Category: Journal

Rotoscope and Humility

Last week, without fanfare or explanation, I had a shot taken away from me and given to a more senior artist, and in one day he made the shot look finished when I had been struggling for a week to make it look presentable. This upset me, because my pride was wounded and my vanity shown up, but also because there wasn’t any warning, any notes that the shot wasn’t up to snuff. I mean, I knew it wasn’t looking right, but nobody said anything and I was scared by the thought that I wouldn’t have warning when I was failing,  that nobody would be catching me here, and that I might not be good enough. This was a frightening realization.

And the next task didn’t make me feel any better: roto mattes for moving foliage and foreground characters. In a grainy night-for-night shot. As a long-term career choice, becoming the best matte puller in the shop is not where I want to put my focus. Rotoscoping mattes (and painting plate fixes) is the sort of high-skill, low-creativity labor that won’t be done by high-paid domestic labor for long. It’s too easy to outsource and software may even make it trivial before long. But it is something that a VFX artist needs to know how to do competently.

So I had work for my hands while my mind reflected. Rotoscope can be a meditative process, especially when lots of hand-roto is required. Something in the nature of this straight-forward work, where the goal is clear and progress can be marked, is good for the lost spirit. I worked through it fairly fast and focused, too. Even though I’d rather be lighting and comping shots, I think I needed the humility of this task to remind me not to be too prideful, and its utility to the larger project to feel that I was contributing. Finishing it quickly and well helped me regain confidence. At the end I was ready to accept the situation and work with it.

So no-one is holding my hand here (or indeed anywhere in life), and I will need to take the knowledge that I need. I looked over what the other artist had done to see what he had done that made the shot look so much better. Different choices in lighting, changes to the texture artist’s choices, and lots and lots of real-world elements in the comp, for starters. An understanding of the Nuke lens flare tool. In general, he uses an improvisational approach that seems to be hunting out the better image, as opposed to my head-directed attempt to fulfill what I think should be the right approach. I think it’s a good example of letting the hands do the work instead of listening to the head. That, and quite a lot of knowledge and skill built over years and years of finishing shots. There is a wealth of knowledge to be learned here, if I can keep my eyes open.


Two days after installing RescueTime, I’ve already noticed a difference. Today, aware that I would have to answer to an objective record of where my attention went,  I worked really hard at maintaining focus. My brain is exhausted!

Two things I did: while waiting for viewport renders or other quick processes, I kept my eyes on the render and my mind on the problem. Normally I might pop over to Firefox to read something for twenty or thirty seconds. What a way to lose flow. Second, long renders that hold up progress on a shot are usually a perfect excuse to hit the web or go for a wander. The lighting TD’s pal. Today when that happened I switched to a second project, one that was easy to pick up and set down where I could make concrete progress in small steps. I feel like I used my time pretty well. I can definitely feel the difference; I’m considerably more tired and my brain feels worn out.

I have a long way to go, though: I hit my slack limit (goal for time waste) at 2pm. First goal: don’t hit the slack limit before 5. Then I’ll work on shortening the slack limit.

Some readers might think I must be some kind of oaf, crowing about how I didn’t dick around all day, but what can I say? I’ve acquired some unproductive habits over the years (to be honest, I was born lazy), and now I’m going to shake them. After, of course, a relaxing weekend…


Man, my body is a delicately tuned machine. Forgot my allergy meds this morning and spent the whole day with yuck belly. The yuck belly makes me feel hungry which led me to eating two Starburst candies I found in my jacket. Mistake. Even that small amount of sugar was enough to get me on the blood sugar rollercoaster, sending me into a crash right before lunch. I managed ok but felt uneven all afternoon. The blood sugar affects my focus strongly, so lately I’ve tried to avoid sweets at work. A birthday cake? Forget it, I won’t be getting much done that afternoon. Hard to decline, especially if it’s those little chocolate cupcakes from Hey Cupcake. But I really try to hit the low-sugar snacks like pretzels or bananas. Honestly, I do.


Not a bad day, productivity-wise, although I let myself get distracted too much late in the afternoon. I let myself get distracted from blowing a man’s head off with a shotgun! Shameful. Sorry,guy, we’ll finish you off tomorrow, I promise.

A good day

Just a nice  productive day. That’s all.

OK, not 100% show-based productive. I discovered, a really exciting weblog that focuses on non-photoreal rendering, VFX production, and dinosaurs, so I had to explore that a bit. Lots of ideas and connections to a world of NPR I assumed existed but didn’t know about. Huzzah.

The Infovore’s Lament

I love to read. It might even be more accurate to say I’m addicted to reading. I can’t help it. I see words, I read. I will read just about any article to the end. And a bad habit I’ve developed is that I will read to avoid thinking or working. I’ll happily devour news or polemics or tech gossip in order to avoid mental labor. This is a habit I am working on unlearning.

First step: being faster about closing browser windows. I’m trying to bias myself toward NOT reading things. I try to remember to think, “Do I need to know this, right now?” and unless the answer is unequivocally yes, I click close. I’ve not regretted it a single time.

Not reading ties into my larger goal of maintaining focus, short term and long term. I am letting go the need to know and do everything and it’s actually quite a relief. I am slowly making peace with and embracing the fact that I am finite, that the set of things I might like to know about is boundless but making the most of my finite self requires that I focus my time and attention.


Today: tuning various layers on the logo animation. Found a bug in XSI2010/mr wherein it ignores transparency maps on objects appearing in reflections. Seems to be a problem with all material types. Banged on it for a while, decided to fix it in comp.


Lighting logo type again, trying to tune the fineness of a displacement mesh to get the best tradeoff for detail over speed. You can definitely see the difference, so I guess it’s worth the render cost. :\

Render, review, render, review.  Tuning this shadow animation is time-spendy. And, wow, it turns out mental ray renderse ray-traced shadows very differently (in certain situations) if you set the shadow method to ’segmented’ versus ’sorted’ or’enabled’. I wonder what’s going on there? I’ll have to investigate the difference between segmented and sorted shadows.

Weekend work

I moved the ball down the field in my Variations project this weekend, doing a few short work sessions. For now, the half-hour limit is working out well as I remain energized afterward, I know what steps to take next, and it’s not intimidating to start each session. Another advantage is that I have plenty of time to contemplate the next step and problem solve in the interval. That’s not a luxury you get in a full workday.

Another thing I discovered is that tutorial movies, which I normally have a hard time sitting through, are somewhat easier to take watching in bed on the ipod. It’s comfortable, and more important there are none of the distractions I find on the desktop. Small screen’s a downside. Maybe watching a short tutorial video before bed would be a good way to wrap up the day and get through more tutorials.

UVs and NPR

Researching the best free UV solutions, and spent a little bit of time with Maya 2009’s Unfold tool. I can’t tell yet if it’s as powerful as the purpose-built tools like Headus, but it’s a start and it’s built in! The search for a fast fast fast UV solution continues.

Meanwhile, here’s an entertaining bit of non-photorealistic rendering (mixed with IBL, interestingly) that’s pretty entertaining for a student project. I don’t think the style is 100% successful, especially the lack of consistent contact shadows on the ground, but I like how adventurous it is all the same.

Elk Hair Caddis from peter smith on Vimeo.

Powered by WordPress. Theme: Motion by 85ideas.