RGB to RGB555 packer

Legacy graphics hardware stores color as a fifteen-bit RGB555 word instead of three eight-bit channels. This RGB to RGB555 packer takes red, green, and blue integers from zero through two hundred fifty-five, truncates each channel to five bits by dropping the three least-significant bits, and packs the result as red in bits fourteen through ten, green in bits nine through five, and blue in bits four through zero. The unused high bit stays zero so the packed integer always sits between zero and thirty-two thousand seven hundred sixty-seven. Designers, emulator authors, and retro-game toolchains use this mapping when a palette, framebuffer, or BIOS dump expects the historical fifteen-bit high-color layout used by VGA modes, older PC games, and many graphics APIs. The engine is fully deterministic: no network, no random seed, and no wall clock.

What RGB555 packing means and how the fifteen-bit word is built

Eight-bit sRGB, the model CSS and modern paint tools use, gives each of red, green, and blue two hundred fifty-six distinct levels. Fifteen-bit high color, often called RGB555, keeps only thirty-two levels per channel. The historical conversion is truncation, not perceptual rounding: shift each eight-bit channel right by three bits, which is the same as integer-dividing by eight and discarding the remainder.

Pure red two hundred fifty-five becomes thirty-one, mid grey one hundred twenty-eight becomes sixteen, and any value from zero through seven collapses to zero. Those three five-bit fields are then packed into one unsigned integer with red in bits fourteen through ten, green in bits nine through five, and blue in bits four through zero. Bit fifteen stays unused and is always written as zero, so the packed word never exceeds thirty-two thousand seven hundred sixty-seven, which is hexadecimal seven F F F.

This is the RGB555 layout documented by Windows 15-bit color, VGA high-color modes, and many PC game engines—not the BGR555 swap some handheld consoles use, where blue occupies the high five data bits.

Input contract, truncation rules, and validation that never clamps

Outer whitespace on string channels is trimmed. A value is legal only when it is a finite whole number from zero through two hundred fifty-five inclusive. There is no silent clamp and no nearest-level rounding: seven right-shifted three bits is zero, eight becomes one, and two hundred fifty-five becomes thirty-one.

That truncate-only rule matches the hardware path used by VGA high-color DACs and by countless asset converters that drop the three least-significant bits before writing a fifteen-bit framebuffer. After validation the engine packs rgb555 as red five shifted left ten bits, bitwise-or green five shifted left five bits, bitwise-or blue five. Orange two hundred fifty-five, one hundred twenty-eight, zero therefore becomes thirty-one, sixteen, zero and packs to thirty-two thousand two hundred fifty-six, written 0x7E00.

White two hundred fifty-five on every channel packs to the maximum thirty-two thousand seven hundred sixty-seven. Black packs to zero.

Where an RGB to RGB555 packer earns its keep, and what it costs

ROM hackers and asset pipelines use it to turn modern eight-bit mockups into the coarse thirty-two-level-per-channel look of 15-bit high color without standing up a full graphics library. Computer-science courses use it as a clean bit-shift exercise: students can check that two hundred fifty-five right-shifted three bits is thirty-one and that the packed word for white is 0x7FFF. Front-end developers who still ship fallbacks for historical formats, or who document how a canvas export will look after five-bit quantization, can call the same contract from a script.

Pair the free in-browser RGB to RGB555 packer with hex or HSL converters when the next step is a stylesheet, and with the 24-bit RGB integer packer when you need eight bits per channel instead of five. On this page the work runs entirely on your device.