Wednesday, 28 December 2011

Specially Commissioned Banners

Hey, I'm sorting out this blog slowly starting with the advertising, which never made me any money anyway.  I'm installing custom banners which my best friend Julz drew for me.  They'll both link through to her comic, which is rad.

Saturday, 24 December 2011

For Your Spare Time...

Since I'm currently writing this on a laptop that was manufactured no more recently than 2001, I'm unable to tidy up the films I made this term into an uploadable state, and won't be able to until I get in front of a computer that's less than 6 years old, which might be not for a week or so.  In the meantime if you need some alternative programming, which leaves you asking 'good God, what program did she use to make those?', and 'where can I find a tutorial on how to produce something with that kind of cutting edge graphic quality?', I can strongly recommend Stuff & Nonsense, by my sometime housemate and collaborator Julz.  Its rad.  Interrobang using the ask box.

Thursday, 22 December 2011

RCA Term 1 Experiments

I'd been on 4 hours sleep per night for many weeks in a row including weekends, up until last Friday when the 'holidays' started.  The RCA is rad.  I barely have a moment to write about it now, but at some point I'll photograph some of my scribbles from the 60 hours or so of life drawing classes I've been to.  If I get a moment between watching films, catching up on reading, and scriptwriting for next term's film I'll tidy up the proper work from this term and post it too.  In the meantime here are some tests.

30 minutes at the lightbox

Rhythmanalysis tests

Sunday, 16 October 2011

Stop Motion and Memory Drawings

 Two of my daily memory drawings for the drawing elective I'm doing at the RCA.

Thursday, 11:30pm. Two deer caught in my torchlight on my cycle home.
Wednesday, 12am. Occupied bus shelter.

And some work in progress building stop motion ball joints (with drilled out brass flats and some ball headed bolts), and a twisted wire, foam and latex fox.  Used a blowtorch and a thread tap for the first time this weekend.  Exciting stuff.

Animation Process Workflow Recap

I was recently asked to crit a piece of work by someone starting out in animation.  I took the time to type up my workflow notes, not only for sharing, but also as a way of recapping what I've learnt in my whirlwind first year in animation (my Pg Dip at Central St Martins).  These are a mish-mash of different peoples' advice that I've received and noted down over the last year, and come with no particular warranty!

Aside from concerns of how the style relates to the narrative or the function of the piece (form follows function, or a house style), style can partly be dictated by the quality of the animation you're aiming for.  One of the best pieces of advice I ever got was to move away from cutout animation and try drawing.  Cutout animation is realistically only efficient for making limited animation on a fast turnaround - its a real struggle to make 'full' character animation in cutout.  In cutout animation poses typically end up being very weak as you don't have to engage with creating the pose as much as you do in drawing (by default the assembly of the cutout pieces resembles a pose, so people tend to do the minimum posing work necessary, whereas every drawn pose must be created from scratch), and the restriction of the media prevents you from working with proper three dimensional poses, in perspective and so on - you might find it hard to progress your character animation skills if you limit yourself to cutout.  Compositions also often end up being very flat and dull, and thinking spatially, its hard to bring the audience into the scene - the flat look resembles a long lens, and gives a voyeur's viewpoint in contrast to a POV which might be obtained with a wider lens (less flat composition).  Unless you've storyboarded or thumb-nailed strongly its likely you'll make compromises in the animation stage due to the awkward workflow that cutout has.

Limited animation is fine for a certain type of work, but the work might have to rely heavily on dialogue (as for example Hanna Barbera cartoons do - see "Limited Animation Unlimited Imagination" in this set of essays) or on a fast paced story line (for example, "The Simpsons" is an example of animation where story is prioritised over animation), and unless these are really great, you'll have people turning off very quickly.  Try sitting through the whole of the Beatles film, the Yellow Submarine and you'll see what I mean about limited animation failing to hold attention.

Drawing is so important, not necessarily for image making, but as proof that you can observe correctly.  People often say they can't draw, but really the problem is that they can't observe, and if you can't observe, then you stand no chance at all of being able to critique your own work and will constantly be relying on others for advice.  If you can hold a pencil and draw a straight line, a circle and a square, and you are capable of tracing, then there is no problem with your drawing facility, but rather with your observations.  The word animation literally means breathing life into something, and so, no matter how far away from realism your style might be, its still essential to reference back to real life.  Draw lots of poses from life to practice (if you can't find anywhere to go life drawing, get out on the street and draw people who are moving as fast as you can, or try posemaniacs if you want a way of checking how accurate your sketches are).  Crit each drawing you do yourself, draw corrections over the top in a different colour (see John K's advice).  Pick a scene of your favourite film and break it down into key drawings as if you were rewriting the storyboard.  Draw from music to help capture a mood, literally draw everything you want to capture in your animations as it will be a way to proving to yourself that you've processed and understood what you want to capture.  Also physically get into the poses that will be in your animation - feel where the weight is, what aches when you hold the pose for too long, if you're off balance, which way will you fall?  Is the pose natural, would moving a limb make it more natural.  How does the pose make you feel - commanding? subservient? confident? vulnerable? open? and so on.  Draw with both hands, one after the other, and at the same time.  Try to make drawing automatic if it feels unnatural, so you don't have to think or worry about moving your hand over the paper.  Sit a real person next to a life-size doll and draw the scene trying to capture the life in the person and the stillness in the doll.

Act out the scene a few times trying different things, then record it when you're confident you know the scene so well you can act it as if it was real (ie without any trace of your 'acting' showing through).  The director Bresson used to get actors to repeat scenes hundreds of times until he felt they weren't showing any acting - he referred to his actors simply as 'models' who had to follow instructions rather than invent on the spot every time (See Notes on the Cinematographer for more).  You should know the action inside out before you even begin to animate.  Make sure the action is natural - would you do it that way?  Ask someone else to do it without showing them how, without them acting, and without them knowing your watching or recording their movement - you'll have to set it up somehow eg "hey could you pass me that X over there".

Working from the video draw thumbnails of every single frame. (see Rainplace for a great workflow idea).  These can be very rough to start with, but by drawing every single frame, not just the keys, you start to learn how long each action needs to be in terms of frames, and also you feel the flow from one key to the next if you're drawing fast enough (Shamus Culhane used to work frame by frame drawing very fast so as not to loose track of the flow).  If you can't imagine the movement or draw it out from memory without looking back to notes, it will really slow down your animation.  I find that making a neat summary of what I've learned really helps afterwards - here are some examples from my own notes: 1 2 3.

Once I've finished going through stop frame, I'll play the reference back on loop at real time speed and try to tap the rhythm of where I think the keyframes should be (footsteps, head turns, etc) as if I was learning the beat of a song.  If its lipsync, rather than working with a dope sheet with phonemes, I personally prefer to speak the dialogue with my finger on my top lip and my thumb on my bottom lip, feeling when my mouth is most open and closed (the most important part visually), and then on a second pass, where it is widest or narrowest.  For 3D animators Stop Staring has some great advice.

Once I've internalised my reference and planned out my keys either as thumbs, or ideally as full size sketches (I tend to block in 2d, even though I'm animating in 3D as it gives me a much better grasp of how the poses work and really helps to record it in my mind, which I don't think I get in the same way if I do my blocking in 3D), then I start to animate.  Even if you hate disney style animation, you can't ignore the 12 Principles of Animation which are present pretty much in any movement.  Disney chose certain aspects of them to exaggerate (notably squash and stretch and appeal), and certain aspects to underplay (personally I always feel that their staging is sometimes dull and straightforward - I would guess for efficiency and clarity over drama), but if you pay a fairly even amount of attention to all of them, you'll be heading for life-like animation.  The summary on wikipedia is really no replacement for the full version in the The Illusion of Life which is definitely worth a read if you haven't already.  You should really know these principles off by heart and be applying them all at once without having to think about them.  Its not good enough just to be able to list them, they have to be happening automatically in your work without thinking about them.  Its easy to find examples of them all in animation, and the disney book has great illustrations which go through each principle, showing it visually - try copying a few sequences out if you don't get what the text means.

I wouldn't start to animate a scene before I know whats going to happen in it, and have planned it out, at least in terms of major poses and keys, in thumbnails at minimum, but ideally on layouts at full scale.  I know some people like to feel their way through without thumbnailing first or without doing a rough blocking pass, but usually I find this costs a lot more work, and makes the edit much harder.  At the very least you need to know how one shot connects to the next, so you have to have a sort of target pose to reach by the end of the shot.  If you choose to go straight ahead (as I do for most shots - I work in 3D as if its stop motion, animating frame by frame, and completely ignoring the graph editor), then I'd strongly recommend using a type of "pop-through" technique, where you work out the timing on a separate rough pass (either with live action stills or quick sketches), then start again animating straight ahead, but aim to hit the poses on the times you planned out.  You could go the whole hog and film live action and rotoscope to it (even in stop motion or 3D), but ideally I'd recommend reducing a live action shoot to a selection of keys and then sketching over the keys to exaggerate them.  In Bresson's Notes on the Cinematographer he discusses how a photograph of a sculpture has none of the artistic (or monetary!) worth of the original, and as such a video of an actors brilliant performance has none of the brilliance of that performance (in the way it does live in a theatre).  The video needs to be resuscitated - Bresson does this in the edit - as animators we have a second chance to do it with exaggeration.

To take an example of the pop-through method:. if you know you have to hit a contact position on frame 9, and its a fast walk, then you're going to need to be around the up position on frame 7... ie you work backwards from the key you need to hit to work out how close to it you need to be on the frame you're currently drawing or posing.  That way you get the nice flow of straight ahead, but none of the unpredictability.  If you're working in 3D straight ahead you might benefit from some tools to smooth over glitches - tracking motion paths as well as the Push & Relax tools found in blender are very useful.

Once I'm done, if I'm not happy with the result I have a big troubleshooting checklist to go through.  There are tonnes of books out there for learning animation methods, but nobody really tells you what to do when it doesn't work.  Start with the big questions first...

Staging - Are the characters in the places on the screen which would both ease the cut (if that's your aim), and communicate the situation as best as possible? Is the character above or below the camera's "eye" level if they are dominant or suppressed respectively? Can you see a facial expression if you need to?  Have you needlessly close-framed to try to get "added drama" (point 3 in Bordwell's list)?  Is the camera obeying the 180 degree rule (and the 30 degree rule - alternating angles and shot sizes to avoid jump cuts)?  If not are you breaking the rules for good reason, or establishing your own set of editing rules?  Is the shot size and location supposed to convey information about a POV or not? Does the shot clarify or obscure the geography of the scene?  Is there a continuity error in the staging?  Is the camera moving? Why (If you're Ken Burns-ing/Hitchcock Zooming every shot how will you hike up the tension when you really need it - see point 4 in Bordwell's list)?  If its moving check the animation with it still if possible to minimise the number of things distracting you in your error checking process.  If you're not sure which shots to use check out David Mamet's On Directing Film, (sometimes more cheaply available in the compilation book A Whore's Profession which includes some other essays on screen-writing which I haven't got around to reading yet).

The staging can have a big effect on the length of a shot - ie. do the characters spend ages walking around before we see any relevant action?  Is it OK to cheat the position of characters within the shot to avoid lots of unecessary locomotion? (for a great example of cheating staging see from 2m30-2m40 in Tarkovsky's Ivan's Childhood: Ivan walks about 500 metres in the space of half a second but the audience accepts it because it has been done for emotional effect).  Is it OK to cheat physical space across a cut? - see Edward Dmytryk's seven rules of cutting (in On Film Editing) for when its acceptable to do this.  If your staging is wrong you might have to reanimate the shot - get it right in the thumbnail stage if you can, then block at layout scale to check the poses still work when timed up.

Composition - Much like staging, but in 2D screen space (whereas I consider staging to be primarily concerned with the 3D set or world space).  Hit the key actions on the third guides, use leading lines in the set to focus the eye to points where action occur.  Are limbs cut off awkwardly by the cropping of the shot in a distracting manner?  Do the visual shapes the characters make communicate well and integrate into the background well (not just silhouettes - consider complimentary colours and brightnesses)?  Check your title safe area - make sure nothing important happens outside it?  Eyes will be drawn towards areas of high levels contrast - check the animation most thoroughly there - also consider if the composition causes strobing problems - movements which are too fast and in high contrast colours (for example I had a black wing sweeping across the screen over a white background very fast, and it just flashed/flickered due to the extremes of brightness).

Readability - Can the audience tell whats going on - show the shot to someone who hasn't seen it and isn't an animator and ask what they just saw.  Closely linked to this is silhouettes.  If you're working in an overly graphic or cartoony style which lacks clear definition of volumes, check that the shot works in silhouette and that you can tell whats happening from the silhouette.  Be careful though that working in silhouette doesn't cause you to choose a dull flat Disney-esque composition every time.  Use it more if you're communicating something crucial.  In Sculpting in Time, Tarkovsky argues that as a director you must leave things ambiguous rather than emphatic, and not spoon feed the audience as if they were stupid.  Stories are more interesting and exciting if the ambiguity that you create is not so much random, but adds to the plot development emotionally (deliberate confusion or disorientation) or logically (an alternative plot line).  You should also expect a degree of competance from your audience - I think its in Werner Herzog Eats his Shoe where Herzog talks about the problem of visual illiteracy in audiences today.

Line of Action - As well as planning poses starting with a line of action, assess animation by working backwards from your finished animation into sketched lines of action with a grease pencil - are the lines what you were aiming for? can you reinforce the line in the pose? Does the line show either (1) where the force is - if the character is static and/or (2) where the character is moving to/from if its in motion and (3) what the emotional state/mood of the character is?  is it possible to get all three of these elements into your line of action without compromise, or would you be better to focus on just one of these for this scene?  For example, if the plot point is being conveyed using acting within a shot (or an inflected cut to an emotively emphatic pose), rather than locomotion, then it might be better to use the character's line of action to show depression/elation rather than using the line to show where the force propelling the character is.  Has the timing of the animation given enough time for the pose to read?  Eric Goldberg talks about 'attitude walks' in his Crash Course book - where he adds eases around the contacts - the keys- which most express the attitude of the walk through the pose.  If the line of action is right, check the pose - if I'm working in 3d and I have trouble with a pose I'll draw a copy of it by hand.  If I have trouble drawing it, then I know its probably an unnatural/uncommon pose (one I'm not used to seeing in life drawing!) - I try to draw something else then put that in.  Remember even if you're working in 3D, every frame is ultimately a drawing - the audience sees every frame for an equal amount of time - its no good getting just the keys right!

Pacing - Essentially timing, but on the scale of the edit.  Is the story moving too fast or too slow?  If you're working with a three act structure have you spent enough time on each part: the setup, the conflict, and the resolution?  Have you given the audience time to adjust to the shot (half a second of wasted screen time per cut).  Digital editing encouraged whats known as frame f*cking (see Intensified Continuity: Visual Style in Contemporary American Film, a paper by David Bordwell) which is when the editor replays a sequence of shots over and over, each time removing a frame at a time from one shot until the shot is as short as it possibly can be (contrary to Edward Dmytryk's rule 2: cut long rather than short).  The short cuts problem originated with action movies, but the disease has spread so that average shot lengths for feature films are now in the region of 2-4s rather than the 6-20s common in the Steenbeck editing machine days; obviously when manually splicing film, frame f*cking was a much more laborious process.

Timing - This is where you really need to go back to your reference to check, or start tapping out beats on the desk.  Unfortunately the timing from live action and from blocking doesn't necessarily always translate across to animation.  If the timing's wrong, then why is it wrong?  Don't just start shuffling keyframes randomly (see the update to Keith Lango's shuffle key method, or jump to the non-shuffle method), try to quantify where your mistake was so you don't make it again.  Push the timing from real life - there are many examples in live action films where timing has been deliberately contorted for dramatic effect (in The French Connection the car chase was shot undercranked by Owen Roizman - ie. sped up, and throughout the Mirror, Tarkovsky dips in and out of overcranked shooting - ie. slow motion, but often barely noticeable - the best example is towards the end of the movie when the boy is waiting alone in the dark room; see some great notes on his techniques here)  Try timing to music if you have no dialogue to work with - in The French Connection all of the edits in the car chase sequence are timed to beats of Santana's Black Magic Woman.

Spacing - Relates closely to timing, but has more to do with poses.  Are the poses spaced closely enough together to carry the eye from one to the next?  If the spacing stays relatively uniform between consecutive frames then you have smooth flowing action.  If on the other hand the spacing varies dramatically between frames you probably have sharp choppy movements which might not be what you want?  Examining frame by frame spacing (with motion paths) might be too subtle to see errors (ie. at the inbetween level), maybe consider the spacing between every other frame, or between consecutive keys.  Big changes in spacing indicate a big change in momentum (either a large force or a light object), wheras small changes in spacing indicate a small change in momentum (a small force or a heavy object with a lot of momentum).  A change in momentum requires energy - either new energy arrived, and if so where from (eg. someone pushing something), or energy was lost (eg. something hitting into something else and coming to a stop).

Arcs - related to spacing - do the traces of the end of the limbs sweep out smooth arcs and are there enough frames to make the arc noticeable?  If you're moving from point A to point B over 1 frame you can only go in a straight line - its impossible to go in an arc.  If you go in 2 frames, then you can at best go in two straight lines (which will be too fast for the viewer to see an arc).  If you have say 5 frames to get from A to B, then you have 5 straight lines (or chords) which will join together to approximate an arc.  If you have too many frames and your object moves too slowly, then viewers won't notice the arc at all, but if you go too fast, then there won't be an arc to see!  You can reinforce the shape of the arc with follow through - leaving appendages trailing behind in the arc.  Check the arcs from the camera view.  Draw the shape of the arcs over your background - do the shapes fit nicely in the composition of the background element, or do they clash?  Consider the motion paths that your characters leave behind as parts of that composition - think of the vapour trails left in the sky by aerial acrobatics, or the stripes on a frozen lake from figure skaters - the motion is ultimately beautiful because of the arcs it produces.  If you climbed into a pitch black suit in a pitch black room and wore flow in the dark spots on your knees, toes, fingers and elbows, and started walking, an audience would still be able to identify the motion of the glowing spots as being a person walking even if they couldn't see you.  The human brain can identify motions just from the arcs they trace out.   Laban breaks down motions into three types: spoke like, arc like and carving.  if an arc isn't appropriate would a spoke or carve fit better?  Acting for Animators has more on this.

Eases & Acceleration - An ease can be used to emphasize a key - ie. to spend more time around that key, or just to provide a sense of acceleration from one key to the next.  Don't confuse acceleration and speed.  Check the easing between your keys - something heavy will need a bigger ease, something light might not need easing at all.  What caused the ease - is it related to a force, or to a visual hold to help the audience see a key?

Forces, Leading Limbs & Weight - Forces set a structure into motion, or stop the motion.  Forces allow a motion to be changed.  Weight keeps a motion going (ie. gives something momentum) and accelerates it downwards.  Remember that weight never takes a break and will start tipping a character over as soon as a leg is raised.  Don't just consider how weight works in a walk or a jump, think about more complex motions aswell.  A skateboarder can move forward to some extent by weaving from side to side - they lean off balance, their body tips forward gaining speed, then they steer back the other way bringing their board underneath them.  A surfer on a day with bad waves does a similar motion pumping the board from side to side, and by weaving he turns his weight into a forward motion.  Weight doesn't just pull things down, its a key element in moving forwards.  Without gravity, even if our feet were somehow stuck to the ground, we would be shuffling rather than walking and would have to do a lot more work to pull ourselves forwards, because we wouldn't be able to lean into the walk.  By leaning off balance we give ourselves forward speed.  Leading limbs - check which limb moves first - does it draw the eye to the right place? is it the natural limb to move?  For example a dog will usually start walking with one of its rear legs first because in terms of phase difference the rear legs are a half cycle ahead of the front legs.

Secondary action relates to the leading limb.  if the leading limb is the first thing to move, try to work out the chain reaction of what will move next - My brain wants to move my foot, but really the leading limb is my upper leg, then my lower leg follows and my foot drags behind.  Try not to think about what you want to move, but what has to move first.  Working up the chain, my uppper leg no longer has support on the ground, my hips drop on one side and this propagates up my spine - my upper body adjusts to move my weight over my other leg and you get counter-animation (listen to Eamonn Butler on Animation Podcast if you don't get what counter-animation is) around my neck and shoulders to keep my head level - all one after the other, all time offset (ie. overlapping action).  If you combine 'secondary action' and 'follow through' you get successive breaking of the joints (like a propagating wave).  Follow through happens in space (ie. appendages following (and shaped to) the motion path), but also in time (one thing after another).  The way our bodies move is like any structure (say a tower of cards) one thing affects the next.  Check your know which movement is deliberate (independant action), caused by the character, and which movement is due to external forces, or secondary action.  When I started out I was easily confused by the difference between secondary/independent/overlapping action & follow through, and for the most part treated them (wrongly) as one thing, but realising that they are in fact four separate concepts to implement was a big help.  Don't confuse any of the different types, and make sure you've covered all of them!

Squash and Stretch - Self explanatory for cartoon styles.  Always check you've maintained volume.  What is the direction of the squash (in line with the stopping force) or the stretch (following the shape of the drag)?  How quickly does the object squash or stretch - both require energy - is this a high or low energy situation?  For rigid objects you can use the idea of squash and stretch to form the pose without actually squashing or stretching the anatomy of the character.  For example an example of a 'stretched pose' might be if you stand up on tip toes with your legs and back straight and your arms above your head and fingers reaching as high as you can - without distorting the limbs you still get the idea of a stretch.  A squash might be a crouch - imagine you've jumped from a great height - you don't squash elastically like a ball, but rather you crouch to absorb the force and spread your limbs sideways somewhat.  Consider the squash and stretch to work on the bounding box of the character in the case of realism.

Anticipation - With aniticipation you need to check (1) the size of your anticipation and (2) the duration - two separate things - don't talk about a 'big' anticipation, be specific - is it a big movement over a short time, or a big movement over a long time, or a small movement over a long time?  What exactly is big?  In general if the action is physically large, then the anticipation will be physically large.  If the action is physically small, then the anticipation will be physically small.  However if the action which is about to happen will take a long time then the anticipation will be over quickly (so as not to bore the audience, or tire the character - for example someone about to run a marathon doesn't have a dramatic anticipative movement before the start the race).  Vice versa if an action is going to be over very quickly (eg. someone hitting a baseball with a bat) the anticipation needs to be long to give the audience time to read it as the actual hit will be over very quickly.  The size of an anticipation is proportional to the action, whereas the length of an anticipation is opposite to the length of an action.  Sometimes surprise is key though, and giving the audience no warning at all can be the best option.

Model - Finally I check that the character hasn't 'broken' the model or the rig in any way - I do this last because I don't want to be animating around problems, rather I animate as best I can, and hope it works, and if it doesn't I'll try to fix it on the model rather than by fiddling with the animation.  Check details like eyes and stiff fingers - hands can be really expressive if animated correctly.

Phone a Friend!
Only once I've gone through all of this stuff, do I then phone a friend for crits.  Learning to understand/analyse what's wrong is probably the most important skill in learning to animate.  It's very easy to look back at a piece of my own work and says its rubbish, but working out "what smells" (as Jason Osipa puts it) is the difficult bit.  The easiest way to spot mistakes is to have a systematic way of looking for problems (I use this checklist I've typed up which has most of my common mistakes in, but I'm sure I have many other common mistakes I haven't spotted yet), and also to have another way of looking at your work.  If you're struggling with a drawing then looking at it from a different angle or mirrored to see whats wrong.  If you're working in 3D and you've already seen the problem pose from all angles then draw that pose by hand and you'll soon discover whats wrong by reconsidering the pose in a different medium, or at the line of action stage.  Reference everything and don't rely on reference other people have given you - discover it yourself from videos and sketches - that way it'll stick much better in your memory.

That's all for now.  I'm planning on switching to stop motion this year so I can animate faster and hopefully do some dialogue & acting work which I haven't had as much time to do in 3D.

Friday, 23 September 2011

Pass the Flipbook

View scratch on Pass the Flipbook

So Pass the Flipbook is now stable, and can also print real old school flipbooks from your animations.  There's a new interactive 'planets' browsing mode which lets you see how animations relate to each other.  There's also an embed option so once you've finished drawing, you can share on your blog!

I'm working on a Sketch-Roulette mode as well where you only draw one frame, then pass it on.

Monday, 12 September 2011

Beta Version of The Light Box Swingers Club Goes Live

Hey Folks,

Just to say that a beta version of my side project, The Lightbox Swingers Club, has just gone live.  Its an online flipbook and animation collaboration application.  Please try it out and let me know what you think.

Choose an animation you want to extend and get drawing!

For anyone interested in the technical info, its coded in HTML5 canvas, with javascript doing the interactivity, and php with mySQL running the databases.  Let me know if you run into any problems.


Thursday, 8 September 2011

Animating Dogs: Breaking into a Gallop from Standstill

This evening I was looking through my reference footage of Rufus (a friend's dog who I borrowed for a weekend back at the start of the summer) when I stumbled across an interesting gait (or combination of gaits) which Rufus uses when he breaks into a gallop from complete standstill (lying down in this case).  I was actually searching for front view footage of a gallop to study how the legs are abducted under the body, but this accidental discovery has helped explain why a shot I animated last week came out looking slightly unnatural.  How serendipitous! (As an aside, I put that word in for the enjoyment of the non-native English speakers reading this blog as its a phrase which allegedly has few direct translations to any other languages, it's also how Dave Prosser describes his BAFTA nominated short, Matter Fisher!)

Have a look at the image below made up of a few random frame grabs from one of my reference shots.  Notice how the back legs make a bounding motion (as they do in the gallop, but in this case they are completely twinned).  Even if there's no reason to avoid twinning, I find that when I'm animating, I now do it out of habit, which isn't necessarily a good thing.  In fact avoiding twinning at all costs is probably just as much of an animation sin as twinning negligently is!  The rule really should be 'twin appropriately'.

In the image below the two back legs are exactly in sync with each other.  My guess is that this is so that the drive is as powerful as possible, and so that the best use can be made of the folding and unfolding motion of the spine and the hips.  The spine, the rear hips, and the two back legs all work in time, and together they act like a piston expanding and contracting.  Basic physics tells us that for a big acceleration we need a big force.  Anyone who's been in a tug-of-war (or a rowing boat, or pushed a piano up a set of stairs) will tell you that to get the biggest force you need everyone to pull together at once.  This is what the rear legs, hips and spine are doing - lots of working together, lots of twinning, no time offset!

click on the image to enlarge

Right, so if a dog can get the greatest force using both rear legs, the spine and the hips all in sync, why doesn't it do this all the time, in every gallop.  I'm not a biologist but I think the simple answer is that it's too tiring!  Having all the parts of the body working in one huge push together just isn't sustainable... remember the tug-of-war and pushing the piano up the stairs analogies?  A large force gives a large acceleration which gets the dog up to speed, but once the dog is at full speed it doesn't need to accelerate any more, only to maintain speed, so it switches into a smoother more economical motion which involves offsetting the timing of the legs slightly... avoiding twinning.  If this still doesn't make sense imagine your ankles are tied together.  You can definitely jump higher if you jump using both legs at the same time (twinned), rather than jumping with your right leg a few frames before your left leg (un-twinned or time offset), but now try bounding with your legs tied together... pretty tiring right?  You'd probably last longer if you used alternate legs one after the other (more commonly known as running).  For more than a few bounds, having those back legs in sync doesn't make sense, so the dog breaks out into a gallop, but only after it's got maximum acceleration by using both legs together (like a 'bunch' or 'bullet' sprint start).

The strangest part is yet to come though.  Have a look at that image sequence again.  Notice the front legs aren't bounding at all, but rather they're doing a very very fast walk.  At first glance it seems odd that to go as fast as it possibly can, the dog uses a gait which is usually used for very slow speeds!  Check out the posing on the front legs - those are pretty similar positions to the contacts and cross-overs you find on a regular walk, but in this case a stride measured from contact to contact only takes 4 frames, more than twice as fast as a regular dog walk cycle!

Switching back into my amateur biologist mode again (I've read Sir James Gray's Animal Locomotion and How Animals Move cover to cover and can recommend them both for anyone interested!) I believe this use of the walk gait on the front legs only is to stabilise the dog's chest.  While the back legs are generating a huge amount of power to move the dog forwards, something needs to be done to keep the dog's body up in the air.  Normally both pairs of legs do two jobs:
  1. move the body forwards
  2. keep the body up off the ground
So if the back legs are angling their force horizontally (ie. only moving the dog forwards), the front legs must be keeping the dog up in the air, otherwise Rufus would be face-planting the flower beds (!) rather than getting up to speed.  Take a look at how these forces work...

In a gallop the legs spend more time in the air than on the ground.  In a slow walk the reverse is true - the legs spend much more time on the ground than in the air.  However the amount of time the legs spend on the ground or in the air isn't necessarily related to the speed of the gait, but rather to the amount of vertical support needed.  Think of the slowest walk - a horse in a field eating grass has four legs on the ground and intermittently to take a step it will pick up one leg for a short while, move it forward a small amount and put it down again.  In a very slow walk there are never less than three legs on the ground, and the leg which is in the air, isn't up off the ground for very long.  A horse doesn't move very fast when its eating grass, and it doesn't leave its feet hanging round in the air because that's tiring, and legs dangling in the air don't help to support the horse's body weight.  The strange 'fast walk' that the dog's front legs do in the sequence above is to support the weight of the dog's chest.  Notice how in all the shots above there is at least one front foot on the ground and sometimes both front feet are down together!

Look again at the forces diagram above.  Note how the arrow which makes greater forward progress is lower to the ground.  Think of this as being more streamlined.  The more effort the dog is putting into accelerating, the less effort goes into bouncing up and down.  Let's test this theory:  is the opposite true?  If a dog is slowing down it needs to get rid of horizontal energy.  One way it can do this is to dissipate it as vertical energy.  Read the below image strip from left to right - Rufus catches the tennis ball then as he suddenly decelerates he explodes upwards, changing that fast forward motion into up and down vertical motion.  Then gravity and cushioned bounces (Rufus' shock absorbing legs) helps us get rid of the up and down energy in the same way a bouncing ball bounces less and less on each bounce.

click on the image to enlarge

Looking at these images has just made me realise something.  Have a look at the image below.  Who's accelerating and who's decelerating.  Who has a horizontal motion, and who has a vertical motion:

sprinters accelerating and decelerating

Or how about this one:

looks familiar?

So in summary here's the main positions roughed out.  Remember that the chest is generally more stable than the hips (which tend to jog around a bit at the back... think of how "Slinky's" hips move in Toy Story, always left behind and catching up again).

some rough sketched poses

If you wanted to study the clip the first set of images came from, here it is below on youtube.  You might want to use saveyoutube or similar to download the clip to play frame by frame in QuickTime.

The combination of a bounding gait (ie. a twinned or symmetrical gallop) on the rear legs, with an un-twinned gallop on the front legs is technically known as half-bound.  Here's a cat doing something between a half-bound and a gallop from the Muybridge Animals in Motion books.  I've annotated the "1-2" landing of the front legs and the simultaneous contact for the rear legs:

And as always, please fire off in the comments if you don't get something or think I've got something wrong. It's all open to discussion!

See Also:
How to Key a Dog's Walk
Slow Four Legged Walks

Sunday, 4 September 2011

How to Key a Dog's Walk - Animation Notes / Tutorial

Animating on my current short had been great fun until I hit a tiny 2 second segment of shot 24 which seemed to be so very much harder than the rest of the film.  I'd animated 'Blake' (the border collie dog in 'Crows') cantering and galloping without too much difficulty, but I was really struggling with the walk.  This was a huge surprise to me.  I thought that if I could cope with the faster gaits and various varieties of them then I would be fine with a simple walk.  I'd managed rotatory and transverse gallops in rear lead (rear legs landing first, suspension with the body folded, as a horse would normally gallop), front lead (front legs landing first, body outstretched during the suspension, as a dog normally gallops) and double suspension forms (as in a very fast gallop), and right and left lead canters, as well as flying changes, transitions and so on, all without any trouble.

I had some difficulty in timing the speed of the paces in a walk and working out how many frames both front feet were down together, and how many frames both rear feet were down together, my investigation of which is posted here.  It turns out however that this was not the main problem with my walk.  I went right back to my keys and found I had something like this (and mirror versions of the below), which appears in countless animation books:

Fig 1. The Wrong Way?  Is the XO a Key?

While this isn't strictly wrong, its very missing something.  These keys are not extremes of height for either the front shoulders or the back hips.  While the contact position is lower than the cross-over, the contact position is not the lowest position, and the cross-over is not the highest position.  It sounds obvious, but when presented with the above sketch its easy to forget that the high position (after the XO and before the C) is actually the highest, and the low position (after the C and before the XO) is the lowest... just like in a human walk!  If you're unsure on two legged walks I'd recommend reading up on them in The Animator's Survival Kit, which goes into a lot of detail explaining how key animation principles apply to the human walk.

In addition to the problem of not having any vertical extremes, you also have 2 keys per stride rather than the usual 1 key per stride common for a two-legged walk.  If you're like me, you prefer to plan out your two-legged walks stride by stride (contact by contact) rather than the C-XO-C-XO etc sequence which you have to use with the above keys.  In my brief experience it is much less hassle in the long run to plan your 4 legged walks in the same way as your two legged walks - with one key per stride, simply only keying the front legs:

Fig 2. One Key per Stride - The Right Way?

For the time being lock the back legs out (if you have IK/FK switching then you can just flick them onto FK and ignore them for a bit).  Also check that your joint/bone hierarchy propagates out from above the front legs rather than above the back legs ie. your root-most body joint is facing from the front shoulders to the rear hips.  Even if you choose not to work with one pair of legs, read on - there's important info on phasing.

Fig 3. Where to Put the Body Root and How to Animate It.

Now we've simplified the problem.  You can use your front feet on IK, and the body (plus locked out back legs) to animate a normal two legged walk, a bit like this:

Fig 4. Looks Familiar?

So now put in the cross over between the two contacts, building on the interpolation the computer has given us rather than starting from scratch.  Remember the cross-over is slightly higher than the contact, but still not as high as the high position (obviously).  Because its a dog walk there's some really nice breaking of the joints going on at the cross over which you don't see so much in a human foot.  Study this image taken from Dr Stuart Sumida's 2006 lecture on quadruped animation:

Fig 5. Taken from here.

Dr Sumida is simply the coolest paleontologist around and I strongly suggest you check out his work, and the very weird photo of him sitting in a bathtub in the middle of the desert.

Once you've got the contact and cross-over working as you would have them in a two-legged walk, go on and add the high and the low position.  Watch some reference videos on you tube, or see my other dog walk post for a trace-over of the skeleton.  The dog's foot swings through in the same way as a human foot swings from the low position to the high.

Fig 6. Imagine this is a Dog.

The shoulders also need to drive (lead) the movement.  Remember you're meant to be working with forces not just poses.  Check that the shoulders are ahead timing wise (you don't need to shuffle keys, just get the breakdown right).  When you're keying the C-L-XO-H-C series remember that you should be working with the body and both feet.  The ground-bound foot does a really nice successive breaking of the joints sequence as it moves from the front contact to the rear just before the lift off.

Fig 7. Successive Breaking of the Joints and Forces.

As you might have guessed from my set of keys above the head follows through vertically (unless the dog is actively moving its head around).  I've offset this follow through to be completely out of phase, but on slower walks you'll probably find the phase difference is a lot less in terms of phase difference, but similarly lagging when counted in frames.  Obviously if the dog's head is heavier then the follow through on the head will be later (greater phase offset) due to the increased inertia of the head, and if the dog's head is really light the follow through will only be very slightly behind the motion of the body.  The head moves up and down as well as having a forward/backward tilt (which is simply drag).

Fig 8. Approximate Phase Differences in a Border Collie's Walk.

So in a slowish walk the rear legs will be contacting as the front legs are crossing over.  The rear legs are actually ahead of the front legs in terms of phase difference.  So if the rear legs are at the high position with the right foot in the air, then the front legs will be two positions behind at the low with the left foot in the air.

As the dog speeds up into a trot the back legs increase this difference even more so that the contact for the rear legs occurs at the same time as the contact for the front legs, but on different sides - ie. so rear right and front left contact together and vice versa.  This is now four positions head in the chart below.  The rear legs and the tail work in the same way that the front legs and the head worked respectively:

Fig 9. Animating the Back Legs and Tail.

You'll need to study some dog anatomy before you can apply the human walk to a dog's rear legs, but again the same principles apply with successive breaking of the joints (the foot seems to pad down as if its a soft cushion at the contact, then peel off the floor as if its slimy goo just before the lift off).

Take a look at the second pose in the above image.  Two feet on the same side (the dog's left) are airborne at the same time.  The dog is very unstable laterally (left to right).  To make up for this, it happens when the two ground-bound feet are furthest apart and the dog is longitudinally at its most stable.  Remember this simple rule and it will help you make sure you keep the rear legs ahead of the front legs rather than the wrong way around.

Fig 10. The Window of Stability.

Look for the 'window of stability' - when the two airborne legs are on the same side (in this image both on the right), the body weight is longitudinally centred over the front leg (which is forward compared to the body) and the back leg (which is backward compared to the body).

And in summary this simple table will hopefully help you understand the phasing of the front and rear legs:

Front Phase Front AirborneRear Phase Rear Airborne
C [R forward] XO L
XO L C [L forward]
C [L forward] XO R
XO R C [R forward]
C [R forward] XO L
... ... ... ...

In case you hadn't guessed: in the 'phase' column C stands for contact, L for low, XO for cross-over and H for high.  In the 'airborne' column, R indicates the right leg is airborne, L indicates the left leg is airborne, and in the case that neither leg is airborne I've simply indicated which leg is furthest forward in the direction of travel.

See how what's happening to the front legs is happening two keys later than it happens to the rear legs.  For example find a left forward contact for the rear legs on the table above, then search two rows down and you'll see that the front legs have a left forward contact, two keys later!

I haven't talked much about the twisting that's going on in the shoulders and hips, or the tail wag and head nod from right to left because we've only been working in side view.  To give you a clue you need to be looking for a snaking or waving motion running through from the dog's head to its tail when you view it from top down.  Overall its the same old rules as in a two legged walk, the hip turns to favour the leg which is furthest in front.  In addition the head follows through right to left (and with side to side rotation) in much the same way that it follows through vertically.

I hope you enjoyed this extensive dog walking animation tutorial and can now animate a dog like a ninja, or even better, like a dog.  Please feel free to comment below if you have any ideas, requests or suggestions!

See Also:
From Standstill to a Gallop
Slow Four Legged Walks

Saturday, 3 September 2011

Working with the Composer

I spent yesterday afternoon working with James Whittle, the composer for my 'Crows' short.  We got a rough recording finished which I'm really pleased with.  Hopefully sometime next week we'll have access to a proper studio to record the score properly!  After a long break away from the film its really great to be making progress again.

Friday, 29 July 2011

Illustration from Lanhydrock

I found this illustration hanging on the wall in Lanhydrock last week when I was on holiday down in Devon and Cornwall.  It was behind glass hence the slightly fuzzy reproduction - I've cleaned up the glare and reflections as best I can in Photoshop.  Twinning never looked so good.  It's nice to know that those 12 principles of animation are there to be broken, sometimes superbly.  I'd love to know who it's by and how old it is if anyone has any ideas.

Saturday, 16 July 2011

Thank you to 'Anim Dailies'

Just wanted to say a big thank you to Anim Dailies, who kindly featured my work on their site and gave me some really helpful feedback on a shot I've been struggling with.  The comments will be of great use when I return from holiday to work on the shot again in a few days time.

In the meantime you should all go and submit your shots to Anim Dailies for a focused critique and a chance to be featured on their site.  There are some great animators over there offering free advice - its definitely a service worth making use of!

Thursday, 14 July 2011

Slow Dog Walk - Animation Reference

I was struggling with a dog walk in my short so I thought it best to try to find some live action reference and rotoscope it.  When a dog walks slower its feet spend much more time on the ground than in the air, so the contact pose is not the same as the lift off pose as it might be in say a fast 9-frame stride cycle where both front feet might only be on the ground together for just one frame.

On a 15-frames/stride cycle it seems that both front feet are on the ground together for 3-5 frames, and both back feet are on the ground together for 5-7 frames.  In this way the feet spend roughly the same amount of time in the air if the dog is walking fast or slow (around 5-7 frames in the air).  When a walk slows up, the dog takes smaller steps and spends more time with its feet on the ground.  Hence...

F per stride F on the ground  F in the air  F both feet on the ground 
9 ~4-5 ~4-5 ~1
12 ~6-8 ~5 ~2 (front) ~3 (back)
15 ~8-10 ~5-7 3-5 (front) 5-7 (back)

If you leave the feet hanging round in the air too long the walk starts to look stiff, like the dog is stuffed and artificially held in pose.  Remember the dog 'weighs' something and it needs to drop that leg to support itself on.

Here's some rotoscoping I did on a 15 frames/stride cycle.  I then adapted the the walk to a 17 frames/stride cycle which divides up much easier for 2D animators who don't have the computer to in-between on 5ths and 3rds.

Wednesday, 13 July 2011

Animation Dailies - an open group

I've created an open animation dailies group / mailing list for any character animators out there (especially students working alone) who want to share work in progress on personal projects.  We'll see how this turns out!  2D, 3D and stop motion are all welcome!

Subscribe to animation-dailies

Tuesday, 12 July 2011

Dailies and Old Work

To keep myself motivated I'm hosting dailies over on blenderartists.  Join in if you like, criticism is always welcome!

Secondly, its always really nice when someone posts some of your work, even if its an old piece that you're not particularly proud of any more, its good to hear that people are still enjoying watching it.  Cycling blog Two on Four Wheels featured one of my pieces from last year, Au Soleil.

Thursday, 7 July 2011

Copying Poses with Constraints in Blender

Often I find that I need to use the copy and paste pose buttons to get a smooth transition when animating a ChildOf constraint.  Blender's built in copy and paste pose buttons ignore constraints so I coded my own.  Basically select the bones you want, run the 'copy pose to buffer' script, then change frame to after your constraint influence has been changed and run the 'paste pose from buffer' script.  This way you can get nice smooth transitions across those constraint changes...

Monday, 4 July 2011

Last Bit of Modelling

Well that's it all done.  Every set modelled, and finally Jonny's 'leaving home' rucksack finished and weight painted.  Click to enlarge to see the topology.  Just animation, rendering and compositing left to go.  After nearly 3 hours of careful weighting, I'm going for a run.  Then I'm going to tidy up the rig, make a quick low poly version, then get back to animating the last twenty or so shots... two weeks to go!

Friday, 1 July 2011


Just to let you know, the slight service intermission was caused by me being busy recalibrating at Glastonbury Festival.  I went there with Simon, working for Oxfam, which is definitely something you should look into if you like festivals, are skint, or want good karma.

I'm back at the grindstone now.  I spent yesterday trying to make fixes to the first 22 shots which I completed for my college final submission.  I got great feedback from fellow tube intern Henri Hebeisen (of glowing robot lights and mushroomer fame).  Today I've been prepping backgrounds and modelling the last 13 sets.  Here's a screenshot of working up the alpha channel for a camera-mapped shot, which I thought looked quite nice.

Sunday, 19 June 2011

Meet The Composer: James Whittle

I'm absolutely thrilled to announce that James Whittle is producing the soundtrack for my current short, due to be completed by the end of next month.  James is an incredibly talented composer, instrumentalist and singer, and has just finished his degree in Music at the University of York.  James has worked on some fantastic projects and its an honour to have him on board, so draw the curtains, turn the lights down low, put on your best headphones and take a listen to one of his recent works, 'Violent Cannon'.

VIOLENT CANNON by James Whittle

Wednesday, 15 June 2011

Things That Go Bump In The Shed

Thank god for Bullet Rigid Body Sim, saving animators' time.  Only 38 crashes and two corrupted files (thankfully I had backups) later and I've got a second and a half of falling spades to show for it (I've hidden all the characters by the way, so as not to spoil any surprises).  Sometimes I think there are demons in my computer.

Those warped meshes on the right aren't me forgetting how to model, they're just shadow catchers for the comp over the matte painted backgrounds which should be pretty funky.

Big thanks to this post on blenderunderground which reminded me how to use Bullet - its been a long time.  Looking forward to a merge of Aligorith's Bullet GSOC soon.

Edit: on second thoughts, this had my little brother and I in stitches... we're easily amused.

Tuesday, 14 June 2011


Jut to say that I'm following Bassam's example over at the tube project and now using blender-aid to keep track of which files are linked to which in this ever-sprawling project!  Above is 'Blake' the dog's list of dependencies. Thanks Jeroen!

Monday, 13 June 2011

Freestyle Rendering

Got a shock to find this one staring back at me today.  Some wrongly connected nodes and textures giving a freaky render.

Here's a preview of the node graph I'm using to mix up the AO, Shadow, Flat Color and Freestyle Passes, otherwise known as the 'spaghetti system'.  It masks the freestyle pass using objects from the sets, edge detects to give a line round the shadows to match the Freestyle line, and adds a slight transparency to the characters when comping them over the backgrounds.  The backgrounds get a slight blurred underneath and just around where the characters are to help the characters stand out a bit.  I'm also mixing in ink and paper noise as freestyle's textured stroke rendering isn't supported yet in blender as far as I can tell.

I'm working with three Freestyle shaders (and a common functions module) which I coded based heavily on T.K.'s fantastic examples in the blenderartists thread.

I'll post the source when I've got them finalised.

Back to animating.