Epoch and the Cassette Vision (1997)

I. Masayuki Horie and Epoch

—To start things off, can you give us a rundown of the different consoles you worked on at Epoch?

Horie: I joined Epoch because they were a company where I’d be able to develop table tennis (pong) games. For the first year I worked in sales. Then the block-breaking craze got underway, and I was assigned to game development. The first thing I worked on was the Digit-Com 9 LSI (large-scale integrated circuit) baseball game. After designing this and helping out with the basic layout of the circuitry, I moved over to console game development.

Epoch’s second creation was the System 10 console. They were designing it while I was working sales, and once I moved to the console team, I worked on Terebi Yakyuu Ge-mu (TV Baseball). It sure looks primitive today. (laughs) I designed it. (laughs) That was my third game at Epoch. By the time Epoch started making TV Vader, I was no longer doing the programming myself; I was now in a supervisory role. Same with the Cassette Vision. On the Super Cassette Vision I did two games, though.

—Why did you take a step back from software development after TV Vader?

Horie: I was designing other stuff, PakPak Man and our other handheld games. Starting in 1982, I worked on the Super Cassette Vision hardware.

—Wow, you designed PakPak Man?! You can still find people on messageboards online today who will tell you that PakPak Man was the best LCD Pac-man adaptation ever made.

Horie: I put my heart into that one. (laughs) It sold nearly 3 million units. I made two more variations on that game, but PakPak Man was still the most fun. (laughs) Did you know there’s a pellet under the bridge?

—I did. (laughs) If you forget to get that one, you get all nervous, “Huh? I thought I got everything?” and then you have to hurry up and grab it! (laughs) Horie, what kind of games did you play in your childhood?

Horie: I played those baseball pinball toy games. An older classmate of mine had one at his house, but he didn’t have the ball, so we cut off the top of a little globe attached to some tongs and made our own makeshift ball.

—From metal tongs? Wouldn’t that be steel though? How did you cut that?

—You graduated with an engineering degree; what did the other engineering students of the time think about video games?

Horie: I actually have no idea what they thought.

—Back then “pong” and “video games” were pretty much synonymous, but did you have a sense of what different kinds of games would be possible in the future, or what kind of game(s) you’d like to make yourself?

Horie: At first, I just knew I wanted to make video games, rather than having a specific game in mind that I wanted to create. At my interview with Epoch I was asked why I wanted to work there. “Epoch made a TV Tennis game, and I want to create video games,” I said. Then the interviewer, with the President sitting beside him, responded, “Oh, we’ve stopped making those.” When I heard those words, my heart skipped a beat. (laughs) They asked if I wouldn’t mind working in Sales/Admin for a few years, and since I was determined to join Epoch, I said yes—and true to their word, they really put me in Sales!

In a way, though, I was lucky. I wasn’t told this during the interview, but they were actually right in the middle of developing the System 10. It was released in 78, and I helped with the sales then too. The thing is, you see, no one had ever connected a video game to their TV before, or knew exactly how that would work: not Epoch, not our customers, not the industry. Today we have video inputs but before that you had to use the switchbox and turn to channel 2, where the signal would be broadcast wirelessly. We all know that’s how it works today, but at the time we were gripped with anxiety on a daily basis: what if consumers turn it on, and nothing shows up on-screen…?

In an era before the Famicom, therefore, I was actually breaking new ground in my job. We ran an amazing info campaign to help people understand how these new games worked. I was working out of the Osaka sales office then, but if a customer was having problems, we’d almost always go to their home directly. It was too difficult to explain how to fix it over the phone. We worked in the area from Osaka to Wakayama. There’s so many nice people in Kansai. Talking to them on the phone I’d sometimes feel a little scared, but the people I visited were very kind. They would even treat me with beer and dinner sometimes!

Old television screens were smaller than they are now and the corners were rounded off, so lots of times the entire game wouldn’t fit on screen. I often had to open up the TV and adjust the horizontal and vertical pots. I’d have to get back behind the TV, use a vacuum to get all the dust off, make my adjustments and then move everything back in place. Nowadays we’d call it an occupational hazard. (laughs)

That Osaka sales job was the best place to learn those technical things, though. I had to write a lot of reports!

II. The Era of TV Tennis and the System 10

—TV Tennis wasn’t actually manufactured by Epoch, correct?

Horie: I believe that’s right. I’m not sure who made the original. It’s different from NEC’s model.

—What was the response to TV Tennis like at Epoch?

Horie: We thought it was groundbreaking.

—I see. There’s a two year gap between TV Tennis and Epoch’s next console; I imagine that time was spent developing the System 10?

Horie: Yeah. TV Tennis had been made entirely from discretely linked TTL chips. The System 10 used hardwired logic chips, not a microprocessor; it was just one big integrated circuit made from TTL chips. When you were making a new game, you didn’t compile a program, you engineered a circuit, which was extremely challenging.

—I disassembled the System 10 once to look at the inside, and I saw it was an LSI (large-scale integrated circuit) game.

Horie: Yeah, it wasn’t until TV Baseball that we finally started using microprocessors. Everything, including the video, ran off one chip.

—I know you didn’t participate in the System 10 development, but what was your impression of it?

Horie: I thought it was very interesting. As an employee, too. I played it at home a lot. When my family, the four of us all got together and played it, we had a really fun time. I played it with my friends, and at the office too—non-stop squash.

—The System 10 had a lightgun peripheral, too.

Horie: There’s a funny story about this. The trigger switch on the prototype was experiencing some kind of electrical interference. When you squeezed the trigger three times, instead of registering each shot individually, it was firing three shots at once. (laughs) Autofire! (laughs) I quickly added a condenser and got it working right again.

—Sounds like the Saturn Virtua Gun. (laughs)

III. Advent of the Microprocessor

—With 1978’s TV Baseball, you were finally doing game development. You also designed the hardware. I imagine that process involved going to Akihabara or somewhere similar and buying a bunch of components to then put together back at the office?

Horie: No, the hardware construction was mostly pre-determined by other factors. I did design the external layout, took measurements, etc. Then I would have designers render that, and share the blueprints with everyone at a meeting.

—Was it NEC’s idea to use the microprocessor?

Horie: You could say that. The System 10 had used hardwired (discrete) logic, and was released in September. By that point we had already independently begun developing its successor, the “Super 10.” It was an upgraded version of the System 10, and it was nearly complete—that’s where the microprocessor came in.

Basically, everytime we wanted to make a game using discrete logic chips, it would take us nearly 2 years to complete, and it was extremely difficult too. On a software system we’d be able to make more games more quickly, and we saw a growing demand for home console video games, so we switched to the microprocessor. The design was done by NEC, of course.

—What were some of the new possibilities open to you now that you’d switched to software development, where games could be “programmed”?

Horie: Hmm… well, we could do diagonal lines now.

—Oh, like the foul line in TV Baseball! Or the back of the PakPak Man character.

—(laughs) So that’s why his back has those spaces there! (draws picture) How about sound? What was it like back then?

Horie: Simple monotone. I think it got better once we started using microprocessors, though—once you could program in a counter, you could play different musical intervals. I think the System 10 could make two sounds? A “pi pi pi!” and “po po po,” basically a high beep and a low beep. It was fixed, but with the microprocessor we could actually make different melodies.

IV. Developing Games at Epoch

—How many people were on the development team then?

Horie: When I first joined the console group, it started out as me, my boss, and one other person. Just the three of us. The next year, when the digital games department was officially established, we added a lot more people. They worked on PakPak Man, and other games that used VFDs (vacuum fluorescent displays).

—Were roles divided up, like one person in charge of game design, etc?

Horie: People worked on their own different games… that is to say, it wasn’t “game design” in the same way it is today. In TV Baseball, for example, we spent a great deal of time and energy just figuring out how to make basic rules that would be playable: how should we handle fouls, home runs, etc. Just the simple rules of “3 outs, then change sides” was difficult because of the counters required. (laughs)

—Nowadays there’s planning meetings, voluminous planning binders, elaborate presentations involving multiple people… what was the planning like for TV Baseball?

Horie: You silently worked on the design specs by yourself.

—So you’re telling me that if I opened the door to the Epoch game design room, I’d just see a bunch of guys with their heads down working quietly? (laughs)

Horie: That’s right. First you had to do your research, gather your materials. I knew the basic rules for baseball, but there were some things I didn’t know too.

For the programming, we had something similar to exception handling, so I spent a lot of time doing analysis with that. You had to do graphics yourself too—there were no graphic designers, and of course no graphics tools, so everything had to be programmed in by hand. For our planning docs we would use tracing paper that was a little bigger than normal graph paper. We’d draw a square representing the game screen and then figure out what would happen in different situations. By the latter half of the TV Baseball development, we added more staff.

—But everyone still worked quietly at their desks?

Horie: Yeah. The thing is, even if you have a great game design conversation with someone, you’ve still got to condense and translate those ideas into an actual game. I don’t think much gets worked out by discussion alone to be honest. So many of those ideas where you think “that’s great!”…well, when the programmers look at them later, they’re full of inconsistencies and contradictions. I think it’s best to do the research on your own first to see if your idea is feasible, and how it could be fun in a game.

This kind of situation happened a lot when we were doing Famicom development later. Someone (from outside) would be all excited about some game idea, but when we developers heard it, it was like “what the hell are they talking about?” Again, the ideas were just full of contradictions. Yes, it all sounds good when you’re talking about it, but I think a lot of those game design ideas should come after a fundamental base has been built. Start with the planning docs… is my thinking, at least. (laughs)

—You said “contradictions”, but could you give some concrete examples of what you mean by that?

Horie: In the early days there weren’t many, because the systems we were working with weren’t very big. The problem then wasn’t contradictory design ideas so much as it was the massive restrictions we were working with. The screen was made entirely of individual sprites, about 21 in total. It was extremely difficult working on the TV Vader game with all these restrictions.

—So after figuring out the design for a game, you’d then talk to NEC and have them make the chips for you?

Horie: Yeah. At this point, we were not yet able to do the programming ourselves—not to the point of a finished program. We did after this, but TV Vader was a special case. NEC had some mock-up chips lying around that we used, and they could only be encoded over at NEC.

—My impression of game developers is that they work late nights everyday. How about at Epoch?

Horie: Not on software development, no.

—About how long did a development take you?

Horie: Hmm, how long indeed. TV Baseball took somewhere between 6-12 months, I believe. We had the prototyping tool to rely on too. We learned a lot from creating these LSI games. Our later developments took 4-6 months, I think? I can’t remember exactly.

—And what kind of development tools did you have?

Horie: We used something we got from NEC, basically a prototyping machine. Nowadays you’d just write to P-ROM and be able to debug right away.

—Was there only one prototyping machine?

Horie: Yeah. It was huge, about this big (gestures to a large cardboard box). For a terminal it used NEC’s TK-80 computer. There was a separate tape reader attached, and once you inserted your tape it would execute. What was loaded into the RAM there corresponded to what we’d burn onto ROM chips for the actual games.

—So the process went something like: write the code out, then convert it to assembler by hand, feed it to the TK-80 via tape and prototyping machine, and then finally you’d have something displayed on screen?

Horie: Yeah, that’s basically how it was.

V. The Path to the Cassette Vision

—Epoch is known for their Baseball Pinball games. Was the thinking, “hey if we can make these, I bet we can make a good baseball video game too” ?

Horie: Yeah. And it turned to be very fun when it was finished. There were no other baseball video games like it out at the time. At the end of the year Nintendo released their Racing 112 game. We had to decide whether to make TV Baseball or try and compete with them with our own racing game.

—I remember that people were gradually starting to get tired of pong tennis games. Manufacturers of such games had lots of inventory, and I’ve heard that they had to sell them to wholesalers. In that market, to think of developing new, expensive hardware… but I guess Epoch was very invested aleady. I suppose retailers and developers must have seen the end of the boom as merely the end of the pong boom, that a new game would sell just as well again.

Horie: I too felt that these games (gestures to pong paddles on Cassette Vision) were all the same, and I’d become bored of them personally. And I thought new software would change things.

—Your next game was “TV Block.” This one used an Atari chip.

Horie: At this time Atari had two kinds of chips. We imported them and assembled the circuit ourselves.

—Before TV Block, some Japanese makers were importing and selling Atari’s Video Pinball. It played sound from the console unit, though.

Horie: They probably didn’t have a voice modulator.

—TV Block was your flagship product at the end of 1979.

—Was it difficult to make the ball move like that?

Horie: Not too hard, I think. TV Baseball used 4×4 sprites, and they moved on an x,y grid. To make the parabolic movement of the ball in TV Block, we needed to make movements in smaller denominations, in single pixels and scanlines. By TV Block we were able to do that, though it’s now common practice.

—You also released the “System 10 M2” at this time. I originally thought it was just a lower priced version of the System 10, but it actually has a different ball serve speed! That suprised me. I assume a simple re-issue of the System 10 would have been cheaper, so why make that change?

Horie: Was it different? Did we change it? I’ve completely forgotten. Now that you mention it, though, I do remember we had some things we wanted to fix from the first System 10. The interval between serves was too long, and the score was too big and goofy looking. Of course we thought it looked cool (and was a selling point) to have large numbers then, to emphasize how novel the System 10 was.

—Next came the Atari VCS. This was imported by Epoch and rebranded as the “Cassette TV Game.”

Horie: The VCS was really expensive.

—I wanted to play the Invaders game on the VCS soooo bad back then.

Horie: The VCS Invaders game wasn’t part of the original Atari console; it was made at our request. The fact that a big American company like Atari listened to our request like that was, in and of itself, groundbreaking. I was amazed, and when the chip arrived in my hand, it was a very moving moment for me.

—That’s a great story. Without Invaders, the VCS probably wouldn’t have made it in Japan.

Horie: You might be right about that. Every other Atari VCS game that we imported didn’t do very well in Japan. For some reason Superman was an exception.

—Moving into the 80s, Epoch released TV Vader. Finally, I could buy a copy of Space Invaders for myself! (laughs) My copy went “on-tour” as I took it to all my different friend’s houses. (laughs) Everyone was crazy about it, although it only had 8 invader enemies on-screen at a time.

Horie: That was due to the sprite limits. The architecture at this time wasn’t so much about the number of sprites you could have on an individual line at once (like the Famicom), but more simply about how many you could have on-screen at once.

It was a real challenge. That’s why we had that excuse, that there were “invisible invaders” hiding behind those first 8. (laughs)

—I remember that. (laughs) Maybe that wasn’t the best decision, in retrospect. (laughs) Maybe if you had said they were retreating when they get hit by the bullets, it would have gone over better. To a kid’s mind, I mean. (laughs) The illustration on the backside of the box showed 48 invaders, so it was a bitter pill to swallow. (laughs)

—And were you, too, obsessed with Space Invaders?

Horie: I once blew through 100,000 yen (about $1000) in a single month on Space Invaders. (laughs)

VI. The Cassette Vision

—Finally, then, we get to the Cassette Vision. It was an era where cartridge-based systems were starting to become commonplace; was this factor the starting point for the development of the Cassette Vision?

Horie: I think so. The single-game consoles were so big, they took up a lot of space in a home if you owned more than one.

—Did you take any hints from the Atari VCS?

Horie: In a sales sense, yes. If it sold well, then it meant people would buy game cartridges.

—As an aside, did you know that the Cassette Vision wasn’t actually the first cartridge-based system in Japan?

Horie: What, really? It wasn’t?

—There was the Video Cassette Rock and TV Jack 5000 Add On. They used LSI game cartridges. The Video Cassette Rock was by Takatoku, and the TV-Jack 5000 was released by a GL, a merger of Bandai and Tilt.

Horie: The games weren’t made in Japan though. (laughs)

—Well… (laughs) But the Cassette Vision, when looked at from a hardware perspective, isn’t particularly exciting itself. The Cassette Vision was a “1 chip” computer, but didn’t you think about creating a genuine microcomputer system, with separate ROM and RAM?

Horie: Part of it was that we had released similar machines before the Cassette Vision, and we wanted to capitalize on our knowledge and experience there. And unfortunately, we weren’t able to separate the ROM and CPU. The LSI chips used in console games at the time were 1-chip systems, and with our technology of the time we weren’t able to use a bus to add separate ROM.

The Cassette Vision had a lot of bits (48 bits). So it was already as fast as we needed it. With one instruction we could quickly jump to addresses, perform other instructions, or read the contents of a register. Had we added an external ROM, we would have had to use 8-bit ROMs, which would have been 4x slower.

There was another thing we were worried about: the ability of the bus itself, when you were dealing with external ROM. Not only was it very expensive back then, but the bus was also very weak and prone to noise, which would lead to errors. There was a big difference, you see, between a 1-chip system in which all the electricity is flowing within, and a system where the parts are separated by over 10cm.

It was because of this design approach—re-using our old tecnology and expertise—that we didn’t switch to a real computer system until the Super Cassette Vision. By that time we’d reached the limit of what was possible with the Cassette Vision, and though we had wanted to make something that used better hardware earlier, it would have meant stopping the Cassette Vision early. So we decided to let the Cassette Vision run its course, and then let the Super Cassette Vision take center stage.

The Cassette Vision had been doing quite well in sales, you see. We didn’t really have any rivals. Well, until the Famicom. (laughs) So we couldn’t just stop selling it. We were doing a lot of research on its successor during that time, though.

—Interesting. So would you say it was one really one long chain, from TV Baseball to the Cassette Vision?

Horie: Yeah. The Cassette Vision, in its own way, represented the perfection or maturation of that product.

—When people talk about Cassette Vision games, the first name that gets brought up is Kikori no Yosaku. Were you aware of the arcade game, SNK’s Yosaku?

Horie: Of course. (laughs) We went to all the AM Shows (“Amusement Machine” Shows, aka arcade expos) in those days. Nowadays they’re filled with nothing but otaku-ish people, but back then none of the attendees were players. No one played—seriously, it was only industry types, game center operators and the like. We would decide on what the next game would be by watching how young people play, but we ourselves were not players. So going to the expos, we’d just have conversations with each other, like “what do you think will be popular this year? Crazy Climber, maybe?”

—The Cassette Vision version of Galaxian felt closer to Moon Cresta though, don’t you think?

Horie: No doubt that was down to the peculiarities of the individual developer. Some guys were making some pretty weird games…

—Oh? So the Cassette Version games were all made by different people?

Horie: Let me think… the person who made Astro Command also made the Space Invaders handheld and Nebula for the Super Cassette Vision.

—”Baseball” was the same game as the earlier TV Baseball; you made both games. Battle Vader, likewise, was a port of TV Vader. As for Big Sports 12, I’m going to guess that this originally must have been the game intended for the Super 10?

Horie: Correct.

—Hah, I solved it! (laughs) Well then, from Battle Vader to PakPak Monster, there’s a bit of a gap in releases. Were you developing something else then?

Horie: Actually, because we only had that one prototype machine, we had to make one game at a time.

—Ah, so even though Yosaku and Galaxian came out at nearly the same time, that must have been because you’d created them earlier, and had them finished and prepared for release already?

Horie: That’s right.

—PakPak Man had to be controlled with the four buttons that were laid out horizontally on the Cassette Vision. It was really hard until you committed those controls to muscle memory. (laughs) Next was New Baseball. By the way, did you develop all these games in the same order that they were released?

Horie: That’s right. New Baseball went very quickly because we were able to use our know-how from the previous game. By adding more rules, we were able to make a sequel that took up roughly the same amount of memory and did more of the things we had originally wanted to do.

—You weren’t part of the New Baseball development though, right?

Horie: Yes.

—How about Monster Mansion? Was this meant to be a Donkey Kong clone?

Horie: It was a combination of that and our earlier LSI game, Monster Panic.

—Then there’s Astro Command, which had scrolling despite the lack of such capabilities in the Cassette Vision hardware.

Horie: This was all done with sprites. There was no hardware background layer with the Cassette Vision; you simply moved the x,y coordinates of the sprites.

—(laughs) I do know that Grand Champion was never released because of all the bugs… but I think enough time has passed that its tale can be told. (laughs) What were these bugs, specifically?

Horie: I’ve forgotten now. It was things like you couldn’t make a goal, or the game would keep going after you made a goal. We discovered them after the game was finished. We eventually fixed them, but by then the Cassette Vision was already nearing the end of its lifespan.

—Next is Monster Block. For the longest time I thought this was your creation, since it resembles Sokoban. You’ve said before that you ported Boulder Dash to the Super Cassette Vision because you thought it was a fun game. I get the sense you’re a fan of puzzle games?

Horie: This was actually done by the same person who made PakPak Monster and Elevator Panic.

—With Elevator Panic the carts changed color and became cheaper. Was this to make room for the Super Cassette Vision?

Horie: I think so. The Super Cassette Vision games were very different, so we wanted to distinguish them in both price and appearance.

—Were you thinking you’d keep making Cassette Vision games for awhile yet, then?

Horie: Yes, but unfortunately, we were at the limits of what could be done with the prototyping machine. We only had one unit. It took a lot of time to use, and money to maintain. Simply checking whether the prototype machine was working correctly or not was an annoying task in and of itself.

VII. Looking Back at It All

—The Cassette Vision ended up selling around 400,000 units in its lifetime, I believe. As the #1 seller before the Famicom, it was a big success.

Horie: Yeah, for our company, too.

—Unfortunately the Cassette Vision often gets dismissively summarized as “that console that lost to the Famicom.”

Horie: The thing is, we weren’t trying to compete with them. The Cassette Vision was a success in its own right: we paved the way in showing consumers the appeal of a home console system that you hook up to your TV—”this is how you do it.” (laughs) So when the Famicom went on sale, that aspect was already taken care of for them.

—And yet the media barely ever reports on the Cassette Vision. This is just my personal theory, but I think there’s three reasons for that:

1) they expect that the makers of a machine that lost to the Famicom wouldn’t want to talk about it,
2) readers today wouldn’t be interested, and
3) “old games are boring.”

The first is just narrowminded. As for readers being uninterested, there is a concern that it would be too niche for most readers, but the truth is that people who have written about the Cassette Vision previously have done a very superficial job, and the articles come off like “wow, they sure worked hard back then”—and obviously that isn’t very interesting to readers.

The third reason, though, I think I understand. It’s sort of like the post-war situation and rationing. The generation that lived through that would say “Don’t be wasteful. When I was your age, we didn’t have anything to eat—we had to eat locusts!” But something about those admonitions doesn’t reach young people today. They’re like, “That’s nice Grandpa, but if you’d had pizza to eat back then, I bet you’d have eaten it, right?!”

It’s similar with classic games. With Space Invaders we all wanted to play “the real thing”—an arcade perfect version—but we had to endure and make do with TV Vader. And no one wants to relive those hard times. I think that accounts for the fact that even those who experienced the Cassette Vision firsthand are content to ignore it.

That’s why I wanted to use this interview as a chance to shed some light on the circumstances of this period. And I was especially interested in the amazing efforts of Epoch.

Horie: Well, to be honest, we weren’t working that hard. It was a long, steady effort.

—I see. I did get that impression from our conversation today. In the end, I guess the Cassette Vision was a thing of its time. It shouldn’t be put on a pedestal, but it also shouldn’t be denigrated unjustly.

Horie: You know, the last time I was interviewed about the Cassette Vision, it gave me a lot to think about. I wondered, why has there been a resurgence of interest in these machines recently? Personally, even though it was hardware I worked on myself, when it came time to move on, I left it behind me. (laughs) I wonder how it would be if I showed these games to my kids today? They probably would just say “Wow, you made this Dad? Amazing.” But I don’t think it would leave an impression beyond that.

On that note, when I was young I loved the kid’s magazine “Kagaku to Gakushuu” (“Science and Learning”). I read it religiously as a grade schooler. Back then, hobbyist components such as mini light bulbs and chemicals were very hard to come by, so when I finally got ahold of them, I was really happy. But for my kids today, if you give them an electronics kit with a battery and mini light bulbs, they aren’t going to be very excited.

That’s what I think, when I look back on those days. I think it’s similar for you and your memories of the Cassette Vision—in a sense, those childhood memories are imprinted on us. I mean, when I see a mini light bulb I still get excited: “ah, it would be fun to run a current through that and see it light up!”

So yeah, when you look at the capabilities of game hardware today, like the Playstation and N64, it’s just an order of magnitude different. Still, my memories of how much I enjoyed those old days remain.

—One thing I found interesting recently, I was reading the results of a video game survey we gave young people—those who never experienced the Cassette Vision and similar technology. They didn’t want to go back to that old technology, but they definitely showed an interest in these games.

Horie: I’m not surprised. These are our roots, after all.

If you've enjoyed reading this interview and would like to be able to vote each month on what I translate, please consider supporting me on Patreon! I can't do it without your help!