Convert an image to black and white

Five conversion formulas, a contrast control and three classic toners (sepia, cyanotype and selenium), all computed in your browser.

Drop your files here

.jpg, .jpeg, .png, .webp · up to 50 MB

Choose files

How to Convert to Black and White

1
Add your image
JPG, PNG or WebP, up to 100 files at a time. Transparency in a PNG survives the conversion untouched.
2
Pick a conversion formula
Luminance BT.709 is the natural-looking default. Average, desaturate and single-channel each give a different reading of the same photo.
3
Click Convert and download
The result keeps the original format and pixel size. Several files come back as one ZIP.

Why five formulas instead of one

"Black and white" is not a single operation: it is a choice about which colours become which greys, and different choices suit different pictures. Luminance BT.709 weights green heavily and blue barely at all, which matches how the eye judges brightness and keeps skin and foliage where you expect them. BT.601 is the older television weighting; it lifts reds slightly and is often kinder to warm interior shots. A plain average of red, green and blue treats all three equally, which brightens blue skies dramatically and is the classic "flat" look. Desaturate takes the midpoint between the lightest and darkest channel, the same maths as the HSL lightness channel, and produces a lower-contrast, gentler image. Single-channel throws two channels away entirely, which is the digital equivalent of screwing a coloured filter onto the lens: the red channel smooths skin and blows out red lips, the blue channel turns a blue sky nearly black and makes clouds leap out.

The conversion happens in linear light

Image files do not store brightness proportionally: a pixel value of 128 is not half as bright as 255, because the values are gamma-encoded so that the limited byte range is spent where the eye is most sensitive. Averaging those encoded numbers directly, which is what most quick grayscale filters do, makes shadows muddy and washes out skies, because the arithmetic is being done in the wrong space. This tool converts each channel to linear light through a 256-entry table, does the weighting and the contrast adjustment there, and converts back afterwards. Contrast is applied around the linear midpoint, so raising it opens up the tonal range instead of clipping the highlights.

Toning, and why the shadows stay neutral

Sepia, cyanotype and selenium are darkroom processes, and each has a characteristic colour: a warm brown, a printer's blue and a cool purple-grey. A naive tint multiplies the whole image by that colour, which stains the blacks and makes the picture look like it was printed on coloured paper. Here the tint is scaled by the grey value of each pixel, so a pixel at brightness 10 barely shifts while a pixel at brightness 240 takes the full tone. Blacks stay black, highlights carry the colour, and the result reads as a toned print rather than a colour cast.

Nothing is uploaded

The file is decoded, converted and re-encoded inside this browser tab with WebAssembly codecs. There is no server, no queue and no stored copy; closing the tab is enough to clear it. The EXIF block is not carried into the result, so camera details and any GPS coordinates are dropped along the way.

Frequently asked questions

Does the photo get uploaded anywhere?

No. Reading, decoding, converting and encoding all run locally in your browser. Nothing is sent over the network at any point, which also means the tool works with the connection off once the page has loaded.

How do the formulas actually differ?

They differ in how much each colour channel contributes to the grey. BT.709 uses 0.2126 red, 0.7152 green and 0.0722 blue; BT.601 uses 0.299, 0.587 and 0.114; average uses a third each; desaturate takes the midpoint of the brightest and darkest channel; single-channel uses one channel and ignores the rest. On a colourful photo the difference between the extremes is easily twenty levels of grey. A blue sky can be light or nearly black depending on which you pick.

Why is the black-and-white file smaller than the original?

Because the colour information is genuinely gone. JPEG stores brightness and colour separately, and once all three channels carry the same values the colour planes compress to almost nothing. A drop of a quarter to a third of the file size is typical at the same quality setting.

How do I make a sepia photo?

Convert as usual, then set Toning to Sepia. Strength 60 is a subtle warm print; push it to 100 for the strong, obviously-toned look. The same control gives cyanotype for a cold blue and selenium for the cool purple-grey of a selenium-toned silver print.

Does PNG transparency survive?

Yes, byte for byte. The alpha channel is copied through untouched and only the three colour channels are recomputed, so a transparent logo comes back transparent with grey artwork.

Can I get the colour back afterwards?

Not from the black-and-white file: the colour information is discarded, not hidden. Keep the original.

How many files can I convert at once?

Up to 100 files, 300 MB combined. Each file is converted independently with the same settings and the results come back as pulsarkit-black-and-white.zip. A file that fails to decode does not stop the rest; you get a note saying how many were skipped.