Fred Tepper, VFX

“Doing CGI back then was tough because computers were so slow.

On one computer it could take anywhere from half an hour to several hours per frame, and this was only at the old video resolution, which is about a quarter of HD…”

“Since Steven Spielberg had always used ILM for his films and shows, that was the initial idea for seaQuest. But the show’s production designer Richard Lewis convinced Steven to allow a CGI test.

Richard brought on Joe Conti, whom he had worked with before, and a test was made using off-the-shelf LightWave 3D on an Amiga 2000 computer. That test convinced Steven and Universal that CGI could achieve the look they wanted and be cost-effective. That was in the spring of 1992.  At the time I was living on the East Coast but I happened to visit Joe, who I had known since I was fourteen, just as the test was completed.

Joe and I had been doing little visual effects films since the beginning, but obviously not with computers, so when I saw what was possible, I was amazed and wanted to learn it. (I had sworn off doing animation for years, since a time in college where I spent an entire month doing an animated short, only to get the film back from processing and finding that due to a torn sprocket which had apparently happened on the first day, all I had to show for all that work was about 16 frames.  I said I’d never do animation again until I had proper equipment.  Little did I know that “proper equipment” would come to mean “a computer.”)

Joe said if I learned how to use LightWave he could hire me, because at the time there just wasn’t an existing pool to hire from. So I moved to Los Angeles in September ’92, bought an Amiga 2000 and borrowed a Video Toaster (the hardware from NewTek that came with LightWave) and started learning.  This was long before YouTube, Udemy, books on the subject, dedicated forums, and all of the other things available to help you learn software (or anything else) now. There was just the software manual, and that’s it.  Luckily the manual– printed and in a three-ring binder– was good.  So back then being self-taught really meant self-taught.

During this time Joe was setting up what would later be called Amblin Imaging, on the lot at Universal Studios. I visited there a few times that fall.  There were only a handful of people there, maybe six total. But the models and R&D they were doing looked amazing. In fact it looked so good that I decided I had to be sure that I was clearly on that level, so no one would think Joe hired me just because he knew me. (Well, he did, but I had to make sure that was irrelevant.)

So rather than just learn the software, I decided to make a little project, and came up with the idea to make a little scene of the sunken Titanic.  I wouldn’t be able to create an animation though, because on one computer all I could render per day was one or two frames. So an animation would have taken forever– about a week per second. So I just made the models, texture, lighting, camera composition, etc. as good as I could –


“This is the sunken Titanic I made while learning CGI.   It later became the King George, and the Titanic for the movie (but not for 1912 “new” shots, not for present day wreckage shots…”

In mid December I brought the images in and showed them to Joe and a few others there.  Everyone liked them enough to say I should join the team, so in the beginning of January I did. Aside from Joe, only one of the people there had done any visual effects work for film or television before. Joe had found most of them through NewTek – they were LightWave hobbyists.

Side story:  One of the people hired a month before me, Eric Barba, had initially come to Amblin Imaging to demo SGI computers. Joe told him he wasn’t interested in the computers, because they were tens of thousands of dollars (compared to around $1200 for a fully outfitted Amiga 2000) and there was no way the production could afford it, but that he was impressed with Eric’s enthusiasm and computer knowledge and offered him a job. Eric took it, and his first task was to go to Steven Spielberg’s house to get the SGI Indigo computer that SGI had given Steven. He was letting us use it, and that’s what the initial modelling of the seaQuest, Stinger, and Darwin were done on. Years later, Eric won an Oscar at Digital Domain for Benjamin Button. These days he’s a VFX Supervisor at ILM.)

My first assignment though wasn’t doing visual effects, it was creating graphics for the screens on the seaQuest and other sets. I worked under Ron Herbst doing this. Ron was great at designing screens. One of the main edicts from him was to stay completely away from the current look of operating  systems. I’m not sure how dated the screens may look now, but they’d have absolutely looked badly dated by now had it not been for Ron –

The set up on the stage (primarily Stage 28) was done through a company called Playback Technologies, run by Steve Irwin (still is!) and involved miles of cables, all of which led back to a small room built into one of the opera boxes from the 1925 set for The Phantom of the Opera, which was still standing on half the stage. The main screens of the seaQuest, and any screen bigger than twelve inches or so, were projectors, Barco for the main screens and Eiki for smaller ones. The screens on the navigation seats were very early LCDs. We were also on stages 14 and 15, which were connected to each other and could be made into one big stage. To me these were cool because they were the Abbott & Costello stages –


The set for the SeaQuest Bridge on Stage 28 framed by scenery remaining from Phantom Of The Opera

I worked on-set for the pilot and parts of the next two episodes. The best thing was that the pilot was directed by Irv Kershner, whom the entire VFX team was a huge fan of because he’d directed The Empire Strikes Back.

Side story: On the Thursday of my first week there, Irv came by to meet everyone and decided that the best way for us to all get on the same page and see what we were up against was… to go with him to the Amblin screening room and watch Empire with him. It was amazing! He gave a live commentary the whole time and answered any questions we had. Then when the pilot started shooting he’d always come and sit in the playback room when he had time to take a little break, so we got to chat quite a bit.


“This is the season one’s video playback room on Stage 28 (located inside one of the opera boxes from the Phantom set). This is where the screens were controlled for the Bridge, Lucas’ room, the MagLev, and the area with Darwin’s tank. I believe this was taken during episode 4, Games…”

Side story of a side story: At that screening, sitting next to me, was another VFX artist, Richard Payne. I’d only met him when I first visited in October, but he and I were, unknowingly at the time, at the same theater at the same showing on opening night of The Empire Strikes Back in 1980.  And now here we were, watching it again but with the director!

The art department shared the facility with us, so we worked closely with them. And with the great designs on the show, from Production Designer Richard Lewis and Art Director (originally) Jim Lima, and especially illustrator John Eaves (check out his Star Trek book on Amazon), plus the above-the-line talent involved, we thought this would be the next Star Wars. Unfortunately when we read the script we realized that it was still just regular old network television. Understandable, sure, but still a letdown.


A peek into the SeaQuest Art Department reveals blueprints, continuity Polaroids and a 3D model of the Stinger (upper left)

All of the screen graphics for the bridge and other stage 28 sets were controlled live from our control room using a program on the Amiga called Elan Performer, which let you assign animations to different keys on your keyboard. So you’d just click on a key and it would play the animation assigned to it. We had a full set of Amiga computers as well as several 3/4 inch video decks. And we had a video feed from the main camera so we could see what was going on, for timing purposes. For other sets we had a “portable” system. It was about as portable as a barn.

Shooting the pilot was a lot of fun, and everyone on set was excited. For Roy Schieder, a production this big was old hat, but for most of the others in the cast it was the first time they’d done something of this magnitude. I don’t recall any tempers on the set, ever.

Some random memories:

– At one point we were shooting a scene in Stark’s submarine set with Shelley Hack. The camera assistant slated the take and everything was quiet. Then I hear Irv: “And cut.  I mean action!

– We were shooting Bridger’s quarters but the sound mixer, Will Yarbrough, said he was hearing some kind of buzzing. He suspected it was a cable, so he switched it out and asked for quiet so he could check the new one. Everyone shut up and stopped moving. It was totally silent. And then we hear Irv say, “Annnd action.” The first assistant director, John Kretchmer, hold his laughter, says “No, we’re checking a cable.” And then everyone started laughing – including Irv.

– Roy would get his lunch brought in from Jerry’s Deli on Ventura Blvd.

– On some days the craft service table would have the largest bowl of shrimp I’ve ever seen, and really good shrimp cocktail sauce to go with it.

– One night after wrap I thought it would be funny to put one of the then-popular car security devices, “The Club” onto one of the Nav chairs on the  Bridge so they’d find the ship “secured” in the morning. I figured there’d be a little laugh and I’d take it off. But I was a little late the next morning and when I went in, they were, desperately trying to figure out how to remove it. The person who got the biggest laugh about it was the Director of Photography, Kenny Zunder.

– One of the producers and I used to go on a seaQuest newsgroup all the time. (Remember Newsgroups?) No one knew who we were, even though we used our real names, and we’d participate in chats, usually, for us, laced with inside jokes that no one else would get.

– One day, about halfway through the season, we got a visit from Dick Van Dyke. He was a fan of the show, but more a huge fan of the CGI. He told us that during the production of Mary Poppins he got bitten by the visual effects bug, so when he finished on that he set up a little bluescreen studio in his garage, and would make little films with his kids, and then grandkids. When CGI came on the scene, he got into that, and he used the same equipment and software we did. He as easily as excited to see us as we were to see him. And I remember how odd it was to be talking to FREAKING DICK VAN DYKE about CGI and having him as super in-depth questions and fully understand the answers. We told him that if he wanted a job we’d gladly hire him, but he said he was in the middle of Diagnosis Murder– and he actually seemed bummed. For a couple of years at the big Siggraph computer graphics convention, he demoed LightWave for NewTek, before a crowd of very surprised computer graphics artists. Several years later CBS did a Dick Van Dyke show reunion, and in it Dick’s character was retired and just making little animated films for himself as a hobby. The clips they showed really were Dick’s.

– Another visitor who came by because he wanted to get into CGI was model/actor Fabio.  I’d worked on a film with him in New York a few years before, so it was fun that he remembered me and was surprised to see me there. We also had visits from Walter Williams, the creator of Mr. Bill on Saturday Night Live. He had decided to do some CGI Mr. Bill in LightWave and would come by to chat about it.

– Since we worked for Amblin, we were invited to the cast and crew screening of Jurassic Park up at Universal City Walk. After the screening, I saw Ray Harryhausen, my idol and inspiration from childhood, walking around on City Walk. I stopped him talked to him for a while (he was surprised anyone would recognize him). It turned out that he had just seen the screening too. He loved the film and I asked him what he thought about the CG dinosaurs. He said, “You know, I never really paid much attention to computers before, but they’re really quite wonderful!” He gave me his card and we stayed in touch.

– Steven Spielberg has an arcade in the main Amblin Building, and we used to go there during breaks to play. But eventually we were banned from there for being too loud.

– After the seaQuest pilot was done, Irv was planning to do a remake of Forbidden Planet. He was really into it and it was going to be serious science-fiction, delving far more into the Krell civilization than the original film had. He hired me and a few others on the Amblin Imaging team to do a few short proof-of-concept shots of the new Robby the Robot and a couple other things. (The new Robby was a John Eaves design and it was fantastic.) He was never able to get funding for it, though every couple of years I’d see something about it in the trades, but the last time was probably twenty years ago. I think Jim Cameron was attached to it at some point, but I don’t remember exactly.

– One of the most fun things in the seaQuest days was that we had golf carts to get around the studio on. At lunch a couple of friends and I would go exploring the backlot, sometimes having a little picnic a the Psycho house or Back To the Future square. We’re in a LOT of tourist photos. There’s a ton of history on that lot and it was great to see it and explore it like that.

– During the second episode we hired someone else to do the on-set playback  and I went back to being mostly only in the office. I continued to work with Ron Herbst doing graphics, but also did some VFX, culminating in blowing up the seaQuest as it sat in the lava in the season finale. Plus the Titanic I made, back when I was learning LightWave, ended up being the King George. (And later it was the basis of the digital ship I co-created for Jim Cameron’s Titanic, which several of us went on to after Amblin Imaging closed.)

Doing CGI back then was tough because computers were so slow. On one computer it could take anywhere from half an hour to several hours per frame, and this was only at the old video resolution, which is about a quarter of HD. We were using, as mentioned, Amiga 2000 computers, with an 040 accelerator card. The software was LightWave 3D, and other than always having the beta versions with the latest features, it was just off-the-shelf software that anyone could get. We didn’t even have any plug-ins or extensions, because at that time none existed. Not only were they slow on rendering, but there was a limit to the number of points a CG model could  have. CG models at the time were made of squares and/or triangles, so each square was four points and each triangle was three.  Making something as detailed as the seaQuest took a lot of points, so the high-res version of it had  be broken into pieces: you’d only load what the camera would see. If you were far enough away to see the whole ship, then, by definition you didn’t need it to be high-res, and you could load the low-res complete ship.  In reality you wouldn’t even be able to see more than about ten feet unless you’re right at the surface of the ocean, so even just seeing something like the entire head of the seaQuest is a major cheat.  The network never would have allowed a more realistic visual falloff, because you wouldn’t be able to see practically anything. We ended up varying the falloff on a per shot basis– sometimes you just needed to see more than other times. We also had to battle with how dark it should be. Sometimes, because there was no time to fix it, shots ended up too dark.

For the first season we had a render farm of twenty computers, and this meant that we could get several shots rendered per day. Still, there was never time to get more than one or two takes rendered. You had to get it right the first time, so you’d render individual test frames to see how it looked, and hope that in motion there were no problems. Once a shot was rendered, there was no way to play it back on our computers– no Quicktime videos, or MP4, etc. We had a videodisc recorder instead. It would take about thirty seconds to a minute to transfer one frame to it, so, again, things were painfully slow. Even though the live-action shoot was regularly way behind– in the first season I don’t think they finished any episodes in much less than double the scheduled time– there were times when we delivered shots on Saturday night that were to air on Sunday. When this happened, editorial would slug black into the edit and that’s what would get sent to New York for prep to air. Once we sent the shot, they’d put it in. If we didn’t get the shot done, a black screen would air. Luckily it never came to that. Another thing that was tricky, though it might be over most peoples’ heads was that except for the last season, and only a few times even then, all of the compositing was done in the 3D software (LightWave); we didn’t have compositing software. So if there was a live action background with CGI in the foreground, that was done in one pass, rather than bringing both elements into a compositing program. This meant that once a shot was rendered it was completely finished. But it also meant that getting it to that point was much harder. As a side note, we also only had one seat of Photoshop (on our lone Mac) and that was it. That was what the seaQuest’s skin texture was created on, by Jennifer McKnew, working with input Jim Lima. For most drawing, texturing, and for all screen graphics, we used Deluxe Paint on the Amiga.

Each year the render farm grew, but of course each year what was asked of us grew too. The worst thing was when the Northridge Earthquake happened, early Sunday morning on January 17th, 1994. Since the render farm was on and running, all of the hard drives got destroyed, because they were spinning during the shaking (until the power died). (Side note: the production shut down for a while because the stages had asbestos insulation in them, which had to be cleaned out.)

So how did Amblin Imaging do on its promise to be more affordable than ILM doing it with models and film? We delivered the entire first season for only  about 50% more than what they had budgeted for just the pilot. The FX may not hold up very well today, but considering the era they were made in, they’re  not bad. In 1995 we started transitioning to PCs, which were much faster than Amigas. We tested a 90mhz(!!!) Pentium and it rendered almost five times faster than an Amiga. I’m writing this on a 3.2ghz computer, so it’s theoretically the equivalent of about 177 of the Amigas. The seaQuest high-res model– which can now be loaded in its entirety– renders almost in realtime now. (I also remember us, in 1993, getting a 1 gigabyte hard drive– ONE gigabyte– and it cost just over a thousand dollars. And I remember saying to Ron Herbst, “We’re never going to have to delete anything ever again!” Up to that point, 200mb was “big,” and the Amiga I learned LightWave on had a 52mb hard drive. My current cell phone has a THOUSAND gigabytes– a terabyte. And forget about how much my computer has.)

During the years seaQuest was in production, we also did other projects I did screen graphics for the show Space: Above & Beyond, and for a Rockford Files TV movie, and some shots for the M.A.N.T.I.S. opening titles, and Earth 2. As Amblin Imaging we did some work on an episode of Baywatch, some episodes of The Adventures of Brisco County Jr., some of the Weird Science TV series, Sliders, and most importantly, the first season of Star Trek: Voyager, for whose pilot we won an Emmy for. There was talk of Amblin Imaging continuing beyond seaQuest, and when Dreamworks SKG was announced, Spielberg told us we’d be a part of it.  But that didn’t happen. Interestingly though, many years later, after we finished Titanic and left Digital Domain to start our own company Station X Studios, Dreamworks came to us to do a direct-to-video sequel to their film Antz, but the project ended up getting cancelled.

These days I’m back to mostly doing screen graphics, and only sometimes do I do visual effects. The Amblin Imaging/seaQuest team is still in good contact with each other, and I believe everyone is still in the industry in some way…”