in-Betweens

One thing that I love about my job is there’s always something new to learn. I’ve been proficient in Maya for some 10 plus years and have been animating for many of those years. When you work in an industry with directors, producers and others whole have a hard time fully understanding what a shot will look like while in a blocking phase, re-timing a shot can become quite a burden.

I previously wrote about a great tool that our company’s lead TD created called Timewarper. Using this tool has been a time saver and best of all it’s non-destructive, that is until your have to merge to curve. Since the Timewarper tool connects to the input time on each object or controller, the only way to get a real view of your keyframes back is to bake everything down. While fine with pure mocap data, it becomes overkill when you have hand keyed data or layers.

A fellow colleague of mine has always used breakdown keys in his animations and it used to frustrate me because I would always inadvertently mess up an animation that he did if the breakdown keys weren’t converted to regular keys before hand. It’s always been a running joke about him using them, and me not understanding their usefulness. The issue was because I always did my time scaling in the dope sheet editor (this was also a running joke among our group because no one used the dope sheet except for me). After using the Timewarper tool, we once again had a discussion about scaling keys, breakdown keys and timing shots. It finally dawned on me the usefulness of such keys.

Simply put, breakdown keys scale evenly between two normal keys. What does this mean in a practical sense though? Say you have a long animation with dense keyed data. You have a section right in the middle where things need to either be slower or faster, but all timing after and before those keys needs to stay the same. By converting all the keys between your two anchor keys to in-between keys, you will be able to re-time that section without having to slide the chunk of keys following the ones you wish to scale first, then guessing how much you want the section to be scaled by. My workflow in the dope sheet editor would be this:

1) locate the keyframe to be my anchored key, then select it and everything after
2) move those keys an arbitrary distance so that I have enough room for the next operation
3) select the keys I wish to re-time and scale them accordingly

The biggest problem with this technique is you must keep track of the distance from your anchor key and the last scaled key. Usually this would be solved by setting a key one frame before the anchor key. The other issue is that it makes iteration on the timing of the section bit difficult since each time you wish to re-time the section, you have to make sure the selection highlighted and remember where your anchor keys are.

Using breakdown keys, you can effectively re-time the section over and over until you have reached the desired speed. After you’re done you can convert the breakdown keys or stamp new ones at whole frame numbers either by hand or by baking the section. The bonus is that you can quickly check your timing from the original source by using a buffer curve and swapping your curve back and forth.

The one thing I’ve learned over my career in animation is that time is finicky. This isn’t to say that I have a hard time with timing an animation. But in the industry, there’s always something you can count on and it’s that the director, producer etc. will almost always want a shot re-timed. The change usually isn’t something as simple as “make it 30 percent slower”, but rather on the heels of making a fraction of the shot slower while speeding up another portion.

Being in the industry before the layering system with Maya was ever implemented, there were times when repositioning and retiming a shot meant shifting and modifying groups of keys. Add in the fact that games use a healthy amount of motion capture data and it meant for very tedious work.

Fast forward to Maya’s latest versions and there are finally a new host of tools that borrow from it’s acquired brethren like Motion Builder. Though still not as robust as MB’s scene assembly and layer blending, adding the functionality described above in a limited form breathed new life into an otherwise aging animation system. With Maya’s layering system we are now able to manipulate massive key positions while still maintaining their timing and original “feel”.

Yet the problem still remains with retiming. Even if you’re using layers to modify our scene, you still need to shift all keys in the base layer and subsequent layers to re-time your shot. Add in the fact that you may be dealing with baked key data and that makes for tedious work once again.

Maya 2011 introduced a new piece of functionality previously offered in it’s failed “trax view editor”. This concept was time warping and it allowed you a seventh type of animation curve (3 translate and 3 rotational curves). This new curve can handle the time values of a shot and allow the animator an easy way to re-time without the need to scale and shift keys by hand. Unfortunately for those of use NOT using the latest version of Maya, it would seem that your only option would be to assemble the animation using Trax.

A brilliant tech artist I work with named Eric Pavey informed my animation team that Maya has always had attributes to edit time with a curve, they were just never hooked up. A couple of days later and the “Timewarper” tool was online offering the ability to shift and re-time our shots easier than before.

I’ll say that’s it’s not all roses since working with the keyed data either requires a healthy amount of layered animation or that you bake your keys to transfer the modified time of the shot. The tool can be used in a variety of ways to even pre-visualize the timing adjustments leaving you with a blueprint on what keys to shift and how much. Since the time curve can be turned on and off easily, it’s non-destructive.

I will caution that this tool could easily be used as a crutch, and like with any tool that tries makes the core foundation of animation easier (i.e. full body IK reach, motion capture, simulations) it’s also a good way to of making a badly timed shot look horrible. Like what 3D animation did for 2D, It’s intention is to supplement your existing abilities and make mundane and tedious tasks a bit easier to swallow.

All I have to say is… unbelievable! I thought for sure that the acquisition of Palm by HP was a clear signal to the design of WebOs, the Pre and everything in-between. But against all odds HP managed to not only breath life into the now defunct Palm brand, but also raises the bar for Apple.

The TouchPad tablet is an iPad in every way except that it runs Flash and has a whole host of integration with it’s sister hardware, the Veer and Pre3. After viewing all the videos of the Veer multitasking I have only one thing to say to HP… Bravo!

I love the PS3. Despite the fact that as a developer it’s a difficult platform to develop for, I find that the games I’m most interested in consistently are either exclusive titles or are also made for box XBox and PS3. Yet for all that Sony seems to be on the ball with their developers and titles, I have to question their ineptness at announcing a product almost a YEAR ahead of the actual launch date!

I get that they would rather not leave the developers in the dark and want as many launch titles confirmed as they can, but I’m not sure it’s the right move. The specs might seem impressive right now, but as with any technology, the landscape is ever changing. What is cool now may not be cool down the road or may be used in a radically different way.

As a developer I can say that having a clear roadmap to create your next title against is wonderful. But I fear releasing this hardware so publicly gives the competition more than ample time to get their ducks in a row.

So many may know that I worked on the original Dead Space title while I was at EA, before I joined Sledgehammer Games and Activision. Some may not know that I and a good portion of the team that formed SHG actually worked on a good chunk of Dead Space 2 before leaving. Now that the game has finally been released I got a chance to see exactly how much was changed when I left.

I worked on the revamped Zero G mechanic that let’s Isaac fly through Zero G rather than jumping from surface to surface. While the end result is certainly a good idea, a better one would have been what I had originally worked on, which enabled you to boost and break in mid-flight. While I’m touched that the original assets were at least kept in some form, I fail to see the reason for the cut. Maybe when I finish the game I’ll see the reason.

In a few months I’ll be sure to post a new reel with some of the work that was done while I was still employed, as well as my impressions of the game and where the series might be going. It is, after all, a 90 rated game so far!

You know I work for Activision, you know I’m working for Sledgehammer… well it should come as no surprise then, that I’m rooting for Black Ops (part of the Call of Duty franchise) to do very well. I’ve played a lot of the game and believe me, it’s worth picking up. Though Treyarch has not been very well known for their game changing work in this arena, I’ve witness some amazing amazing things of late from them.

On the animation side of things, there’s not much to talk about. They use Motion Builder for their animation pipeline, do a LOT of motion capture work, but also have some interesting ways of comping their shots. Aside from that, their story is a bit unique for the Call of Duty universe and I’ve very interested to see what happens next.

Well at least for me things are coming to an end very quickly. I’ve been looking at pictures of babies, asking friends who have kids for advice, all because my wife and I have a little bundle if joy coming into this world. More on this soon!

In other news, my company is hard at work on our game and I simply can’t wait for the announcement. Rest assured I’ve been having a blast animating for this project and every shot I have worked on has turned out so much better than Dead Space. This is due to a couple of things. Most notable is the workhorse of machines we have at our disposal. Working for EA meant a lot of handed down tech and sub-par computers. Sledgehammer Games has all the latest equipment, crisp, large monitors and the tech for our game has some of the most innovative tools I’ve seen in the industry. In months after the game has shipped, I’ll be posting shared animation knowledge as well as case studies on how shots were produced. Till then, keep checking back for more updates!!

If one thing bodes well for me, it’s that the Internet is always there. I apologize to all readers out there for my lack of updates for the better part of 3 months. Of late I’ve been working on the newest project my company, Sledgehammer Games, is producing. We’ve released our latest website filled with job postings, pictures and info on everything in and around the office. Check it out at www.sledgehammergames.com
I’ll posted more about the ongoing project I’m involved in as soon as I can.

Heading out to the Game Developers Conference this past weekend was filled with so much useful information. I attended a talk from Naughty Dog’s lead programmer, Travis Mcintosh.

The game was basically designed around getting very lifelike animations in game with tons of variety, all while trying to maintain a runtime animation budget. A lot of information was learned from this talk and it was perhaps the most informative I’ve ever been to at a GDC. Props to the Naughty Dog team to a brilliant presentation!

Tons on new stuff in Maya 2011, animation and UI improvements are at the top of my list for sure!

http://bit.ly/aM3LGX