The Toaster and Tim's Vermeer
COW Library : TV & Movie Appreciation : Tim Wilson : The Toaster and Tim's Vermeer
When I heard about the documentary "Tim's Vermeer" at the 2014 Palm Springs International Film Festival, it sounded kind of interesting. Here's the description from the festival program:
Any good mystery begins with a series of unanswered questions that get under the skin. In the case of Tim's Vermeer, the fascinating and quirky documentary by magician team Penn (who produces and talks) and Teller (who directs and stays mum), the questions piled up fast when Tim Jenison, a rather obsessive inventor friend from San Antonio described his current fixation with the 17th-century Dutch master painter Johannes Vermeer: "How did Vermeer paint with a photo-realistic detail that seems practically impossible to render with the human eye?"
Duplicating a Vermeer, Penn & Teller, a witty catalog description, a maniac named Tim: this combination was a no-brainer.
But wait. This "rather obsessive inventor friend" Tim isn't just any obsessive inventor. He's one of the most important figures in the desktop video revolution. You may not know his name, but you know the name of his most famous invention.
The NewTek Video Toaster.
Calling the Toaster the first shot in the desktop video revolution is an understatement. The Video Toaster added high quality, high performance video cards to an Amiga personal computer, and a software suite that included LightWave 3D. There had never been this combination of video and graphics power in one place, at any price -- and it happened that the price was incredibly cheap, only $2399.
NewTek Video Toaster
When the Video Toaster was released in 1990, there was no Adobe Premiere. No Avid Film Composer or Media 100. This was before CoSA After Effects, before Photoshop was released, even before QuickTime. It was long before Final Cut Pro, of course. In fact, when Tim founded NewTek along with Paul Montgomery in 1985, the Mac was still limited to strictly black and white, on a 9-inch screen.
For a while there, Toaster was very nearly the entirety of the desktop video revolution.
The revolution of Toaster wasn't just that individuals could create video on desktop computers. It was that they could create 100% broadcast spec-compliant video. The combination of features, price and exceptional image quality led to NewTek's 1993 Emmy Award "For Developments in Television Engineering," singling out the Video Toaster.
Tim is the first to admit that the Toaster is also remembered for its cheesy-looking graphics, and its distinctively playful video effects and transitions, but, as he notes, "they were broadcast quality cheesy-looking graphics."
(As someone with a video production company in those days, I can testify that the cheesiness was more endearing than anything else.)
In Tim's mind, quality wasn't just a prerequisite for broadcasting. It was, and is, an obsession.
Tim has had a lot of obsessions along the way. Over the course of our conversation he variously described himself as a film geek, a darkroom geek, an electronics geek, and an audio geek (are you sensing a theme?), but, ironically for this tale, most certainly not an art geek.
Still, it's inevitable that a photography / film / electronics / video / image quality nerd would look at art a little differently than most other people. In Dutch master painter Johannes Vermeer (1632-1675), Tim saw a kindred spirit: somebody interested in capturing the world with uncanny detail.
The more closely Tim looked at Vermeer's work, the more he became convinced that Vermeer was using technology to aid the creation of his art. He became certain that this included Vermeer's variation on camera obscura, a system of lenses and mirrors to project images. The idea had been around for centuries by Vermeer's day, but better lenses were making camera obscura more useful for painters. These new, improved lenses allowed them to project high quality images on a board or canvas, to use as a guide for creating their paintings.
Tim was certain that that's how Vermeer was able to create such photorealistic pictures. So certain that, having never painted before, he became confident that he could paint something very much like Vermeer, and very nearly as well. He decided to try and duplicate The Music Lesson.
The Music Lesson
As crazy as that sounds, it gets a lot crazier. Tim built a full-scale replica of the room in the painting. The details went all the way down to the pattern on the rug and the paper decorating the exact model of harpsichord in the picture -- which Tim built himself. He lathed and milled all the furniture. He made his own paints and brushes using materials that Vermeer would have used.
Tim even ground the lenses himself, using 17th century methods.
Spoiler Alert: Tim pulls this thing off.
With no real painting experience, he successfully recreates The Music Lesson, with closer to Vermeer-like quality than you'd ever imagine -- and art world chaos ensues. Could Vermeer possibly have worked this way? If he did, did his reliance on technology make him less of an artist?
Actually, I'm not giving anything away at all when I tell you that Tim succeeds. As Tim's Vermeer goes along, the film becomes a little less the story of Tim painting a Vermeer. It becomes the story of the obsessions that drove Tim to take his ideas as far as possible, and then a few steps past that.
Think of this as a Penn & Teller-style reveal. The way they show how they're about to perform the trick before they show you the trick. The reveal before the trick actually adds to the magic, because you know exactly what's coming, and you still can't believe your eyes.
There's no smoke in Tim's Vermeer, but there are mirrors -- which makes the result no less magic. You now know exactly what you're going to see, you're going to see it, and you're still not going to believe it.
Until you do.
'Tim's Vermeer' Trailer (2014): Tim Jenison, Penn Jillette
The unlikely story that became the unlikely documentary Tim's Vermeer has had an overwhelmingly positive response, with festival plaudits from around the world, and a 93% Fresh rating at Rotten Tomatoes.
You've likely picked up by now that I'm a Tim Jenison fan, and a Penn & Teller fan. I also adore the movie, Tim's Vermeer, that these three made together. This leaves me anything but objective. Trust me, though. It's going to work out fine. You're going to be amazed by this wild ride through the obsessions that tie together Tim, the Toaster, and Vermeer.
Creative COW: Long before Tim's Vermeer, the place that most of us got to know you was through the Video Toaster. What was the origin of that?
Tim Jenison, NewTek: I was a film geek, and an electronics geek as a kid. I had an uncle work for IBM, and my dad was an electrical engineer, so I learned by osmosis. I was also a darkroom nerd and made some 8mm movies, so when video equipment started to become cheap, I latched onto it.
When personal computing came along, I glommed onto that too. I always felt that video and personal computers belonged together, and that at some point, personal computers would become useful for video production.
They just weren't up to the task at the time. The IBM PC was only a few colors, the Mac was black and white, and neither was fast enough. The Amiga computer wasn't fast enough either, but it was the first one that was video-centric, so I was able to graft on some high-performance video circuitry.
You weren't just interested in performance, though.
I knew that quality was important, because the gatekeepers of video at the time were the networks and television stations, and chief engineers were not easy to please. I figured if there was anything wrong with the video, it would be a great excuse not to allow a Toaster on the air.
I dug into all the specifications, like the RS-170A spec, which describes what video is supposed to be like for broadcast quality, and I made sure the Toaster would meet that spec. You know, there were some cheesy-looking graphics on the Toaster, but they were broadcast-quality cheesy-looking graphics.
The acid test came just after we shipped the Toaster. We went on a network television show, Computer Chronicles on PBS, to demonstrate it. While we were setting up, I said to one of the engineers, "If you want, you can take a cable from the Toaster, and plug it into your switcher for a clean feed."
He said to me [in a condescending tone], "You've got to understand, computers are not compatible with broadcast video." I said, "Well, this one is compatible with the RS-170 spec," and he looked at me like, "Ah, the monkey speaks English!"
Behind him was a guy with a baseball cap and a coil of coax, and he plugged us in. As he does, he looks into the control room where there's a commotion going on. I ask him what's happening and he says, "A bunch of guys made bets whether this would actually meet the RS-170 spec" -- and the guys who bet against us lost the bet!
You started NewTek with other products, but did you get a sense what you were on to when you came up with Toaster?
Yeah, I did. I grew up in the days of three networks, and by the time we came out with Toaster, it was 36 channel cable, and it looked like that trend was going to continue, that there would be more and more video production. But it was inherently very expensive. I knew that if I could do something that was 10 times cheaper than the traditional way that it would be successful.
And the Toaster was successful, and a lot of people got into video production. Once the internet came along, and streaming video, Pandora's box was opened.
That's where TriCaster starts to fit in, right?
We called it "tri," meaning three, because your program output can go three places. It can go to standard video, like to tape or a hard drive. It can go out to a projector for image magnification for live events. The third is streaming to the web, and it can do all three things simultaneously.
Also, "TriCaster" sounds like "Telecaster," which is one of my favorite guitars. It also sounds like "Tricorder" from Star Trek. It was a combination of all those things.
Okay, this is a little sideways from the film, but I want to hear about you and the Fender Telecaster guitar.
I played in several rock bands, for four or five years. I designed a synthesizer, and my very first product for a personal computer was an automated mixdown system for 8-track recording studios, called Mixmaster. I only sold a handful of them, but it was my first production product, and music has been on the side for me ever since.
In his warehouse in San Antonio, Texas, Tim Jenison plays the viola de gamba he used to furnish his Vermeer room. Photo by Natalie Jenison, © Tim Jenison, Courtesy of Sony Pictures Classics.
This actually is connected to the film then, because your side interests take you to a lot of unexpected places. Where does art fit in?
You know, I've never been a real art nut, but I did take an art history class in high school. One day, my teacher, Mr. Thorpe, was showing slides of a lot of different painters. We got to the Dutch golden age, and I raised my hand and said, "These guys are basically trying to make color photographs, aren't they?"
He took offense to that. He took umbrage. "That's not what this is about! This is about narrative, composition...." It was kind of embarrassing. (Laughs.) I don't know why he reacted so strongly, but I sure do remember it. Maybe this Vermeer film is revenge against Mr. Thorpe, because I set out to prove that Vermeer really was making a kind of color photograph.
It's a long way from high school to where the film picks up, though. What happened in between?
It actually started about 2002. My daughter gave me a book by David Hockney, called Secret Knowledge: Rediscovering the Lost Techniques of the Old Masters. He talks about how art changed very suddenly in the late Renaissance, and became much more photoreal. He theorized that it must have been due to breakthroughs in optics that happened at the same time.
Hockney mentions another book, called Vermeer's Camera: Uncovering the Truth Behind the Masterpieces by Phillip Steadman, and that's what got me going.
One of the pictures in Steadman's books is, side-by-side, the Vermeer painting The Music Lesson, and a miniature tableau that he built for his book, sort of a Barbie doll scale version of the room Vermeer painted.
As I looked at the two, I realized I was seeing something impossible, especially the way the light fell off along the back wall in Vermeer's painting. As a video guy, I know that that's a really hard thing to see. Our retinas filter out that level of detail before it gets to our brain. They try to make a white wall look more uniform than it actually is. That got me thinking: Vermeer must have had some way to observe those paint colors that is more precise than the eyeball.
I was lying in the bathtub one night and it occurred to me that you could this very simply, with a mirror.
Let's back up a second. Why did Hockney make the connection between more realistic painting and better optics? What do optics and painting have to do with each other?
Throughout history, even to pre-historical cave paintings, people have attempted realism -- but somewhere in the 1500s, it started to really, really change. Hockney was mostly thinking about perspective, the way that artists got their shapes and angles correct for the first time.
17th century illustration of camera obscura, a system of lenses and mirrors that projected images that artists could use as a guide for their work.
Hockney realized that with better and better lenses, the camera obscura could project an image that must have been pretty amazing to somebody in the Renaissance. It looked like a video projector. Actually, some people would put on plays outside, then project the images through a lens to an audience sitting indoors, in the dark -- basically, watching television, hundreds of years ago.
So when I look at Vermeer, I don't just see something realistic. I see something impossible.
And now you're thinking about Vermeer and lenses.
Yes. The trick is that you have to match every color, every white, along that wall, on the way to almost pure black. Every spot on the wall has a different color. So Vermeer painted according to the laws of physics, the way that light actually behaves.
Here are two enlargements of sections of the wall seen in The Music Lesson. Note that nearly every point on the wall has a completely different color.
This raises a couple of questions for me. The first is, why this particular Vermeer?
I was mostly interested in the lighting. You could also reconstruct the room based on that painting. It's that good. You know where the light sources are, because you can see the windows on the left side of the room. You only see two of them in the painting, but you see the reflection of a third in the vase.
That meant I could build a room of the right size and shape, I could build windows of the right shape and size, and then I could test my theory about how Vermeer worked, in a space that was virtually identical to the one he worked in.
There's my other question. How do you get from an idea in the bathtub about how Vermeer did it, to actually wanting to build that exact room to prove your point?
My first experiment was to take a black and white photograph -- it happened to be my father-in-law's high school picture -- and prop it up on a table. Then just in front of it, I put down a board to paint on. Then I put a small mirror that was exactly at the midpoint between these two things, at a 45 degree angle. If you get the mirror in exactly the right spot, it's easy to tell when you've made the colors exactly match.
It was my first oil painting ever, and what I ended up with was hardly distinguishable from the original photograph. That's when I said, okay, I'm on to something.
You can see Tim's first painting using his invention, the "comparator mirror" and a photograph of his father-in-law, closer to the finished version. Photo by Luren Jenison, © 2013 Tim Jenison, Courtesy of Sony Pictures Classics.
I started searching for more about this online, because I couldn't believe that somebody hadn't done it -- but apparently, no one has.
So I mentioned this experiment to Penn Jillette as we were having dinner one night, and he said, "What are you going to do with this?" I said, "I want to paint something that looks like a Vermeer. I'll get some Dutch-looking furniture, put it in a room, and see if I can paint from it. Then I'll make a YouTube video about it or something."
"That's a stupid idea," he said. "This could be a real documentary." From that point on, it was Penn's project. It took on a life of its own, and it became far more ambitious.
At this point, you now believe that you've understood the physical mechanism mediating between the room that Vermeer saw, and the one he painted. But now, you transition to wanting to actually do it. To actually paint a Vermeer, the same way you think Vermeer did. How did that happen in your mind?
I thought that if I was right, that it was important. An important part of art history.
It also seemed to me that the way I could prove I was right about how he did it was to end up with something that looks very much like a Vermeer. I could have started with a modern scene, with a TV set in it or whatever. It could be Vermeer-like, but it wouldn't make the same connection that it would if it looked like one of the Vermeers.
So I really liked Penn's idea. I thought that the film would be a great way to get this idea out there, where it would cause more ripples.
How did you go about building the room that this particular painting is set in?
First of all, you can tell what all these things are.
The harpsichord in the painting was made by the Ruckers family, famous Flemish harpsichord builders. These are still in museums. It's decorated with printed paper, with a standardized pattern that you can still get. That chair -- those are still in museums. We found a carpet expert who told us exactly about the one in the painting, a Turkish medallion rug, and we were able to find one that was very similar.
Tim measuring the exact replica of the harpsichord in Vermeer's The Music Lesson that he built to furnish his Vermeer room. Photo by Tim Jenison, © 2013 Tim Jenison. Used with Permission. All Rights Reserved.
The harpsichord and the chair, as well as the window frames, those are kind of unusual. You can't buy them. They would have to be built.
I actually tried to get somebody to build them for me, but everyone flaked out. (Laughter). San Antonio is a fairly large city, but I just couldn't find anyone to do it. I decided that I had to do it myself.
Tim Jenison lathes one of the legs for the furniture he built for his Vermeer room. Note the name badge sewn onto Tim's shirt: Geek. Photo by Tim Jenison, © 2013 Tim Jenison. Used with Permission. All Rights Reserved.
We have this machine shop at NewTek that we use to create prototypes of our products. It includes a metalworking lathe, and a big CNC [ed. note: computer numerical control] milling machine, and I knew how to operate it all. I thought if I could get a computer model of that furniture, I could feed it into the milling machine and make this stuff myself.
It sounded pretty easy, but it was a lot of work, a lot of details. I built the harpsichord, and that Spanish chair with the lion heads, and the window frames. It took about a year to do all that, but by the time I finished, I was basically standing in Vermeer's painting.
So now you sit down to paint.
It's not quite that simple. I thought I could just use a lens to project a camera obscura image onto a white card, and paint the same way that Vermeer did.
In theory, that works. In practice, when I set the machine up in the room, the image was too fuzzy and dark. If I'd used a modern lens, it probably would have been sharp enough, but with a simple lens that Vermeer would have had, it was just too fuzzy. I couldn't see the details on the harpsichord.
A magnified view of the harpsichord in Vermeer's The Music Lesson.
At that point, Teller came interview me, and he said, "You think this is gonna work?" I said, "Well, I think it'll work. I've got this problem, though." He said, "You know, if it doesn't work, it's going to be a very different movie." I said, "We're not going to make this film if it doesn't work," and he said, "Oh yes we are."
He's right. There would still be a story.
One that's extremely embarrassing to me, and I didn't want that. (Laughter)
So I sat there, playing with stuff, stressing over it, and one day, I discovered the solution to it, which was to add a second mirror. Suddenly, everything was bright and clear, and from that point, I was ready to paint.
But at this point, you'd only painted that single black and white painting.
I'd also done some experiments using a color photograph, and I did some experiments with paint brushes, but essentially, yes, my second oil painting.
I thought it was a simple extension to go from my father-in-law picture to a Vermeer, but I really wasn't prepared for the sheer amount of work it was.
You mentioned that you needed to build some 3D models. How did LightWave factor into how you managed to execute this?
LightWave was extremely useful. First I made a model of Vermeer's room. In LightWave you can set up a background image to work against, so I used Vermeer's painting and started building objects on top of that image. I laid out the basic parts of the room -- the floor, the ceiling, the corner -- and then started to lay in some of the basic lighting.
Then I built a model of the legs of the harpischord, these lathe-turned spindles that the harpsichord sits on. I didn't have to make a detailed model of the harpsichord itself, because there's a guy in Scotland who has already done this. He restores harpsichords, and I bought the blue prints for this harpsichord from him.
I found one of those Spanish chairs with the lion heads in Holland, and I took a zillion photographs from all different angles so I could make a 3D model of it. Again, I had the lathe-turned spindles, but the tricky part was the lion heads on the top. They're three or four inches tall, and once I built them in 3D, I had to carve them on the milling machine.
Tim has placed a photograph of Vermeer's The Music Lesson to use as a guide for the scene he is building in NewTek LightWave 3D software. Courtesy of Sony Pictures Classics.
Once I had all the models in the computer, before I even started building the room, I could run all these lighting models, seeing how light affected the room in various ways, simulating angles of the sun. The sun does come directly in, in late summer afternoons, from nominally north-northwest facing windows, in both Vermeer's room and mine.
Direct light from the sky, though, did not give me Vermeer-like results. I discovered that everything worked great once I had diffusers over the windows. This is actually mentioned in painting manuals from Vermeer's time, that you could use oil paper over the windows to give you a really uniform light, kind of like using a lightbox in modern photography. LightWave was the perfect tool for working that out.
I had to learn more about LightWave than I ever wanted to, though. You have to turn on every feature of LightWave to get this accurate. I actually had to build a 24-core Pentium machine so that I could work and not have to wait days for every test render.
But it was fun. I had a lot more confidence going in because I started in LightWave.
Tim Jenison assembles one of the experimental optical devices he built. Photo by Shane F. Kelly, © 2013 High Delft Pictures LLC, Courtesy of Sony Pictures Classics. All Rights Reserved.
You pull this off in ways that are impossible to describe, which is part of what makes the movie such a pleasure. It raises the question though, about the difference between art and technology.
What you've basically done is come up with a technological solution to what could be considered an artistic, or even a philosophical question, which is, What is the nature of ART?
You know, this is a big can of worms.
One of the worms is, do you consider photography to be art? Some people do, some people don't. Some people say that what I've described as Vermeer's method boils him down to some kind of photographer. I don't agree with that, but that's a question.
Another question is, are movies art? Because we certainly use technology to create movies. Movies are technology, but they also incorporate all the arts -- literature, music, so on. I think it's the fine art of our day, and we use any technology we can get our hands on to get the best possible image.
Tim Jenison (right) adjusts the wig on Graham Toms (left) who modeled as the gentleman for Tim's painting. Photo by Shane F. Kelly, © 2013 High Delft Pictures LLC, Courtesy of Sony Pictures Classics. All Rights Reserved.
In Vermeer's day, the fine art would have been oil painting, and we know that artists at that time were striving for maximum realism. They said that they were. They wanted those paintings to look like you were looking through a window. I don't think Vermeer would have been ashamed to use this technology.
Long before Vermeer, artists were using mechanical aids to achieve this. Leonardo talked a lot about them [nearly 200 years earlier]. There were manuals about how to do this in Vermeer's day. We may not have found all the things I think he used in his studio, but he might have kept it a secret, not out of shame, but because there were trade guilds. You had to keep some secrets.
Tim Jenison discovers a mistake in Vermeer's original painting of "The Music Lesson." Photo by Tim Jenison, © 2013 High Delft Pictures LLC, Courtesy of Sony Pictures Classics. All Rights Reserved
If this was actually the way he did it, another reason he might have kept it a secret was that it gave him an advantage over his next door neighbor who didn't know the trick. If Vermeer could get a little better result, it would give him a certain edge in business.
We make the point in the film, maybe not explicitly enough, that what we call art and technology are basically two sides of the same coin, which is creative activity. Writing computer software is a highly creative activity, very inventive, but most people don't think of it that way. They think of it as a lot of arithmetic or something. Columns of strange-looking characters. But if you ask a computer programmer, they're inventing things all the time. It's just creativity. Art and technology.
It struck me that some of the people who had negative things to say about the film weren't really talking about the film. They have a negative reaction to this idea that Vermeer was a copyist. In the context of Toaster and LightWave and TriCaster, though, there isn't an inherent conflict between art and technology. You use tools to create art. Vermeer's mechanical aids didn't replace the "art," or create something other than art, just because of those aids.
If Vermeer worked the way I think he did, it's still not cheating. There are no rules to break in art.
So I have to ask, what's next for you?
Well, right now I'm back to my day job full time, and I'm between obsessions. I have too many hobbies, though, and I'm sure that one of them will get out of control soon.
Along the lines of painting, there's an experiment I'd like to do trying to replicate part of a Carvaggio. He painted long before Vermeer, and it looks to me like Carvaggio was using a simpler version of the machine Vermeer used. I'd like to at least try it out and see how close I can get.
I'm not going to make another movie, though. Too much work. (Laughter.)
The young film geek has been satisfied.
Yep. I got that out of my system. But if I do try painting a Carvaggio, I'm sure I'll roll some tape on it and see what happens.
It's interesting that when I asked you what's next, you went straight to an artistic obsession. Without being reductive, Toaster and LightWave originally sprang from obsessions. I don't want to say "hobbies," but you did actually have these as hobbies at some point when you were younger, before they became products.
True. Somebody once said if you can integrate your fun with your business you've got it made. That's always the way I've looked at it. I hate working, but it doesn't seem like working if you're doing what you want to do.
With Toaster, you created affordable, high-quality tools to be used in a broadcast infrastructure. With TriCaster, you've enabled people to do their own high-quality broadcasting. Those are two really big problems to have solved, so what's the next horizon from a business point of view?
First of all, I don't think they're solved yet. We've come leaps and bounds, but it's still hard to do this. I think we can make it easier and easier, and cheaper and cheaper.
I forget who said it, maybe it was Orson Welles, but somebody said that film isn't a true artform because nobody can afford the tools, whereas all you need is a pencil to write a novel.
The moving image is the most powerful medium of communication. At some conscious or unconscious level, people simply believe their eyes. That's almost too powerful. You certainly don't want that power in the hands of just a few people.
I'm really optimistic about the future of mankind because that power is no longer concentrated. We've taken that power and given it to everybody.
For more information on this remarkable project, visit TimsVermeer.com
HEY, LOOK! TWO BONUS CLIPS
The visit that Tim made to PBS's Computer Chronicles to introduce the NewTek Video Toaster in 1990? Here's the full segment! Needless to say, it looks its age, but it's impossible to overstate just how jaw-dropping this was at the time: a full-on broadcast suite for under $1500. Old-timers will also be delighted to see a Kiki Stockhammer cameo.
explanation as the most magical part of the trick.