The History of Light Guns: Galleries to Modern Tracking

About this guide

A sourced history of shooting galleries, CRT lightguns, arcade and console adoption, flat-panel decline and modern tracking systems.

Executive chronology

Date	Event	What changed	Confidence / evidence
Before 1930s	Mechanical shooting galleries	Physical projectiles or mechanical linkages actuated targets; these are predecessors, not optical guns.	Contextual boundary; no “first gallery” claim is made.
1934–1936	Charles W. Griffith / Rayolite and Seeburg-era optical galleries	Griffith filed US 2,007,082 on 12 April 1934; it describes a trigger-fired light flash projected along the sights to a photoelectric target. The patent issued 2 July 1935. Surviving advertising and museum research date Seeburg/Rayolite commercial games to 1936.^[1]^[2]	Patent establishes mechanism and dates; museum research corroborates commercialization.
c. 1952	MIT Whirlwind light pen	A photosensitive pen selected illuminated points on a CRT. It established the computer-input principle of detecting when the CRT’s beam reached the sensor’s field of view; a dark point could not be selected.^[3]	Computer History Museum collection record. A light pen is a direct ancestor, not an arcade gun.
1967–1968	Ralph Baer’s Brown Box lightgun	Baer’s prototype gun played Target Practice on the Brown Box. Magnavox licensed the system; the Odyssey appeared in 1972 and later offered the gun plus four target games as an expansion.^[4]^[5]	Smithsonian object record plus Baer/Sanders patent.
1972	Magnavox Odyssey Shooting Gallery	First-generation home TV target shooting used a gun-mounted photocell to detect a bright generated target. It did not derive a continuous X/Y coordinate and did not need to know the raster time of every point.^[6]^[7]	Primary patent and museum record.
1974	Atari Qwak!	Atari sold a CRT duck-shooting video game with a tethered rifle. It is often called the first arcade video lightgun game; the surviving Atari flyer securely establishes manufacturer and 1974 date, but “first” remains a secondary-history convention rather than a claim printed by the flyer.^[8]^[9]	Date cross-checked with an original flyer and a specialist archive; priority claim explicitly qualified.
1984–1985	Nintendo Famicom/NES Zapper era	Games such as Duck Hunt popularized target-flash detection: after a trigger event, software shows a black frame, then a white target/mask; the gun reports whether suitably rapid CRT light was seen. Nintendo’s patent describes black-picture, white-mock-target, synchronizing-signal filtering and ambient-light rejection.^[10]^[11]	Nintendo patent plus NES hardware documentation. Product regional dates vary, so the range is used.
1992	Super NES Super Scope	Nintendo’s wireless gun still used CRT optical timing for aim; infrared was used for gun-to-receiver communication, not to locate the screen. Nintendo’s development manual documents picture-tube scanning, horizontal/vertical counters and optical/virtual alignment.^[12]	Manufacturer development manual.
1992	Konami Lethal Enforcers	The operator manual documents simultaneous one/two-player play, gun checks and the now-familiar instruction: point off screen and pull the trigger to reload.^[13]	Manufacturer/operator manual.
1994–1996	Polygonal arcade shooters and physical feedback	Virtua Cop (1994) and peers made calibrated freehand CRT shooting central to polygonal rail shooters. Namco’s 1994-priority patent shows a photosensor behind a collector lens plus a solenoid-driven reciprocating slide for blowback-like impact; the 1996 publication date is not the feature’s debut date.^[14]	Game date is conventional catalog history; patent is primary evidence for the mechanism and 1994 priority.
1995	Namco Time Crisis	A foot-operated input selected attack versus defense; releasing the pedal hid the player and reloaded. The patent claims a gun, foot input and attack/defense computation and has Japanese priority of 7 September 1995.^[15]	Manufacturer patent; strong for mechanism and date.
1997	Namco GunCon / G-Con 45	Home systems increasingly tapped the video signal’s sync as a timing reference while the muzzle photosensor detected the CRT beam. Namco’s controller patent has 17 October 1996 priority and describes a controller/video cable arrangement.^[16]	Manufacturer patent; commercial release commonly dated 1997.
Late 1990s–2000s	CRT timing formalized for PCs and arcades	Patents explicitly describe comparing photodetector time with horizontal/vertical synchronization to calculate X/Y, and also describe arcade alternatives using IR sources plus a CCD in the gun.^[17]^[18]	Technical patents from ACT Labs and Konami.
2006 onward	Wii-style IR pointer mainstreamed	The Wii Remote’s imaging device observes two light-emitting markers near the display and produces marker-coordinate data; Nintendo confirms the “Sensor Bar” emits infrared lights. This is a marker-based pointer, not a traditional display-sensing lightgun.^[19]^[20]	Nintendo patent and support documentation.
Late 2000s–2010s	USB IR-gun products	AimTrak places an IR LED strip by the monitor, observes it from the gun, calibrates aim, and enumerates as USB mouse/game-controller interfaces. Multiple guns require distinct IDs and host software must keep devices separate.^[21]^[22]^[23]	Manufacturer manual/product page and MAME documentation. Exact first-sale date was not established from a primary source, so none is asserted.
Late 2010s–2020s	Four-marker DIY/commercial systems	GUN4IR and OpenFIRE use four IR emitters around the display and an IR-sensitive camera in the gun; onboard firmware converts marker geometry to aim and sends commands by USB (OpenFIRE also documents Bluetooth capability).^[24]^[25]	Project/manufacturer documentation. No unverified “first” date is asserted.
2020s	Border-based computer vision	Sinden uses an in-gun camera and a software-drawn rectangular border as the reference. The vendor specifies over 50 calculations/s and under 20 ms end-to-end for its own pipeline, not for the whole game/display chain.^[26]^[27]^[28]	Manufacturer documentation and patent; performance wording kept vendor-attributed.

1. Three different families, often given one name

1.1 Gun emits light, target senses it

Rayolite-era systems reverse the direction assumed by most video-game explanations. The trigger briefly lights a bulb in the gun; a lens projects that pulse along the sights; a photoelectric cell in a physical target closes the scoring/animation circuit.^[29] The gun is an emitter and the target is the receiver. Mechanical scenery, belts, relays, counters and sound effects create the gallery response.

This family does not measure screen coordinates and is not display-dependent. Its optical link merely answers “was the photosensitive target illuminated while a valid shot window was open?” Anti-cheat circuits were historically important because a room lamp could otherwise operate a photocell.^[30]

1.2 Traditional video lightgun: gun senses the display

Brown Box/Odyssey, Zapper, Super Scope, GunCon and many arcade guns place a photocell, phototransistor or photodiode behind a lens in the barrel. The display emits; the gun receives. However, “CRT lightgun” hides two materially different schemes:

Brightness / target-flash detection. The software blanks the screen, then presents one white target or a sequence of white masks. A light/no-light result tells the game whether a target occupied the sight line. Nintendo’s patent deliberately checks a black interval before the white mock target and filters for a signal related to television sync, reducing false hits from steady ambient light.^[31] This can identify an object without ever returning a full coordinate.
Raster-time coordinate detection. The game shows a bright raster (often a white frame). The sensor produces a pulse when the scanning beam crosses the small region seen through the barrel. Hardware measures elapsed time from vertical frame start and/or counts horizontal scan lines; the phase within a line gives X and the line count gives Y. Nintendo’s Super Scope documentation and later PC-lightgun patents explicitly describe horizontal/vertical counters or sync-fed processing.^[32]^[33]

Some products combine ideas: screen flash ensures a strong, known optical signal, while beam timing determines where in the raster it was seen. Therefore “all old guns only looked for a white box” and “all old guns returned X/Y from scanline timing” are both over-generalizations.

1.3 Pointer systems: gun observes references, not target pixels

Modern display-independent systems usually place a camera in the gun and make the gun observe geometry:

two IR clusters or points near the display (Wii, AimTrak);
four IR points around the display (GUN4IR, OpenFIRE);
a visible quadrilateral border surrounding the image (Sinden).

Calibration estimates a mapping from camera observations to screen coordinates. With four corners or a border, software can solve a perspective transform (homography) and compensate for viewpoint/rotation more directly than a single bar. With fewer markers, firmware combines known marker spacing, tilt estimates and a user calibration model. These are absolute optical pointers wearing gun ergonomics. They work on LCD/OLED because tracking references, rather than CRT phosphor timing, define aim.^[34]^[35]^[36]^[37]^[38]

External-camera systems are another pointer class: a camera near the display can track LEDs, colored shapes, retroreflective markers or the controller body. The geometry is inverted—camera outside, marker on gun—but it still estimates pose and intersects the gun’s pointing ray with a calibrated screen plane. Computer vision may add marker recognition, contour detection or pose estimation; it does not turn the display into a CRT-like optical target.