Crunch Mope

This is one of those never-quite-finished entries that's been long-lingering due to lack of time and attention — in this case it's been months (there are some that are older... what can I say?) — the original save date was early 2005, and it's lingered in "draft mode" ever since.

P1000609.jpg

I was digging around on del.icio.us one afternoon and saw that after a lot of web traffic and game-industry furor back in November 2004, ea_spouse was still drawing hits from across the del.icio.us spectrum.

Now, I know a lot of folks at EA, I deal with them and know that a lot of them are happy, that they keep moving on from one project to the next, from group to group, production to production, and they're doing fine. If the general picture were really as bleak as ea_spouse paints, then I doubt very much that anyone would work there for more than a few months. And that's not the case — plenty of people at EA have been there, happy, for years.

Just the same, there can be problems in the industry. I regularly see people griping on message boards, usually about pay, hours, and credits. I read the recent article on IGDA purporting to be lessons for EA managers (and game managers in general), & I felt that the authors had gotten a few things wrong.

I don't think this is really an EA issue at its core — maybe "ea_spouse" hails from there, but EA is just a big, easy target for journalists. Rather, it's an industry-wide issue. There are many other companies, better and worse alike...

(Cartoon aside: VG Cats)

The games industry, spawned from the general computer graphics and animation industries, is by its nature dynamic and changeable. More than that — at least in its earlier days, it was mutable. It was really possible for a very small number of people — sometimes just one dedicated and brilliant person — to create things that would radically change the industry or create whole new genres. When Carmack & co made the Doom engine, that small group changed everything. When Blinn simplified the Cooke-Torrance illumination model, it changed everything. Toy Story? Photoshop? GeForce 256? Jurassic Park? Every one a work created by some small group of smart dynamos that blew-open the prevailing notions about what was possible.

(Folks like Clayton Christensen & Tom Peters might argue that it still is mutable — we just don't see how yet)

When you're hot on the trail of a Big Idea, time is immaterial. You do it out of love for craft, love of what you're doing, love of the idea that you're out to change the world. It's not a job, it's What You Do, it's a part of who you are and who you're becoming, what you're discovering and dreaming. That's true either at the office or out, depending on the person. For the people spending 50 hours a week after work carefully airbrushing the side of their van, for someone carefully tweaking their website scripts every other day, or someone sure that they can just get the rig on that character just so because it's never been done that way before.

The beauty of Big Ideas is that no one complains about them — those hundred-hour weeks spent on finding a way to get cool reflections aren't a waste, they're a part of life's flow. The nasty part is, sometimes those Big Ideas make money. Sometimes they make a lot of money. And then the trouble starts.

The business-statistics sources IGDA quotes aren't inherently bad, they're good sources. A lot of the same sources can be found in books like Ed Yourdon's Death March or even systems books like Deitrich Dorner's The Logic of Failure. Probably my favorite of all such resources is on the web: The Software Program Manager's Network is a great source on project management, risk assessment and management (and most of it has been already paid-for by our US tax dollars, so its resources are free to all).

The problem is that such sources don't see the surrounding business environment as highly dynamic — rather, they see business as stable, as a broad zero-sum or low-growth background. If your business is expanding at a rate of 50% per year and bad management is costing you 12% productivity per year, well, guess what? You made money anyway. And what's more, human nature being human nature, there's more than a slight chance that those bad managers will be convinced that the reason they made money was because of their own sage and masterful leadership.

In a slower, more-stable business environment with 5% or 10% growth (or less), that 12% drop will get a manager fired. In a boom market, it won't even be noticed. It may seem counter-intuitive, but successful innovation is an excellent breeding ground for bad managers. Good managers can exist there too — a good manager in a boom market will do better than a bad one, but the bad one has a good chance of survival — even encouragment — simply from the dumb luck of being in the right place at the right time. Anywhere else, they'd be dead.

Second, honey and money attract flies. When you look at the growth of many industries and niches, you'll see that once money appears, the companies will swell — not only with the core creators and producers of products, but with layers of bureacrats who really had little or no knowledge of the XYZ business before they realized that Forbes had identified it as growing. I've heard animation producers say they saw no difference between making films and making cardboard boxes. They were only there for the money, and had no interest in anything like Core Values or purpose.

So now imagine, the stage is set: the innovation of XYZ, based on a Big Idea that its creators were pursuing out of the love of doing it, has suddenly produced great results, and the money is flowing in. The structural layers of the companies are swelling with people who weren't really part of the XYZ innovation in the first place. The XYZ business is slow and project-oriented — you work for years and you either win big or you lose big (large projects are common in many innovation businesses: cel phones, gaming, biotech...). The people who lose big go away, while the people who win big — well, it must have been the great accounting and production liason lunches that made everything so successful, right? Because the cycle of the entire enterprise is such that, by the time a game or film is released and the money is being counted to determine whether the project was successful or a failure, the creators are no longer part of the process. Bean-counting, yes. But animation? Design? Modelling? Research? That was done months ago, maybe years. Those people are working on a completely different project, potentially at some completely different company (back in November 2004 it was claimed EA had something like a 50% turnover annually). In the meantime, the bean counters are patting each other's backs for work well-done.

(Illustration in point: some years ago I was working on a TV commercial and needed to check some client paperwork. I couldn't find my producer or his assistant. After a while I discovered that they were at an afternoon party, because another commercial I'd done some months earlier — modeled, textured, animated, lit, rendered, and delivered — had won an animation award. Despite my name on the award, the agency and producers were busy having a party without even bothering to tell me — eventually one of the agency guys thought it would be fun to come back to the production office and visit, which was the moment at which I discovered that the award even existed and had been won: when someone I'd never seen or heard of before came by to say thanks for all the great work "we'd done together." Two years later, I won the same annual award, and found out about it because a colleague had attended the ceremony the previous evening, where they saw an exec I'd never heard of accept the award in my name — even though I was working only a few blocks away, no one had bothered to tell me. I would attribute these sorts of stories (and I have plenty) to individual people, rather than a broad pattern — if it weren't for the fact that they happened with completely separate companies at opposite ends of the country)

So there is a confluence of forces, between two groups with very different expectations:

  • Producers who have seen the sales figures for Half Life or and plan to replicate them. They see that the people working on those innovative, personal projects were burning 80 and 100 hour weeks. The idea of doing this for love never enters a producer's head. What does enter their head is that a crew of 20 took two years of overtime without overtime pay... hmmm... 20peoplex80hoursx100weeks... and the crew's motivation isn't taken into account. They just think that this sort of work requires 80 hour weeks without considering what inspires an 80 hour week.
  • New artists/TDs/employees coming into the business have a completely different idea of what they're going into, compared to their counterparts of previous projects. They are expecting a "job" in the classic corporate sense — health and dental, ESPP, retirement plan, two weeks of vacation. They went to school and got trained in Maya and Softimage and C# and paid for the privelge of that education, but in the end those schools are training people for "jobs in the industry" which is usually a far cry from what the people who started the industry thought they were after.

(To be continued when I'm less cranky, heh)

Posted April 27, 2006 | Comments (0)

Should Have Ripped This Off Long Ago

P1010108.jpg

Yes, I should have ripped this off,... but it wasn't there to be ripped.

They say that creativity is the subtle theft of others' ideas.

Posted April 27, 2006 | Comments (0)

Smoked Honey

P1010150.jpg

So I'm in my family room, busy reviewing Sketch Proposals for Siggraph 2006, and hear a buzz at the window. "Dang it," I think, "more flies from a couple of days ago," when a door was left open for hours. I ignore the buzz, but when I take a review break I go to the window to find it's not a fly, but a bee.

With two other bees. Great.

I get the fly swatter, dispatch the bees, but as I'm putting the swatter away I see another bee on the living room picture window. I walk over there, and see he's got friends. Seven friends. Wak Wak wak with the swatter, but there's still a faint hum... I turn, lean over to hear it faintly coming from the living room fireplace just as another bee descends out of the chimney and into the room.

Oh boy! Bees nesting in the house! So I make the obvious anti-Santa move — start a fire in the fireplace. Smoke 'em out. A small fire, with the goal to make more smoke than flame. It seems to have worked — haven't seen a bee for the rest of the day. But I'm uncertain of just what will become of their nest. There was no smoke backup, so the chimney's not entirely clogged, but I expect there's something or another up there.

Anyone know a decent chimneysweep?

Posted April 26, 2006 | Comments (1)

More on Presentations, GDC, PDF, PPT

Garr's recent post on slideuments got me thinking about my current Powerpoint method.

I sometimes do fall into the "write the presentation on the plane while traveling to the conference" method, but after taking it in the face a few times I've tried to avoid it. It's worth thinking on your feet though — a couple of years back I traveled some 14 hours to give a presentation to a group, only to discover that just before we began they had "warmed up" by watching one of my presentations on video: the same one I was about to give. Fast backtrack! Happily I had colleagues and enough extra material to come up with a useful session on the spot (and spend part of the time provoking audience response and asking them a lot of questions).

Being the lazy sort, I use one set of slides for as long as possible while preparing a presentation, rather than separate stage and "slideument" versions — my current method is to make the slides and write all of my expected verbal presentation in the notes, jokes and all — I use the notes as my script for all rehearsals, and so after the presentation, a PDF of the PPT notes pages usually makes for a reasonably representative slideument.

The notes also contain extra info in the way of statistics, source code snips, and so forth.

In making the PDF, I often have to split-up duplicate copies of some slides to accomodate the narrative effects of animation: moving items, fade-ins (new items that cover-up previous items), etc. On occasion the spoken text is just too long for a single printed PDF page, those get split into duplicate slides too. So after the conference, there's a second version of the PPT marked "such_a_conference_for_web.ppt" and THAT gets used for the web PDF. The recent FX Composer 2.0 presentation is the latest example.

If a conference administration wants slides well in advance for printed conference proceedings, I comply, but sometimes just barely — a short "teaser" presentation with just a half-dozen slides or so containing key (and leading) points (and contact info). I don't want people to already have read what I'm going to say before I've had a chance to say it! So far no conference has complained about this "incompleteness" (though people did get an opportunity from GDC to update slides after the conference but before their final publication — the GDConf folks are savvy)

Rehearsal is key. In the documentary Comedian, I love how Jerry Seinfeld is rehearsing standup-routine lines all day, chattering them at people in elevators and so forth, working the lines into his daily routines. I wish I could be 2% as prepared. Every minute spent rehearsing always — always — pays off in comfort level when presenting. And ensures that the slideument is a valid representation of the actual delivery.

Some kinds of presentations, particularly super-tight performance-centric ones (e.g. Lessig style) seem to defy easy transition to a slideument. Is podcasting the best answer for these? I'm curious to try it.

Posted April 25, 2006 | Comments (0)

Hacking the LX1

P1000976.jpg

After some time using it daily, I can recognize my own way of working with the LX1, so it seems time to share some rambling notes. Operationally, electrically, and optically the camera is identical to the Leica D-Lux2 — these notes apply equally well to both cameras.

Let me start why saying why I bought the LX1. I knew that it would be slower to use than a DSLR, but I wanted a high-quality compact. A friend at work was raving about his DLux2, and I checked out the Panasonic alternative but wasn't feeling a need to buy anything at the time. A couple of weeks later, I saw a very cool camera at Fry's and realized that this was the same camera I'd been web-browsing. Besides the pleasant feel of the camera in my hand, it had an actually-wide wide-angle (28mm equivalent) and native 16::9 aspect ratio.

16::9 aspect exactly matches the format of newer widescreen movies and the Sony PSP, which I'd found to be an excellent venue for showing short video clips and lots of widescreen photos during the previous week's Game Developer's Conference (where I'd been toting my Canon SLR) (weird trivia — the camera itself is 16::9-shaped , which can be useful for thinking about composition even with the camera turned off (hold it up in front of you for a sec) — and the camera is almost exactly the same size as the PSP screen). I read all the reviews I could find, looked at photos made by it, and then bought it. I have been happy.

This is the first non-phone camera I've had without an optical viewfinder. I know that some folks have gone ahead and superglued a 28mm Voigtlander Minifinder to the top of their DLux2 — I understand it, that desire to raise the camera to your eye is a basic one. The camera is, after all, an extension of your eye...

Yet as much as I enjoy camera hacks, this seems overboard. I like the LCD, though it is not nearly as immediate as a eye-level finder. Responding quickly to what's in front of the camera can be tricky — it takes time for your eye to shift back and forth from the scene and the finder LCD, racking your internal focus back and forth as you do it. It can be a sort of barrier, and I'll mention how to help that further on.

LX1 Hack #1 - Tape Controls
I have gently hacked my own camera, using about $0.05 worth of gaffer tape (you knew I would get here eventually, right?). Gaffer tape can definitely help the LX1, even if you get the camera in the svelte and low-visibility pewter. The first illustration shows two key aspects of the camera that can be enhanced by a little of the old G.T.

At the back, there's a triangle of black tape placed overlapping and below the little "grip nubs" that Panasonic has already provided for your thumb. The fabric of the tape is much better to grip, and it doesn't shift around so much when you're also trying to use your thumb on the controls. It's an improvmenet for two-handed and one-handed operation — in fact the camera is almost too slippery for one-handed operation without this enhancement.

At the top, I've added a tiny strip of tape to the mode-select dial, laid in the blank space. It's directly opposite the "P," so when the camera is in "P" mode you can feel the line of tape squarely under your index finger. This makes it straightforward to set the camera not only to "P" mode but to any mode, once you're used to the order of the neighboring symbols. With the tape, modes can be selected purely by touch — no looking at the dial or the LCD. This is invaluable for use in quick-paced or darkened situations.

LX1 Hack #2 - Tape GripThe front of the camera can likewise use a little grip improvement — placing two narrow pieces of tape as shown here improves the purchase of your middle finger when gripping. The difference in finger-slippage is really dramatic, compared to the smooth plastic surface under the tape. Remember, the LX1 is a very light camera — it's easy for it to shift around in your hand without a good grip (unlike big metal 35mm's of old).

I got a great deal on a red Leica-badged Crumpler case, which is sturdy nylon/cordura and has a wee pouch in front for carrying an extra battery and/or SD card. It can be carried on a shoulder strap (it's almost too light for that), on your belt, or the camera strap can stick out and you can just carry it as a "loose" case. The Crumpler is similar to the Lowepro Ridge 30, which is somewhat generic.

That work colleague whose DLux2 helped inspire me to buy my LX1, has one of Luigi Crscenzi's luxurious LeicaTime cases, which like the Crumpler is a snug rectangle, but unlike the Crumpler is fashioned from heavy leather. Slightly less pragmatic-looking, lacking the same sorts of clips and mountaineering-inspired velcro bits, but beautifully made and the machine is well-protected.

Luigi also makes a "never-ready" style case, which you can see here — Panasonic likewise makes a slightly simpler one for about $30, along with their own pouch models.

I'm glad I got the Crumpler — of all the cases I've seen it seems the most durable and useful.

The LX1 has a lens cap. This can potentially make it less-than-pocketable, compared to cameras that fully retract their lenses. I've yet to have trouble with this. A danger, when pocketing the camera sans case, is that the cap will come loose in your pocket and the lens will be scratched. A solution I used before receiving my Crumpler — keep the pocketed camera in a plastic ziploc sandwich bag, which keeps everything snug and doesn't let fabric or loose pens catch on the cap.

The lens is the star of the camera, I'm really impressed with its lack of distortion and crisp results even wide-open. Pictures speak to its color and sharpness. It's only f/2.8 but the addition of "OIS" Image Stabilization gives it hand-holdability at pretty low speeds.

I leave my OIS in "Mode 1" (always on). The "Mode 2" (on just when you press the shutter) slows the camera down. Mode 2 supposedly saves battery life, but a second battery is trivial to carry. I've rarely needed it, even after a full day of snapping.

(It's been pointed out to me that Mode 2, though it does add shutter lag, is sometimes more stable than Mode 1, since Mode 2 always starts from the center "zero" position)

The controls are cramped. That's what happens when a device is tiny. The grip tape helps a lot, though occasionally I accidentally press the self-timer button (the < arrow) at inopportune moments. Likewise, when reaching for the AF/AE lock button, it's easy to accidentally press the joystick (which jumps you to the "quick menu" — invaluable when you need it, but camera-stopping when you don't). Practice is the only thing that can really help here.

At least one user on Galbraith's has complained that the AE/AF lock button is disabled when the camera is in manual mode. They want it to act like Canon's CF4 custom function. Happily, in practice this is a non-issue.

The reason it's a non-issue is because unlike most any other compact digital I've seen, the MF mode is set by a slider on the side of the lens — much like the AF selector on an EOS lens, but much easier to manage. Not a push button that resets when the camera is switched to another mode, nor a (yuck) menu item. It's physical, and it stays where you put it.

So if you're shooting in manual exposure, it's easy to just ignore the AF/AE-L button, and instead let the camera focus in AF and then flip the slider to MF, which effectively functions as an AF lock (with manual override, unlike the "regular" AF lock). It's easy to use by feel which is a big deal to me. I want to be looking at the scene, looking at the picture — not busying my eyes reading menus and other on-screen junk (and that includes looking for the little AF/AE indicator). The more you can know about the camera's state without looking at it, the better. Let your eye look for pictures, not mode indicators.

The manual focus has another rare feature for a compact — a depth of field indicator, indicated as a yellow band against the white range guide. Keen.

Using MF, the camera is very fast to respond to the shutter button. All good. Manual exposure, easy zone focus thanks to the DoF markers, fast shutter & 28mm view — my kind of camera.

The flash is low-powered and uses a preflash, like Canon et al. I've rarely used it. There's no connection provided for controlling external strobes. This hasn't had much effect on my shooting either.

A lot of web-review trumpeting has been made about sensor noise at the higher ISOs in this camera, and return trumpets have sounded about how great Noise Ninja is at correcting for the noise. Both contentions are true. I like using RAW mode and correcting noise problems later, I have a good workflow set up for this.

A feature of the camera is that when in RAW mode, a full size JPEG is stored — some folks have complained about this using too much space on the card (hey, extra 2GB cards are... what, $50-60 these days?). I like it. The JPEG serves as a good reminder of what you actually intended. I like that I can set the color process to B&W or sepia and the JPEG is stored using that setting, while the RAW file is... well, RAW, and contains all the original color information.

Beats me why the camera only supports USB 1.1 output. Ooops. I use an external USB 2.0 card reader, which is much, much faster.

I've finally become a convert to Adobe DNG — it makes life simpler for people with multiple cameras to keep all the files in the same format. So the LX1 RAW files, the CRWs from the G5 or the DSLR, all can live together in peace. My digital workflow goes:

Card Reader —» hard disk —» DNG Convert —» Adobe Bridge

and from there to Photoshop and an archive disk. I keep some LX1-dedicated Photoshop actions around, particularly one that labels pix with LX1-related flickr tags and copyrights, and another that can take 16::9 photos and ready-format them for the PSP.


Here are some links to people I've found using the LX1:

Posted April 22, 2006 | Comments (2)

Random Ugly Colors

Yes, I'm screwing with the CSS and setting the colors wildly to see where they'll go.... sorry for the temporary eyestrain :)

Posted April 12, 2006 | Comments (0)

GDC:06 Presentations Online

My two official-schedule GDC:06 presentations are now available online in slide-ument form — one presentation about general techniques in texture painting, applicable to anyone building models and painting them in games or movies; and the another specifically about the upcoming FX Composer 2.0.

As for more informal presentations given at the NVIDIA and Sony booths, those were variants of the FX Composer 2.0 talk. If you're a registered Playstation3 developer, you'll find the software and more info already packed-up in your SDK :)

Posted April 03, 2006 | Comments (0)

LX1

P1000096-807.jpg

This past week I received a Panasonic LX1, my first non-Canon digital (other than a phone camera). Tiny, 8MP, Leica lens (essentially, it's identical to the Leica D-Lux2) — and 16::9 aspect ratio, which was the Big Deal for me. So far: though the pace is definitely slower (and the ISO's lower) than using a DSLR, as a pocketable high-quality camera: fantastico. Loving it.

Posted April 02, 2006 | Comments (0)

 

All content on botzilla.com is 1994-2017 by Kevin Bjorke. All Rights Reserved.