metafaker

MetaFaker is a static browser application for loading one or many images, re-encoding them as JPEG, and either stripping EXIF metadata or replacing it with a generated metadata set. The app is fully client-side. There is no backend, no upload step, and no cloud storage in the current version.

Current behavior

• Accepts multiple images from drag and drop or the file picker.
• Builds a local batch queue with previews and per-image state.
• Parses original EXIF locally in the browser.
• Re-encodes each processed image as a new JPEG.
• Either strips metadata entirely or injects a generated EXIF block.
• Downloads processed files locally and keeps the original uploaded filename when possible.

Architecture

The application is split into small browser modules:

• index.html Defines the UI shell, upload zone, preview panel, action buttons, metadata panel, and the single metadata tab.
• css/style.css Provides the layout, responsive batch queue, preview card styling, metadata table styling, and toast notifications.
• js/main.js Bootstraps the app, initializes theme state, and wires drag/drop plus file input events to the controller.
• js/ui.js Owns application state, queue management, rendering, downloads, and the high-level processing workflow.
• js/exif.js Generates fake EXIF data, parses source EXIF, verifies written metadata, formats values for display, and normalizes GPS output.
• js/data.js Contains the static camera database, US location database, and lens mapping data used for internally consistent fake metadata.
• js/helpers.js Provides shared utilities such as ID generation, byte formatting, GPS conversion helpers, blob/data URL helpers, and the canvas export pipeline.

Processing pipeline

For each loaded file, the app follows this flow:

1. The browser reads the input file locally and creates an object URL for previewing.
2. exifr parses the original metadata from the input file.
3. When the user clicks Randomize Current or Randomize Entire Batch, the image is drawn into a canvas and exported as a fresh JPEG.
4. generateFake() builds a camera profile, lens data, timestamps, exposure values, and a US-bounded location from the static datasets.
5. piexifjs writes that EXIF block into the freshly exported JPEG.
6. A second GPS normalization pass rewrites latitude and longitude tags to conventional US coordinates and references.
7. The app reads the resulting JPEG back once more to verify that metadata was written successfully before making the output available for preview or download.

The Strip Current Metadata path uses the same canvas export step but skips EXIF injection and leaves the output JPEG without metadata.

Metadata generation

The generated EXIF data is built from static profiles rather than arbitrary free-form values. The intent is internal consistency between related fields.

Examples of fields that are generated together:

• camera make, model, and software
• lens make and lens model
• shutter speed, aperture, ISO, focal length, and 35 mm equivalent focal length
• white balance, flash, metering mode, exposure program, and exposure mode
• capture timestamps including sub-second precision
• GPS latitude, longitude, altitude, date stamp, and time stamp

The location dataset is limited to US cities in js/data.js, and js/exif.js clamps the final GPS coordinates to a US bounding box before writing them.

Export format

• Output files are always JPEG.
• If the uploaded file already used a .jpg or .jpeg extension, that filename is preserved.
• If the source file used another extension, the basename is preserved and the extension becomes .jpg.

The canvas export pipeline in js/helpers.js creates a fresh JPEG bitmap and may resize large inputs before export. This keeps the processing path uniform and avoids depending on the original container format.

Runtime dependencies

The app is static, but it relies on two browser-loaded libraries:

• piexifjs from jsDelivr for binary EXIF writing
• exifr from jsDelivr for EXIF parsing and readback verification

There is no build step. Opening index.html through a static host is sufficient.

Local state

The app stores only theme preference in localStorage under the metafaker.theme key.

Processed images, previews, and queue state live in memory for the active tab only. Resetting the queue revokes object URLs and clears that in-memory state.

Development notes

• The app is meant to run as a static site.
• There is no server-side API surface.
• The queue and preview logic are optimized to avoid rendering every thumbnail at full cost for large mobile batches.
• The metadata panel always shows either original EXIF, generated EXIF, or an explicit stripped state for the selected image.

Hosting

Because the project is fully static, it can be hosted on any simple static host such as GitHub Pages. No environment variables or API keys are required for the current local-only version.