Christmas day photos

I took a few pics on Christmas day. Some looked pretty so I put them here

Frozen window

Old mans beard

Icy snow

Tree in field

Mysterious gateway to... another field

Another tree and another field

Snowy ruts

My bro throwing snow so I could get a picture

Making a Yew Self Bow

In October last year, a good friend of mine gave me a really nice yew stave to make a bow out of. Since then I have periodically picked it up, looked at it, wondered what kind of bow to make, and then put it down again.

Having made a few long bows before, I wanted to try something a bit different. All of them could shoot an arrow, but they would kick quite badly or broke after a couple of hundred shots. So I decided, after reading the bowyers bible, that I wanted to make a flat bow (Confusingly Americans call it a long bow).

The difference in design it that the limbs (the bendy bits of wood that give the bow its power),  are much wider and shallower than the English style Longbow (which I had based my earlier bows on). This means the extremes in compression and tension within the wood are reduced but the extra width in the limb maintains the the bows power. Also the handle section on a flatbow is stiff, this reduces the shock that can be felt when the arrow is shot.
 
So I started by removing the bark, checking the wood underneath, as there seemed little to worry about, I marked the bow up ready for cutting. Below you can see the basic design I followed

I then took the bow to a friend of mine who cut out the marked up shape with a band-saw. I was a little apprehensive about using a band-saw on the bow stave because; a - it can ruin it very quickly and b - I have only used hand tools on bows previously.

Yew has a unique property that makes it an ideal wood for making bows with. The heartwood (the wood that grows in the center of the trunk) resists compression very well. And the sapwood (the bit that grows on the outside of the trunk) resists tension. So when making a bow you need to put the sapwood on the "Back" of the bow, this is the bit that is facing away from you when you point it at a target (The part facing toward you is called the Belly).

So to get the bow working as efficiently as possible I needed to reduce the amount of sapwood on the stave so the finished bow will contain both heart and sapwood (the sapwood is too thick otherwise).

To take full advantage of the wood's natural strength, I wanted the back of the bow to be one single intact growth-ring. This will reduce the risk of a ring lifting and causing a spilt and ultimately causing the bow to break. So I did this by selecting a ring that looks wide enough to work easily and is far enough into the sapwood to serve as the back of the bow. I then used a draw knife to peel and prize away the wood down to about 1 ring above the one I selected. When I get to knots, I leave them at this stage. Maybe take of a few rings around them but it's very easy to chop straight through the selected ring.

Growth rings

working around a knot

I then use a scraper to work through the last ring, I work around the knots with a small knife until I have got the whole stave's back down to a single ring. Now this is done we don't wont to do any more work on the back of the bow, unless its cosmetic.

finished the back of the bow

The next stage was to remove the wood from the belly of the bow, this is a very time consuming process. I use a draw knife to reduce the the limbs to about 1" deep, I don't want to go much further with the draw knife as it doesn't deal very well with knots etc. You can end up ripping large swathes of material from the bow and before you know it - you have gone too far.

This is the point I stop using the drawknife

I then use a dreadnought file to reduce the belly further in to a gentle taper towards the tips, when reducing the limb I try to work so that the growth rings "point" down the center of the limb and to the tip. When this happens it means you are reducing the wood evenly and should have a nice even limb, if it starts to point or waver off to one side, work the other side to compensate.

Growth rings on the belly of the bow all pointing roughly down the centre

The next stage is a process is called "Tillering". The aim of tillering is to get the bow to a point where all of the wood is working evenly, ensuring that one part of the limb doesn't bend more than another (known as hinges), so creating an even curve when the bow is at full draw. 

As soon as the bow can be bent, its worth checking the tiller very roughly to ensure you don't have any hinges that need to be delt with immediately. This can be done by putting the tip on the floor, holding the other end and the pushing on the handle, don't push too hard, just so you can see the curve of the limb. If this looks OK carry on evenly reducing the belly. This isn't a very accurate indication of tiller but it does help in the early stages. If you can see any hinges, mark the area that is bending more than the rest of the limb. Then when you are reducing on the limb, just don't take anything off that area until the limb stops hinging.

Once the bow can be bent a bit more, I then use a tillering stick, a piece of wood about 40" long, with a cradle at one end to hold the bow and a load of notches down one edge to hook the string on to.

I then cut some nocks in to the end of the limbs, I use a rat tail file for this. They want to be about 1/2" - 3/4" from the tips of the limbs. I use a long string (a bit longer than the bow) with a timber hitch at one end and a loop at the other.

Put the string on the bow so its slack, put the bow on the tillering stick. Pull the string so it starts to bend and hook the string on to one of the notches. Stand back and inspect the bow checking for even bending and hinges. Use a pencil to mark areas that are bending too much a carry on reducing the limbs leaving the hinges.

Its worth noting that you should never draw a bow to more than its what you want its final weight to be. You can check this by putting the butt of the tillering stick on some bathroom scales when you are pulling the string back. 

You also need to make sure the both limbs are bending evenly, if you have a large piece of card you can mark out a chart on to it indicating limb tip deflex, or you can use patio doors( if they are fairly symmetrical).

using the tillering stick and patio doors method

So I just repeat this process, shortening the string, by adjusting the timber hitch, until I have got the bows brace height (the distance between the belly and the string when the bow is strung but not pulled) to where I want it (about 6 1/2 - 7"). Then until I can get the string back to my drawlength. 

This will take ages to get right, don't rush it. As you get closer to your desired weight and draw length only use a cabinet scraper to reduce the the belly.

After tillering, I used a heat gun to heat the last 8-10 inches of the limb whilst its clamped in to a reflex shape. Not too much as it will add more strain in these bits of the limb. Leave it to cool whilst clamped. This just adds a bit more poke and makes it a bit smoother to draw. 

adding reflex to the tips

Then I used superglue to glue layers buffalo horn with other pretty bits of wood (in this case bamboo) to make the knocks. I then shaped them with some files and sandpaper 

gluing up tips

Nice tips

Sand it back going from p 100 to p 320 then some fine wire wool, I finish it with 3 layers of danish oil, then some pre-cat lacquer, and then some beeswax furnature polish

Then add a leather handle

Slightly dodgy looking handle

Take a picture of it in the snow

bow in snow

 Then Shoot it

Me looking very serious

The final bow is about 65 lbs @ 30" and 71" long but It will probably reduce in weight a little (I hope) as I shoot it.

P.S. This is how I make a bow and by no means definitive , I am not a trained bowyer, I have read bits and learned bits myself, then wrote it here so other people like me can read it and geek out a bit. 

P.P.S.S. Be careful when you are making bows, a limb exploding has quite a lot of oomph and could hurt. Also yew is poisonous so use a mask when working with it, especially sanding.   

Out in the cold

These just some pictures of a load of things with ice on them.

Championsheeps

Built by Aardman Digital, Championsheeps is a suite of games tied into the bafta award-winning animated series; Shaun the Sheep. Though principally aimed at the same demographic as the series, the games have proved very popular with adults and older teenagers, offering as they do, both instant pick up and playability as well as more difficult, technical challenges.

The five games offered within the suite spanned a number of genres, though those I built, Flock Together and Sheep Stack, are both physics-based puzzle-ish games. The former, a Tetris variant, challenges the player to drop the leotarded ruminants into colour coordinated piles, whilst the latter is more of problem solving game involving levers, balances, hungry sheep and an underpants catapult.

Initial animation test were done by Robin Davey and Phil Webster which looked fantastic and made everyone involved in the project very excited. Some mock-ups of initial designs by Gavin Strange  further added to the anticipation. Then the prototypes were developed for us by Gaming Your Way and we could see that this was going to be really great once it all came together.

At this point I thought I would make a lot more work for myself and decided to broaden the scope of the games I was making from simple, single-level progressive games to multi-level challenges with a greater variety of problems and obstacles for the player to negotiate.

As this were my fist stab at physics-based games (I had dabbled before but nothing on this scale), Championsheeps presented me with a new challenge in learning how to easily design and edit levels that would work in Box2d. Chris Underwood pointed me in the right direction, using a technique which allows easy production and editing of levels using the Flash IDE.

Championsheeps has been live for three days now and has already clocked up 90, 000 views. Despite being one of the more stressful projects I've worked on, I am especially happy with the results and the reception it has received.  

The Farmyard

Leader boards

Sheepstack

Flock Together

Now a geeky bit

As well as the games, I also built, with the assistance of George Crabtree and Richard Davey, the main structure of the suite environment. This handles scores, user data and loading and unloading the games. As expected it presented quite a few of its own challenges, one of which is as follows.

The problem with loading one swf into another is the communication between the two. This can be done reasonably well using bubbling events, but that relies on the dispatcher being on the display list which I don't really like. You could write a shed-load of interfaces, or if you don't mind remembering what all of the method names are and leaving everything untyped you could do it that way (gets very messy though). But I am lazy and I like Flash Develop to fill out all the bits it can without me writing any extra code.

What I wanted was a single place I could dispatch events from and listen to.

So I started playing around with static vars in classes and experimenting with what overwrote what when one swf was loaded into another and they are referencing the same class. I discovered that the host swf always wins, and they do both end up referencing the same value.

This meant I could set up a commonly shared registry (just a class full of static vars) in which I instantiate a listening post (essentially a class that extended EventDispatcher). I used this to communicate from game to manager and back again. It also had the functionality to turn on and off events from all of the games except the one that was playing, so no erroneous events firing and games starting, stopping or quitting when I didn't want them to.

As long as I didn't reference anything in either the games or the manager from the commonly shared classes I wouldn't end up compiling one thing in to another.

There were plenty of other challenges that arose in building this, but I thought this was of note because its quite a nice way of dealing with multiple swfs - all with their own logic and all needing to communicate with one another.

update  18/03/2011
look

On the bbc homepage

The Tate Movie Project

The idea of the Tate Movie Project is fantastic; to get kids to contribute their ideas to an animated movie. They'll create every bit of the film, from drawing the characters to writing the plot and recording the sound effects. The site, built by Aardman Digital, takes the form of a film studio beautifully illustrated by Robin Davey. You can go to different departments to make various aspects of the film and you will be guided around the set by the crew; characters developed and animated by the ever-so talented Felix Massie.

My role as lead flash developer was to put together all of the beautifully crafted bits and bobs to make them work as a microsite whilst maintaining the subtlety inherent within the design.

I also built the tools that allow the kids to play around with, or submit their ideas and creations to the film. They can write scripts, make their own animations, record sound effects, send in pictures.  They can also save their ideas and add to them later before getting them just right so they can be submitted to the director for inclusion in movie.

There are loads of hidden treats waiting to be discovered within the rooms and awards to be won for finding them. One example is that you can knock musical dandruff out of the composers hair, pick it up  drag it to a stave which will start to play the music you are creating.

I am very proud to be a part of the team that put this site together, why not take a visit and look around.

Oh and BTW this recently won a BIMA award for best website.

Just some nice photos

These are just a few photos I thought I would share. Some are taken in New Zealand, some in Britain

Show Me Show Me

The Show Me Show Me microsite supports the cbeebies show and made by Aardman Digital. Based in a playroom high in the sky (the top of a ten-storey tower block), it offers a number of games and puzzles for children aged 3-6. 

I worked with the designer Dave Gillet and together we created a warm, safe but exciting environment for kids to enjoy.

Each activity was loaded in from a separate swf and was originally designed to be deep-linked into, and navigated a using the browser history. However due BBC hosting restrictions this functionality was lost

The playroom

Make a picture, a place to add your own things to a picture

A dance along activity where you can choose what dance you want to do

Counting quiz instructions

On finding the correct number of items
you are rewarded with a fun animated sequence

Choosing a story

Searching the sky in cloud hunt to find things hidden in the clouds.