applets.

ASCII Camera

← All apps

About ASCII Camera

This turns a picture into text: every character on the screen stands for a small patch of the picture, and brighter patches get characters with more ink. Pick a sample, OPEN a photo (or drop one on the screen, or paste one), or START CAMERA to watch yourself live. STOP keeps the last frame so you can copy or save it. Keys: C camera, O open, I invert.

Your pictures never leave this tab. Your picture isn't uploaded or stored: close the tab and it is all gone. The camera only switches on when you press the button, and STOP (or leaving the tab) switches it off again.

THE STORY

Pictures made of letters are older than computers. Typists were making them on typewriters by the end of the 1800s. In 1966 Leon Harmon and Ken Knowlton at Bell Labs scanned a photograph and printed it as a grid of small symbols, each standing for a level of grey: up close you see symbols, from across the room a picture. It became one of the best-known early pieces of computer art.

In computer rooms of the 1960s and 70s people printed pictures on line printers, on long sheets of fan-fold paper, sometimes typing over the same spot twice for a darker grey. ASCII, the code that gave computers a shared set of printable characters (95 of them since its 1967 revision), was first published in 1963. The IBM PC (1981) added shaded blocks ░ ▒ ▓ █, and Unicode has had all 256 patterns of eight-dot braille since 1999, which pack eight tiny dots into one character. Green and amber screens came from the phosphor coating inside the glass of the terminals of the time.

HOW IT WORKS

The picture is shrunk to one pixel per character (for braille, two across and four down). A character is about twice as tall as it is wide, so the picture gets half as many rows as its shape suggests. Each pixel's brightness is 30% red, 59% green and 11% blue, as eyes see green as the brightest. Auto levels stretches the darkest and brightest 1% to black and white; then contrast and brightness are applied, and the brightness picks a character from the set, from the emptiest to the fullest.

DENSE is a 70-character ramp from Paul Bourke's well-known page on the subject. Dither uses Floyd and Steinberg's method (1976): each rounding error is passed on to the next pixels, so smooth greys survive even with few characters. Braille always needs it. Edges uses the Sobel filter (Irwin Sobel and Gary Feldman, 1968) to find where the brightness changes fastest and draws the line with - / | \ along it.

Text you copy loses its colour and looks best in a monospaced font. On white paper use Invert, so the dark parts get the ink. The PNG is drawn exactly as the screen shows it, without the glass.

FURTHER WATCHING AND READING

DROP A PICTURE
ASCII CAMERA 80

PICTURE

Pictures stay in this tab: nothing is uploaded or stored. You can also drop or paste a picture.

CHARACTERS

TONE AND SCREEN

EDGES