A History of LEEP
How We Got Into VR
by Eric Howlett — February 16, 2006
We were in the photography market [in the 1980's], and designed and made a very nice wide-angle stereoscopic photography system consisting of a viewer and a matching camera to make pictures to view. There was a solid invention making it possible, and a patent issued in 1983. We had 70 early orders from stereo enthusiasts (average folks thought "stereo" just meant two sound channels) at a price, after we finally raised it several notches, that should have been fine. Trouble: we couldn't make the cameras. The thing was very complicated — lots of bells and whistles — and devilishly hard to put together. My seat-of-the-pants manufacturing engineering was just not good enough. We couldn't afford the 100 million dollars or so that Polaroid or Kodak would have spent, and I thought I could do it myself. Well, I couldn't. And I wouldn't have tried if I hadn't first failed to interest both of those companies in taking it over. Kodak copped out by saying they weren't sure of the market (See "The Kodak Story," below). Land didn't try to explain, and none of his employees could figure out why he said "No." (See "The Polaroid Story," below.)
After several years we had a camera and viewer designed and tooled, parts and materials for 70 [cameras], completed 20, and finally shipped the only three cameras we could get to work properly. One was stolen and two are still, so far as I know, working, albeit with at least two repairs by ingenious owners. So I had to let the 4-6 people working on it go in '85/'86 because an unrelated product that had been supporting us had dried up and my second mortgage was maxed out. But we could make the viewers. A NASA Ames engineer (McGreevy) heard about it. He came by in '85 and was enchanted (as most people were) by the images he saw in the viewer. He described the NASA "VIEWS" system to me and began ordering viewers. He also advised VPL, whom NASA had been supporting, to use LEEP lenses in the HMDs they were developing. So NASA and VPL became our major customers. Disney came in later. Some stereo buffs bought viewers with sample slides. Ulrich Figge and I shipped, in total, between one and two hundred viewers with six slides, and lens sets, at prices ranging from $840 to $3,500.
Above: NASA VIVED Used the LEEP Optics (1985)
I asked one of the NASA guys on the phone, "You mean to say if I made a system like the ones VPL makes, I could sell it for over $10,000??" He answered to this effect, "Sure. There are lots of academic labs and federal organizations that would snap it up at that price." That led to the original Cyberface — I knew more about electronics than I did about optics. We released the Cyberface in March of 1989. In June, VPL released the Eyephone, which also used our optics. We didn't sell many Cyberfaces — maybe because we were using the same low-resolution monochrome LCD that NASA and VPL were using. (VPL wasn't selling many, either. Their biggest investor, a Frenchman, was getting restive. "...they're just burning money," he said when visiting us in Boston, "..BURNING money.") We recently added a product page to the website that describes the original Cyberface. Have a look for some interesting facts.
With the Cyberface2, the poor quality changed. By (cleverly) diverging the optical axes of the viewer lenses we were able to crowd in two higher resolution color LCDs by Sharp. Though you could still see pixels, they were small enough so you could ignore them and pay attention to the image comprising them. These units sold better than the original Cyberfaces. The Cyberface2 was comfortable if you didn't move around too actively, but the nuisance of putting it on and taking it off limited its use in many applications. It still used the same "Featherweight" counterpoise system of the first Cyberface.
The Cyberface3 was a good solution to the donning and doffing problem for people who worked at a desk or in a limited walk-around. See the Wired Magazine review we linked on our website. We sold a number of these and I believe that with a greater marketing effort (like that of Fakespace), we could have made a business based on them. We could have made good use of a more serious investment at that point, but the venture types were losing interest in VR.
At the end of 1994 we (finally, third try) won a NASA SBIR contract. This one was to produce a head-mounted display system with radio links from a trailer on the ground to matching helicopter-borne cameras. The system had high resolution inserts in the center of the field for both eyes. The object was to make it possible for a pilot on the ground to fly a remotely controlled helicopter for combat or surveillance purposes with a view approximating the one he would have if his head were in the drone. This was a major project that was not completed before the fixed budget ran out. We were unable to find engineers qualified to do the ground trailer for the antenna and pilot station or to design the signal modulators and airborne camera servos.
The First Public Demonstration
by Eric Howlett — February 14, 2006
Above: Eric Howlett shows off the LEEP optics for the first time in 1980 Las Vegas.
In March of 1980 I first showed the LEEP demonstrator publicly at the Photograph Society of America convention in Las Vegas. It received an enthusiastic response and, on the part of those who could understand it, some amazement at the technical facts. But the business people there also had the response that, no matter how fine it was, if it was stereo, there was a very limited market for it, so I shouldn't waste my time. Of course I couldn't accept the idea that such a fine product, offering imagery striking to almost everyone who saw it, could not be marketed if the price were in the consumer range. In fact, part of the value of the invention is that the components of the system were very inexpensive — from an injection molded body to two plastic lens elements smaller than a quarter for each eye of the camera. Fred Spira, founder and president of Spiratone, invited me to his hotel room and waxed enthusiastic about the invention, while also offering his opinion that the stereophotography market would never be revived.
The Beginning of LEEP
by Eric Howlett — February 14, 2006
The design of the custom-molded LEEP camera/viewer lenses (they had to match) was done in 1984. I started with a TI 35 programmable calculator and a college optics textbook. While incredibly fast and accurate compared to the time-honored way of computing lenses with log tables and a mechanical adding machine, the TI didn't have enough program memory to complete the calculation of one single ray through the six surfaces of the viewer magnifier. A second program had to be entered by hand to complete the calculation and bring each ray to its final destination. While I gained a little insight into what the lenses should look like, it was impossibly time-consuming to investigate a reasonable set of lens shapes.
There were lens design programs for minicomputers that would have made short work of it, but I could afford neither the programs ($5K) or the computers ($11K) and so bought a Sinclair ZX80 for a few hundred dollars. This little machine was inseparable from its operating system and Basic program, both of which lived on chips inside. The monitor was a 14 inch black-and-white TV. I found the input to the video amplifier of the TV and hooked that up to the Sinclair video output. There was a membrane keyboard that forced me to type with one finger. If I wanted to save some code, I hooked up an audio cassette recorder with a “data-certified” cassette and waited a while. I soon wanted an additional 16K memory that they said would be available in a few weeks, so I split for an additional $160 and ordered it. It took a few months to arrive, and I gave up waiting. A kind computer consultant (Bob Schuldenfrei) was also browsing at our local Radio Shack, and, warming to my story, guided me to a TRS-80 running DOS Plus. The thing was fast, a good keyboard and two floppy drives, so you could make backups and and saved at lightning speed. And 48K! Infinite memory! I plugged a dot matrix printer into the parallel port and it worked immediately. I wrote a letter (in Basic) of thanks to the guy who had sold me the printer.
Now I was ready to roll. I asked Ulrich Figge, who had been working with me in on the LEEP stereophoto system for a couple of years, if he thought we should stick with the available-lens viewers we had been using or push the design of a special magnifier as far as it could be pushed. He opined that we should push the design as far as we could. It was years before I came to realize how right that answer was. I said he might have to mind the store for a few weeks. He said OK.
I don't recall clearly how the next six weeks was divided up. But I was full-time at home night and day writing and tweaking a Basic ray-tracing program and improving my understanding of, and the performance of, a novel lens design with a large eye lens that promised a very wide field of view with adequate flatness of field over ± 70° measured at the center of the eyeball. You saw most of the field without moving your head to “peer” from side to side. (The program couldn't handle peering, and that led to a good design for HMD applications.) The program would trace four or five rays at increasing angles through the four or six surfaces of the lens and print out x's on the dot matrix printer showing the shape of the radial and tangential focal planes of the lens. All by itself — hands free! But it took 20 minutes to do that, so I had a great many 20 minute naps over the next six weeks. At the end of it. I had cartons full of rolls of perforated fan-fold printer paper, and a design that I was willing to risk paying for injection molds to try out, so we ran with it.
That was my programming career. It began, peaked, and ended in six weeks. But even now, I'm proud of what I managed to do with the means at hand. It also explains my admiration for what you do.
An unrepentant collector, I still have that TI 35, that Sinclair and two Trash 80's. Alex and his brother Ian first got their first taste of real computers on them.
Above: LEEP Photo of Alex and Ian
When we got the first batch of lenses back from the molds made according to those specifications though, we had a shock. The eye lens was just too powerful, and not quite large enough to allow a casual movement of the head at a comfortable distance from the first surface of the eye lens. (The program didn't do that) We were supposed to show the thing at the Museum of Science in Boston for a period of several months starting in a couple of weeks. Pure luck saved us. There was in our grab bag of lenses some glass ones that might work, and in fact one did — very well. Six of them (or maybe four) were pressed into service for the Museum of Science exhibit. A new mold for the eye lens based on what we had just learned was ordered. The magnifier housing was redesigned and the lenses and the magnifiers we made with that eye lens over the next several years were remarkably good. They became the standard of the virtual-reality industry. As we got into LCD displays for VR systems, Rend 386 and Sense 8 came along and with IBM 486's gave nice computer-generated worlds. C++ ruled. But I wondered where it would go. Images from cameras with LEEP format lenses — giving "telepresence" — were always better. The wide FOV was hard to fill with a computer unless the shapes were pretty simple and boring. Most users of the viewer lenses didn't even bother to format the image correctly — leading to all sorts of mythology about “distortion”.
The Kodak Story
by Eric Howlett — February 14, 2006
Of course I wasn't ready to except Spira's opinion, and upon returning home. I started badgering Eastman Kodak people for an opportunity to demonstrate it to the right people there. After time, I got in touch with a man who said he could assemble eight or ten photographic engineers, many of whom were also stereo enthusiasts, to look at my invention. I took the demo to Rochester and was warmly received, the demonstrator was thoroughly admired, and a number of other engineers were called in to look. My contact asked me if I would leave the demonstrator so that they could evaluate it further, especially so that the boss of them all could see it. (I think he was on a business trip.) Naturally, I happily obliged, and had them sign a receipt for it.
(There was a special reason for approaching Kodak first. Kodak made all the 70mm motion picture release print film made in this country. That film would have been ideal for the LEEP format, except for the perforations, which made the usable vertical size smaller than regular 120 camera film. But they wouldn't lift the perforators for less than a carload lot, at an astronomical price.)
I returned to Massachusetts, waited a week or 10 days and called my contact to ask what was happening. He said that the boss — a very senior optical and photographic engineer — did not think it worked. Even after he looked at it he believed it was not doing what I said it was doing, my contact said, but don't despair yet, the younger engineers are working on him. I then waited about three or four weeks until receiving a call from someone in their "Submitted Ideas Department" (!) who congratulated me on the demonstration of the invention, saying that everybody had finally been convinced that it worked as I said. Then she asked, astonishingly, why I had brought it to Rochester to demonstrate. After a gulp or two I responded that I was hoping to persuade Kodak to manufacture and market the product under exclusive license from me. She said effectively, "Good. I was hoping that was the case. The next step is to run it by marketing." After that conversation, "elation" does not begin to describe my feelings. Over the following weeks, about six of them, I think, my excitement began to be matched by anxiety. Finally, there was a letter in the mail, rather brief and formal, thanking me for submitting my idea to Kodak and declaring that it was the opinion of the company that there was no longer a ready market for stereo photographic products, and that they would therefore have to decline to take me up on my offer. The demo was returned in good condition shortly thereafter. I suspected that the real reason was that the effort they were putting into the disc camera — a product whose benefits to the customer were hard to discern — might be straining their product development resources, and that might have influenced the decision.
Well, the hell with Kodak, I thought. Polaroid is closer, anyway.
The Polaroid Story
by Eric Howlett — February 14, 2006
A much shorter story. I took the demonstrator into a development laboratory that was across town from their main offices and showed it to engineers working on an instant movie system, who were also charged with looking at ideas people brought in. The first to look at it called a few more and they called a few more, and so on until in the course of a few hours probably twenty engineers and marketing people had seen it. They asked me if I could leave it, and of course I agreed, as with Kodak. That was a Tuesday. On Thursday I called my first contact there and he said he had sent the demo across town so that Dr. Land could look at it, but he didn't know what Land had to say about it yet. Call tomorrow. I called the next day and learned that Land had said "No." No explanation. Dr. Land didn't have to explain anything to anyone. But it only took four days, compared to about four months at Kodak.
(An ironic and graceful note was struck a few years later at the Boston Museum of Science, where a LEEP viewer and slides were on display and I was fixing a lens that had been knocked out of one of the viewer heads. The room guard told me that Land had been by the day before with the museum director in tow and that Land had read them the riot act for letting the damage occur.)
The Inventor's Folly
by Eric Howlett — February 14, 2006
After Polaroid I was comparing myself to Chester Carlson as he trekked around the country trying to interest people in some cockamamie thing called "xerography" that his company, "Haloid" or "Haloid Xerox" was developing. At least mine was a lot easier to make in a practical form. That was when I made the fateful decision to do it myself. That decision wasn't wrong. My proprietorship, "Pop-Optix" had enough cash flow from some optical devices and a "computer overheating" shut-down system to keep a small operation going. What was wrong was the bell-and-whistle nifty features I kept adding to "justify" the price. I failed to appreciate that a photo fan's first LEEP pictures would justify the price. Fancy features could have waited.
Ulrich Figge, a top stereo photographer, joined me, and Lennie Polucci, an incredibly talented auto-didact, and Greg Rivera, an ingenious artist. In about two years we were making viewers, we had parts and materials for 70 cameras, completed 20, and finally shipped the only three we could get to work properly. One was stolen and two are still, so far as I know, working, albeit with at least two repairs by ingenious owners. That's the end of the LEEP stereo photography story, except for sales of some viewers with slides. After that came NASA, who persuaded us to do it in video, and the Cyberfaces were born.