Skera

The problem

Every application that can open a camera RAW file wants to own the library first: import the shoot, build a catalog, work inside it from then on. That is a fair trade while editing and a bad one while simply looking at what you shot. The lightweight alternative — the operating system's own viewer — stalls on a full-resolution file and cannot read RAW at all.

The approach

Skera keeps the viewing and drops the library. It points at a directory on disk, sorts it the way a file browser would, and starts decoding — no import, no catalog, no database, and nothing written back, since rotation is a view transform rather than an edit. The speed comes from the engine underneath: a byte-budgeted image cache, decoding on a priority thread pool off the UI thread, and neighbouring images read ahead before they are asked for.

What it does

RAW without an import step

Sixteen RAW formats across ten manufacturers — Nikon, Canon, Sony, Fujifilm, Olympus, Panasonic, Pentax, Adobe, Sigma and Hasselblad — decoded through libvips on macOS and Linux. The version 1 Windows build ships without that decoder and handles standard formats only.

The whole EXIF record

Camera body, lens, exposure, ISO and GPS, read locally through exiv2 on all three platforms and shown in a panel one key away.

Side-by-side compare

Pin one image as A and sweep B beside it, with zoom and pan linked across both panes. Both images are pinned in the cache so neither is evicted mid-comparison.

A cache measured in bytes

An LRU cache budgeted at 768 MB of resident decoded pixels rather than a file count, so a single 60-megapixel RAW cannot quietly evict forty JPEGs.

Image-first by default

The toolbar and filmstrip float over the photograph and hide on a keypress, leaving nothing but the picture on true black. Every action has a key.

Pricing

Free. The application source is private; binaries are published as public GitHub releases.