After a weekend of working on the animation, I'm much happier than I was last Friday. I decided to make it easier to treat my keys, by using character subsets. I created character subsets for the knights arms and spine. This made it easier to look at the controls for these in the dopesheet. I was able to use the dopesheet to remove unwanted keys that had been put there whilst blocking out, and highlight points at which the animation was going out of control. I could then find this in the graph editor and tweak the movements appropriately. Once I was happy I was able to offset the keys of the spine controls and arm controls to create overlapping animation. Its not quite there yet, but its worked up enough to continue with the next stages. I spent yesterday and today creating the facial expressions for both Knight and Horse for the first parts of the animation, I found I could make my knight look like an angry Steven Segal by increasing the negative value of his surprised blendshape! It make him a lot more angry and nasty looking than I had previously allowed which is all good. Today I also made a start on the soundtrack. I've trawled the archive here at CSM and found quite a few suitable sounds, and Dan Hazlehurst AKA Midimanvandann (www.myspace.com/midimanvandann) finally got back to me and said yes to using his tunes! He'd been away for a week. So cheers Dan. Right now I'm tackling the laborious task of keying the rein controls. I still haven't thought of a nifty way of doing this quicker, so am keying the knight holding the rein frame by frame. Over the weekend I did a few experiments using the soft and rigid body dynamics tools with constraints. This may be an option for making the reins dangle realistically but the fact is the underlying keys stil need to be there for soft body deformation goals. You can set a soft body (ie rope) to be affected by gravity, and use goals (ie the underlying ik spline) as a second force. Using spring constraints in the soft body can also make the mesh stretch. I don't know yet if any of this will be useful but I'll continue to experiment. Anything to not have to key all of it frame by frame! As well as this, dynamics don't work well with skin objects, they use the original undeformed mesh to calculate the bounding box, so even if it is possible to make the reins move realistically, intersection issues might let me down. Time will tell.