A light gun cannot see a picture. Inside the barrel is a single light sensor behind a lens, and all it can report is whether the spot of screen it is pointed at is bright. Yet from the late 1960s to the end of the 1990s, games used that one signal to tell which duck had been shot, which gangster had been hit and exactly where on the screen the player was aiming. They managed it because a cathode-ray tube television never shows the whole picture at once: it draws it one line at a time, fifty or sixty times a second. A game that controls the picture, and knows when each part of it is drawn, can work out a great deal from one flash of light.
Before video games
Guns that fired light at targets came first. The arcade maker Seeburg sold a shooting game called the Ray-O-Lite in 1936. The first video light gun came out of Ralph Baer’s work at Sanders Associates in the 1960s. His team built a “light pen” for quiz games, then turned it into a gun, and by June 1967 their TV Game Unit 2 had a target-shooting game played with two of them. When Baer’s design became the Magnavox Odyssey in 1972, the gun was the only accessory ever made for its expansion port. In the Odyssey’s Shooting Gallery, the player fired at a square of light bouncing across an overlay on the television screen.
How a CRT light gun finds its target
Every CRT light gun works on the same fact: a spot on the screen glows brightly for a moment when the electron beam passes over it, and fades soon after. Games used that fact in two main ways.
Flash each target in turn
Nintendo’s Zapper for the NES reports only two things, the trigger and whether it sees light. The NESdev Wiki, the community reference for NES hardware, sets out the sequence games use:
- When the trigger is pulled, the game first checks, during the vertical blank between frames, that the gun is not seeing light. A gun that sees light then is pointed at a lamp or a window, not the screen.
- For each target that could have been hit, the game shows a black screen with a white box where that target is, and watches the light sensor for the whole frame.
- If the sensor fires, that target was hit. If not, the game tries the next one.
So a game with three ducks on screen needs up to three frames to find which one was hit. A game can save time by first flashing the whole screen white and timing when the sensor fires, which shows roughly how far down the screen the gun is pointing, and then skipping targets that cannot match. This is how Duck Hunt works; the Zapper entry has the detail.
Latch the beam position
Sega built the other approach into the Master System. Its Light Phaser, released in 1986, connects its sensor to a controller-port pin called TH. When TH changes, the console’s video chip latches its horizontal counter, which records how far along the current line the beam has reached, while a vertical counter gives the line. As SMS Power!’s documentation puts it, “the HCounter is automatically latched when either controller’s TH line is active”. The game draws a bright shape, reads the two counters when the gun sees it, and has the aim as an X and Y position rather than a yes-or-no answer. SMS Power! advises games to average several readings to find the centre of the shape.
Namco’s GunCon for the PlayStation, which arrived with Time Crisis in Japan in 1997, took the counting into the gun itself. A lead from the gun plugs into the console’s video output, so the gun can see the television’s line and frame sync pulses. It counts from them to the moment its sensor fires and reports the result to the console as coordinates: the horizontal position in 8 MHz clock ticks since the start of the line, and the vertical position in lines since the start of the frame.
Light guns at home
- Nintendo Zapper, sold with the NES from its launch. Duck Hunt and Wild Gunman made it famous. The first guns were grey, and Nintendo later relaunched it in grey and orange.
- Sega Light Phaser, 1986, sold only outside Japan; the Japanese Mark III could not use it. It worked with Safari Hunt, Shooting Gallery and Operation Wolf among others, and Mark Cerny’s Missile Defense 3-D needed both the Light Phaser and Sega’s 3-D glasses.
- Nintendo Super Scope, shown at the January 1992 Consumer Electronics Show as the first accessory for the Super NES. It rested on the shoulder like a bazooka, was aimed through a sight on top, and ran on six 1.5-volt batteries. It sent its signal by infrared to a receiver on top of the television, so there was no cable to the console. Mean Machines listed it at £79.99 in Britain.
- Sega Menacer, the Mega Drive’s reply to the Super Scope, which Sega showed in 1992 and planned to release that October with Terminator 2: The Arcade Game.
- Namco GunCon, for the PlayStation, which arrived with Time Crisis in 1997. Edge, reviewing the Japanese release at ¥7,800, called it “the excellent Guncon”.
- Sega’s Dreamcast gun, which Sega never released in the United States, out of worry about how guns would be seen at the time.
Why they stopped working
Light guns depend on the CRT in two ways, and flat screens break both. The Zapper’s sensor has a filter tuned to about 15 kHz, the rate at which a CRT draws its lines. The NESdev Wiki notes that the filter “helps filter out more slowly changing light signals (e.g. light bulbs)”, but “will strongly attenuate the light signal from many modern televisions (e.g. LCD based)”. An LCD holds its picture steady rather than drawing it with a moving beam, so there is no flash for the sensor to catch.
The second problem is time. Light gun games “tend to expect no effective delay from the CRT”, NESdev notes, but most modern televisions process the picture before showing it, which adds display lag. By the time the white box appears on a flat screen, the game has already moved on and decided the shot missed. Guns that time the beam, like the Light Phaser and the GunCon, are thrown off in the same way, because their method depends on knowing when each line of the picture is drawn.
Sources
- Retro Gamer, Videogames Hardware Handbook volume 1 (Imagine Publishing), “Retroinspection: Magnavox Odyssey” (Baer’s light pen and lightgun; TV Game Unit 2 in June 1967; the Odyssey’s accessory port and lightgun; Shooting Gallery), the NES section (the Zapper, grey then grey and orange) and the Dreamcast section (no US release of the gun).
- NESdev Wiki, “Zapper” (web, secondary): the sequence of operations, the 15 kHz filter and display lag. https://www.nesdev.org/wiki/Zapper
- Wireframe issue 20, August 2019, Rik Cross, “2D shooting gallery”: the Seeburg Ray-O-Lite in the 1930s; how games flash each target white for a frame.
- Wikipedia, “Light gun” (web, secondary): the Ray-O-Lite in 1936; the Odyssey’s Shooting Gallery in 1972. https://en.wikipedia.org/wiki/Light_gun
- SMS Power!, “Light Phaser” development notes (web, secondary): the TH line, the latched H counter and averaging. https://www.smspower.org/Development/LightPhaser
- Sega Master System: A Visual Compendium (Bitmap Books): the Light Phaser (1986; Western release only; not usable on the Mark III; compatible games); Missile Defense 3-D and Mark Cerny.
- Game Informer issue 3, January 1992: the Super Scope at the January CES, “an infrared firing device that allows you to aim and shoot from anywhere in the room”.
- Mean Machines issue 20, May 1992: the Super Scope (shoulder-mounted; six 1.5 V batteries; receiver on the television; £79.99).
- Mean Machines issue 22, July 1992: CES report (the Menacer, due in October with Terminator 2).
- Edge issue 48, August 1997: Time Crisis (PlayStation, Namco) review, with the GunCon (¥7,800).
- psx-spx, “Controllers and Memory Cards” (web, secondary): the GunCon’s video-signal connection and its coordinate units. https://psx-spx.consoledev.net/controllersandmemorycards/