Image

1. 01TIFF opens in the browser — no browser decodes one on its own, and it is what scientific cameras write
2. 02Multi-page TIFF: z-planes, time points or channels stepped through with a Page control
3. 0316-bit and deeper samples mapped to the screen by a choice you make, with the measured pixel range and the applied window shown beside it
4. 04Auto-contrast that matches ImageJ exactly: the same 0.35% saturated-pixel rule, checked against ImageJ’s own code over four real pages
5. 05Saturation is a control, not a hidden constant — drag it to zero for the plain minimum-to-maximum extent
6. 06The measured range is flagged when the data uses under an eighth of its container — the case where a correct mapping shows a black image
7. 07A TIFF compression this reader cannot decode is named, with what to convert it to, instead of drawing a blank rectangle
8. 08PNG, JPEG, GIF, WebP, BMP and AVIF decoded by the browser, so orientation and colour profiles behave as they do everywhere else
9. 09Format read from the file’s own first bytes, so a mislabelled image still opens
10. 10Dimensions, depth, colour model and page count read from the header without decoding the image
11. 11Invert for fluorescence images, which are often easier to read on white
12. 12The same viewer draws the file manager preview and the stage tab

TIFF decoding is UTIF.js. It was chosen over the more actively maintained alternative on a measurement rather than on maintenance status: the alternative refuses PackBits compression, which real microscopy writes. Header facts — dimensions, depth, colour model, page count — are read without decoding anything, and are checked against Pillow and tifffile over a corpus covering every format and colour model above.

Everything runs client-side. Files are parsed in the browser and projects are saved on this machine, so an unpublished sequence is never uploaded.

Why does my 16-bit TIFF look black in other viewers?

Because the pixels use a small part of the file’s range. A camera with a 12-bit sensor writes into a 16-bit file, so the values might span 120 to 4095 out of 65535. Mapping the whole container is correct and shows almost nothing. This viewer measures the actual range, shows it to you, and maps by it — and lets you switch to the full-container mapping when you want pages of a stack to be directly comparable.

Can it open a multi-page TIFF?

Yes — a Page control steps through them. The TIFF format does not record whether its pages are z-planes, time points or channels, so the viewer reports pages rather than claiming to know which.

Is my image uploaded anywhere?

No. It is decoded in the browser from the copy stored in your project.

How large an image can it open?

Files up to 128 MB. Pixels matter more than bytes — an image costs four bytes per pixel once decoded — so the dimensions are always stated before anything is drawn.

Does it read the image’s metadata?

It reads what the header declares: dimensions, bits per sample, colour model, and for TIFF the page count, all without decoding the image. Selecting a 40-megapixel scan in the file manager therefore costs what selecting a thumbnail costs.