Image Metadata Viewer

The viewer organizes everything it finds into five tabs (Overview, EXIF, Histogram, Details, Analysis). Basic file facts come from extractMetadata, which loads the image and reads naturalWidth/naturalHeight, then reduces them with a recursive greatest-common-divisor function to print a true aspect ratio such as 16:9 rather than a rounded decimal. From those dimensions it derives megapixels (width x height / 1,000,000), total pixel count, bytes-per-pixel, orientation (Landscape/Portrait/Square), and a resolution tier matched against a fixed ladder of 8K (>=7680px), 4K/UHD (>=3840px), QHD (>=2560px), Full HD (>=1920px), HD (>=1280px), and SD. None of this leaves the browser, and the file's lastModified timestamp and MIME type are surfaced verbatim.

The EXIF tab is where camera data is decoded. The tool dynamically imports the exif-js library and calls EXIF.getData / EXIF.getAllTags after loading the file through a FileReader into an image element, so it reads the real EXIF block rather than guessing. It then formats the headline shooting parameters: an ExposureTime below 1 is rewritten into conventional shutter notation as 1/round(1/x) (for example 0.005 becomes 1/200s), FNumber is shown as f/2.8, ISOSpeedRatings as a raw ISO number, and FocalLength in millimetres, alongside Make, Model, LensModel, DateTimeOriginal and DateTime. If GPSLatitude or GPSLongitude tags are present the tool raises a red 'GPS Location Found' privacy warning ('This image contains location data. Consider removing before sharing publicly') and lists the raw coordinate values (it does not convert DMS to decimal). Every remaining tag is dumped in an expandable 'All EXIF Tags' list with a live count, and images with no EXIF (screenshots, social-media re-encodes, many web graphics) hit an explicit 'No EXIF data found' state because those workflows strip the block on save or never wrote one.

Beyond stored metadata, the Histogram and Analysis tabs compute fresh statistics from the pixels. The histogram redraws the image onto a downsampled canvas (its width capped at min(width, height, 500)px), then tallies 256 bins each for red, green, blue and a luminance channel computed with the Rec.601 weights (0.299R + 0.587G + 0.114B); it renders luminance as a faint backdrop (0.3 alpha) behind the semi-transparent RGB curves and reports each channel's peak bin and mean value. Color analysis runs on a separate 100x100 sample, quantizing colors into 32-value buckets to surface the five dominant colors and flagging any alpha channel below 255 as transparency. The Analysis tab snaps the dimensions to the closest of eight named aspect ratios (1:1, 4:3, 3:4, 16:9, 9:16, 3:2, 2:3, 21:9), marking an 'Exact match' only within 0.01, checks suitability against five platform minimums (Instagram 1080px, Twitter/X header 1500px, YouTube thumbnail 1280px, Facebook cover 820px), and prints physical output size at 300, 150 and 72 DPI. An Export JSON button saves the file and dimension facts plus a small analysis block (resolutionTier, hasTransparency) but not the EXIF block to a metadata_<filename>.json download.