginate is still sweating a bit. I hope it doesn't interfere with the drying of the latex. I did a test already to see if it would work, and it did. But these hunks of alginate contain alot more water than my test peices.

I bought a couple different brands of liquid latex - turns out they vary quite a bit. One is thick, one is very watery. They both have their uses.

Top picture is the head after the last pour. I found that using alginate on bare clay causes lifting of the clay, and can damage fine edges. Two very thin coats of lacquer fixed that.

After inumerable tests, I finally figured out how to get rid of the fucking bubbles. And yes, that's what type of bubbles they are. I tried a thinner mixture, more mixing, pouring water into the alginate and vise versa, but nothing made any difference. Then I noticed that when I peeled the alginate out of the mixing containers, the undersides were smooth. So, I epoxied a piece of wood onto the clay bust (see second from top photo) and put the sucker in upside down. After pouring in the alginate, I heaved the head up and down like a plunger for a good while.

And it worked! Finally. I can get on to the next step, making a rubber mask out of liquid latex from the alginate.

Here's what I mean by 'sweating'; after 24 hours, the alginate is still soaking wet and slimy-ass. The solution was just to use heat; a blow dryer, specifically. I'm sure it'd dry on its own eventually, but would probably take a week.