Well, it certainly has been a while, hasn’t it?
It is fair to say that the general lack of activity on this blog has been directly related to the lack of significant progress on Vespers. While much of that has been due to the inconsiderate intrusion of Real Life on the development process over the past, oh, 18 months or so, a fair share of the stagnation was also due to an unfortunate combination of bad luck, shortsightedness, and an incredibly exasperating game engine.
I’ll conveniently ignore the former stuff, and expand just a bit on the latter.
As I’ve discussed in earlier posts, most of what I’ve been focusing on since last fall (or so) has been the first cutscene, since the demo will be all of the action leading up to, and including, the first cutscene. Of course, in the original text version of Vespers there is no cutscene per se, but there is a short sequence of events that spans a few turns and which the player has no control over and cannot interrupt. So in a sense, it’s a scene playing out over three or four turns regardless of what the player tries to do, which is really not that different from a short cutscene. So that’s how I decided to implement it, despite all of the discussion on the interwebs over the past few years about how cutscenes are evil and kill puppies.
Rather than play a pre-rendered cutscene movie like some games, I wanted to have one rendered in-game, using the same character models used throughout the first Act, with some creative scene design, lighting, and camera manipulation thrown in. Not really a big problem conceptually, but reality doesn’t often play nice with concept, and this situation was no different.
Now, rather than bore you with the details of all that went wrong over the past six months or so, let’s just say that most of the issues could be traced back to three key things:
1. Some of the character models weren’t originally set up exactly as I needed them to be;
2. The animator used a separate copy of each character in order to make the cutscene animations; and
3. The Torque Game Engine is very fussy about a lot of things when it comes to playing animation sequences.
Basically, in Torque, you create a character model, rig it with a “skeleton,” and then create animation sequences for that skeleton. Different models can use the same skeleton, and any animation sequences developed for the skeleton will work with any model that uses that skeleton. The only requirement is that the skeleton must be constructed precisely the same way, down to the very last tiny little detail, for an animation sequence to work with it. Generally speaking, this is not a problem. But along came #2 above, and with it came some minor but undetected changes to skeletal structures that couldn’t possibly get past #3 above. Some of these were easily correctable, some not so much. Extra time, workarounds, and so on, you get the picture.
In the midst of this came the realization of #1 above, and the need to fix some of these seemingly minor but ultimately important model structure issues. But naturally, making changes to model and skeletal structures led me once again into the endless pit that is #3 above. So not only would I have to update and recreate the models and animations for the cutscene, but the same would apply to all of the animations for these characters for Act I. It became like a revolving door, round and round from the cutscene to Act I and back again.
With the additional constraints of time and life on top of everything else, it began to feel like I was on a Road to Nowhere at times.
Fortunately, I was able to find a few great people to help, people who know a lot more than I do about the baffling intersection between modeling, animation, and the Torque Engine. With a little persistence, a great deal of assistance, and a whole slew of questionable workarounds, I was finally able to push through it. And now we have a cutscene.
It still lacks a background musical score, which I’m in the process of working on, so overall it seems strangely quiet. But it’s all there, it’s all coded, and it runs smoothly from start to finish with all the right camera shots, pans, and sound effects. I’m quite relieved.
Now all I have to do is go back and re-export a whole mess of Act I animations…again.
Enjoyed this article? Subscribe to The Monk's Brew RSS feed.