Dave Jerrard, VFX Artist
February 08, 2007

Dave Jerrard is known to many that frequent CG forums, for his work, his helpful advice, and his habit of coming up with unique signature lines for each and every post that start with "He who..." Widely regarded as an expert on the subject, Dave Jerrard can truly be described as one of the first users of LightWave 3D. Over his 17-year career in the animation industry, his work has been featured on television and in film and he's served as the author of a number of magazine articles and books on the subject. We recently talked to Dave about his history with LightWave and in the industry.

Can you tell us a little about yourself and your history with 3D?
Well, back in 1990, I made the ever so logical jump from pizza delivery to 3D animation, when the pizza chain I worked at started closing stores. I had an Amiga 2000 at the time, which I had just paid off, and now I had some time to work on it. I had DigiPaint at the time, as well as a few others (Deluxe Paint, PhotonPaint, etc.) and Sculpt 3D. I started off in my free time, modeling the Enterprise and a Bird Of Prey, in wonderful HAM glory, and saved a bunch of images of these on a floppy disc. I also did a few weird 2D pieces of artwork, and I shared most of this with friends who worked at the store, where I bought my Amiga. They used a lot of it in a slideshow they had set up on a store demo, and before long, I was meeting people there that had copies of my artwork at home.

A short time later, a copy of that floppy that had the Star Trek stuff on it fell into the hands of the owner of Media Innovations, who was working with Gold Disk, creating a huge multimedia demo of their ShowMaker software, in Toronto where my brother worked. That guy called me up and I started doing 24-bit animations for his company for the next year or so. Those first animations were pretty crude by today's standards, since animation tools at the time were really limited. Just getting them to tape was difficult. At that time, we were using a Mimetics framebuffer and a Panasonic 8" laserdisc recorder, and we had to manually load an image into the frame buffer (which would take several seconds) and press the record button on the recorder to record a single frame. If the button bounced, we had to start all over. It was a painfully slow process.

Since then, I did a lot of corporate videos, then I did some LightWave articles for NewTekniques, which led to a job at Foundation Imaging, then UFO. I've been in California since August 1999.

How were you first introduced to LightWave 3D?
While I was working for Media Innovations, they were talking about getting something called a Video Toaster. I had no idea what that was, other than it had some kind of frame buffer in it. Then about a month before it was released, I saw my first screen shots of LightWave in one of the flyers. That looked so way beyond what I was using, what with that fancy 3D perspective view and all. I was actually a bit intimidated by it, and this was version 1.0!

By the end of that year, I was set up with one of the first two Video Toasters to arrive in London (Ontario). My roommate had the other one. I did the usual fiddling around with live video for a few hours, and then finally clicked that 3D button. I loaded up an old F4 Phantom model I built in Sculpt, and played with some texturing on it. This was entirely new, since Sculpt only dealt with solid colors. I was hooked. All those wonderful textures, and a full 24-bit output to boot! And the best part for me was that I didn't have to select all the points in an object to move it anymore. Within ten minutes, I had my first animation rendering, of a sword being wielded by an unseen warrior before being hurled toward the camera. And I did all this without ever having to decipher any weird icons or glyphs anywhere. This was easy! I never did get around to reading that manual, beyond the introduction to 3D part.

My first professional animation used that F4 Phantom, where it was chasing a Lamborghini along a road with some Vista generated mountains mapped in the background. The next one involved an image mapped Rubik's Cube solving itself in space.

What feature of LightWave do you like the most?
That's a tough call. There's so many to pick from these days. I would probably have to go with its ease of use. That's what got me hooked in the first place. The render quality is a close second, especially with 9.0. I love the new cameras, which provide me with features that I've wanted for about ten years now. The modeler is also pretty good, though it took a while for me to get into it with version 1.

Are there any plug-ins you use on a regular basis?
I don't really use many in Layout, but I have a flash drive full of them for Modeler. I think FI's Weightmap blur is probably the one I use the most. My favorites are pretty much anything by Pictrix, FI or Shift. I also have LW-CAD, but I haven't used that a lot lately. I've been spending most of my time these days in Layout. On the Layout side, I use FPrime for getting lighting right, but recently, I've had to go back to the older F9 tests since I've been taking up the nodal texturing quite a bit.

Do you work primarily for a studio or freelance?
That varies from job to job. It's pretty much even at this point - half studio, half freelance. I actually prefer the studio jobs, since that gets me out of the apartment for a change, and when I get home, I'm definitely not at work anymore. Freelance is a bit trickier to get away from, but then it's usually pretty quiet with no-one to talk to during the day.

How do you find freelance work?
It finds me! Last week alone, I was asked about my availability for about 10 different projects. It gets frustrating sometimes when you know someone really wants to hire you, but they have to wait to find out if a project's going to happen and when. It could take several months, like the spider project I just finished working on, which I was originally approached about last February, or it could take longer, like the one I just landed last night.

Most of it is word of mouth, through people I've worked with before. In these last two cases, it's studios I've worked for in the past. I haven't sent out any reels since earlier last year. I have to admit, this is the most potential projects I've had lined up at one time. It sounds like there are a lot of projects underway out there right now. I guess the news of a possible writer's strike later this year has pushed a lot of work out the gates to beat the deadline.

What are some projects you've worked on recently?
I just finished up a project with BFX Imageworks, where I had to create six giant spiders, on a budget. Originally, these were going to be six specific species, but things didn't work out so smoothly. By the time the studio was ready for effects, there wasn't much time left to do them, and these six spiders ended up having to be variations on a single body and had to fit the same bone rig. While I was modeling and texturing these, my wife worked on the rig and some expressions to get them to walk automatically.

In about two months, we had six spiders, and several morphs built in, in case the client needed to alter their body shapes even more. He was able to animate his shots with the first model I provided, and then swap that model out with the correct one later, before rendering, and everything would work.

In addition, I did some commercial work for Wal-Mart pharmacy and some modeling for Olympus video. I also worked on some very-last-minute effects for a music video, and I'm sure there are a couple other jobs in there that I'm forgetting right now. Other than that, I've been keeping a bit of a low profile.

I've got a few other plates in the air now. I'm just waiting for any one of them to get a green light. In fact, I just got hired this week to do some effects work for an upcoming feature film. That one's been dangling in front of me for over a year now.

Are there any projects that stand out as particular favorites?
I'd have to say that was working for Stan Winston. It was originally going to be a 4-week job, but it grew to three months, and I was brought onto 2 other projects there at the time. I worked on a dream sequence for Because Of Winn-Dixie, where the main character was told that her mother was such a good gardener that she could plant a tire and grow a car. The sequence was a vignette vision of that.

The entire shot, which was about 180 frames, started with a close-up of a mound of dirt that was just watered. The camera pulled back from it as a small CG VW bug popped out of the dirt and began growing to full-size. Matching the CG car to the real one they had on set was easy, but getting it to be dirty was a challenge. I hand keyed several chunks of tumbling mud that the car pushed out of the mound as it emerged, and used particles with HyperVoxels to create a lot of dust falling out from between the panels. As the car grew, I had to make it appear to get heavier. So, as it grew across the lumpy terrain, I had the wheels follow the ruts in the ground, and animated the car rocking on the suspension, with progressively longer, deeper rocking motions. The tires even left ruts in the ground, thanks to Worley's Polk.

The entire shot was lit solely by an HDR image, involved about 13 elements, and was rendered at 2k resolution. The car element alone took nearly two nights to render - this was with LightWave 7.5. FPrime wasn't available yet - it had only just been announced, and I didn't get my copy until about a day after the shot was completed. It would have made texturing go faster, since I had to see how everything looked under the same lighting condition the real car was photographed under - the HDR image. It was a pretty dead on match that had several people unsure which images were the real car and which were CG. The real car ended up never being used in the final.

I did some remodeling of a model they already had to match the actual car, all surfacing, lighting, animation and particles, and the only other person to work on it did the final compositing. All that remained from the original plate (which was a muddy field, a dolly track that had to be removed, a sky and the mother) was the sky and the mother. Everything else was LightWave.

I was the only LightWave artist there at the time, and I was the only one that did the majority of a shot like that. Everyone else was using either Maya or XSI, and several people were usually working on a single shot in those packages - one would be doing tracking, a couple more would be doing lip sync, a few more would be doing lighting, etc... Several seemed quite impressed that a single person was pulling off a shot like mine.

Are there any projects that stand out as particularly challenging?
That would either be Last Samurai - the shot where Tom Cruise's character gets off the ship in the Japanese harbor, which was a long, tracking shot that had to be rendered at about 6k because the plate started out as about 1/4 of the final frame, or the CG Ronald McDonald I worked on at Warner Bros. The Samurai shot would not track properly - every tracking solution resulted in the tracking points having a forced perspective, so I had to fudge the camera motion and view a bit to keep everything as closely matched as possible. That was a long shot, and the only good tracking points were way off to the sides of the frame - the center had a large number of people moving through it, which couldn't be tracked. The shot was mostly set extensions and the addition of a large matte painting.

Ronald was a project that about 6 or 7 people worked on at one time or another, but was mostly handled by four - myself, Jennifer Hachigian, Larry Shultz, and Tom Dickens. Tom and Jen were also working on another project at the same time. I mainly did surfacing, trying to get it as real as possible (they wanted something that looked like Shrek quality), which became a challenge since the model was still being worked on, by TWO different people at the same time. This resulted in one person making UV maps, while another was making endomorphs.

This led to us trying to get these two different sets of vertex maps, on two different models, combined into a single model, which was also being revised by another person. To complicate things further, the client wanted to see how the suit would look with dynamics applied, even though there was no rigging, so someone had to fake it by modeling wrinkles into the suit. This left some pretty ugly geometry for UV mapping. The suit had very noticeable stitching going up & down the legs, and those wrinkles caused all kinds of polygon flow problems. Once I got that worked out, a few renders were done, and then I had to remove those wrinkles again, clean up the geometry for the real dynamics, remake the UVs and continue texturing the suit again.

When that was done, Jen had finished the hair, done with Sasquatch, but it had to be changed for a third time. Jen was now working on dynamics and I created a whole new wig, which took about a week to get just right (again). Larry did the rigging and facial morphing, and the character ended up with about 80 endomorphs, which really pushed LightWave [8] to its limits. Tom did hand-keyed animation that brought the character to life, which I thought was motion capture at first. That project took just over four weeks.

Is there an area of 3D work that you feel to be your specialty?
It would have to be texturing or lighting. That seems to be what most of the projects involved that I've worked on recently. I've done a lot of effects animation as well, but getting stuff photo-real is what I'm getting known for. I guess it goes well with the practical joker in me. I like making images that fool people into thinking they're real. The best compliment is when someone doesn't know they just saw an effect.

You've also written a number of books and articles, which do you prefer?
They both have their strengths. Books have the benefit of a longer time frame, and I can go into far more depth than I can with a magazine article. Magazines have shorter deadlines and color images, but they tend to also have pretty low word count limits. I had to try to squeeze an article for creating trees into 700 words for one. That was a real challenge since my first draft of it, written in a record day & a half, was originally 2200 words, and that was mostly meat. Trimming out two thirds of that, without losing important steps, gave me a new respect for editors.

Do I prefer writing over doing, it actually works out to both. I do before I write, and I tend to write about what I do well. Half the time it's something that someone else is trying to figure out how to do, so I'll take a shot at it. Often, I'll end up taking the wrong path at first, which can frequently lead to useful discoveries. Other times, it's just a project that I'm working on or just me playing around. I almost always try to do something different in every project, which helps broaden my skills, and also breaks the monotony that can occur when you have, say, 20 shots of a submarine - there's not much to see, but you have to make it look good.

I haven't done any real writing for a while, other than posts on the LightWave forums. It can take a lot of time to write material, and with the rapid pace of development going on right now, a book is almost certainly going to be out of date by the time it hits the shelves. I'm still trying to catch up with all the new stuff in v9.0, and v9.2 is already in open beta!

Are writing projects usually your idea or the publisher's?
They're usually my ideas. The NewTekniques articles were all based on things I was trying at home. The first book was something that Advanstar and Joe Tracy (then editor of NewTekniques) came up with, and they asked if I'd be interested in doing it. From then on, just about everything was in response to a publisher's requests. The subjects are almost always my ideas though.

As an active forums user, what is your opinion of the tone of the venue?
Over the past year, I've noticed a distinct upswing in opinions, mostly due to the features in 9.2. There's still a lot of moaning, but it's definitely thinning out in favor of the new tools and bug fixes. Even I had my doubts about LightWave a couple years ago, but now I'm checking my email frequently every day to see if there's a new build to play with. What new toys is NewTek giving me today? In the past month, quite a lot! That's cheering up a lot of the doom-sayers, who are also getting excited about the way the software's improving. Of course, it's never going to be fast enough - development is never fast enough for users - but it's taking big strides in the right direction. I know one particular character animator that was really depressed a year ago because it was difficult to do convincing flesh and renders were taking 30 minutes or longer. Now he's thrilled that he can render stuff in a fraction of the time it took before, and his complaints are now more along the line that decent skin takes over 5 minutes to render. He's had to raise his expectations, which isn't a bad thing.

I've seen other posters do the same, and it shows with the increase in the number of posts now. I used to be able to wade through an entire day's worth of posts in a few minutes. Lately, it's taken up to a few hours to catch up with them. I've also noticed more new faces, or names, appearing regularly as well.

For the most part, people are generally quite helpful. I actually picked up a fair amount of information on the node editor that way, both from Antti's posts and a couple of the node-inclined users out there (I'm one of those firm believers in "when all else fails, read the manual"). I haven't mastered them yet - the math stuff isn't my thing - but I use them almost exclusively now. I still think that's going to be the biggest stumbling block for people, and the forums are definitely helping people stay on their feet. Luckily, no-one's being forced to change their ways. Instead, they have more options available.

Have the recent changes in LightWave impacted the way you work?
Yes. They've forced me to think a bit more! Nodes are the big thing for me now. At first, I didn't get them, until I realized that you kinda have to look at them with your head tilted to the right. They're a bit like layers would be if you tipped them over on their right side. Once I figured that part out, everything kinda fell into place. They still take a bit of thought though, and I'm still trying to find a magic phrase that will help other people get them. I hope the head tilting helps.

Are there any new features of LightWave v9 you've found particularly useful?
Definitely the nodes, though I tend to stick with layers in a crunch. The Advanced Camera is also used a lot. I set up a mesh for it which turns it into a quad-view camera. I use that a lot when I'm texturing an object now. It's faster to be able to render four different views of the object in one render than it is to render multiple views. I can see exactly how light affects surfaces from different directions, which is really important when it comes to subsurface effects.

Do you have a "LightWave tip or trick" others might find helpful?
Well, there's that Quad-view camera trick I just mentioned. That's probably my most recent trick, or at least the most recent one I can remember. My most famous are the Spinning Light Trick and the Laser tutorials I wrote several years ago for the old NewTekniques magazine. I still get questions about those quite frequently. I've posted a lot of tips on the LightWave forums, and I'll probably keep doing that for quite a while.

Do you have any advice for someone new to the 3D industry?
Don't do it just for the money! I've run into people that got into it for this reason and they really aren't happy. I got into this because it was a hobby. Eventually it turned into a job, and then a career. But I did it because it was fun. If you don't enjoy it, you're not likely to stick with it very long. With the way the visual effects industry is (especially with a potential writers strike coming up), I can't advise anyone to take that jump. Not unless you got some backup cash to fall back on. But that's just one of the 3D markets. There's a lot of medical & scientific imaging, architectural, and game design fields out there. But the important thing is you have to like what you're doing.

What's next for Dave Jerrard?
After this current project, I wish I knew. Definitely books. Don't ask when. I'm getting hounded for another LightWave Applied book, but LightWave has been getting so many new features so quickly, it's hard to cover them before the next version comes out. I could fill a couple books with just node techniques alone. I've been asked about my availability for several different projects coming up, so I think this year's going to be a busy one. Plus, I have some personal projects I'm working on when I get the time.

Thanks for taking the time to talk to us, Dave!