01
Metadata first
The fast first pass reads the information already attached to each photograph: camera, lens, capture settings, dates, ratings, picks, collections, and folders. The report becomes useful before image analysis begins.
How LensAtlas measures
Quality profile.
A clear reading of image-quality patterns in your own archive — not a laboratory verdict on your lenses or technique.
Useful quickly.
Meaningful in context.
LensAtlas keeps metadata and image analysis separate so you can see the shape of your archive early, then add quality signals when you want the deeper view.
02
Detailed image analysis
The optional second pass works from bounded previews in background workers. It adds quality indicators while the report stays responsive and continues to update with the latest analysed photographs.
What each signal is telling you.
Every indicator is normalized to a 0–100 scale. The direction on each card tells you how to read the gauge, not whether a photograph is objectively good or bad.
01
Apparent sharpness
How much meaningful detail is visible in the preview.
Higher is better
LensAtlas combines normalized whole-preview detail with a bounded local-focus pass. Looking for meaningful high-detail regions means an intentionally soft background does not unfairly dilute a sharp subject.
02
Low-focus likelihood
The likelihood that the image has unusually low focus confidence.
Lower is better
This is the inverse of the focus-confidence estimate. It is a useful comparison signal, but it is not a verified missed-focus rate and cannot know your intended point of focus.
03
Motion-blur likelihood
Directional softness that is consistent with movement.
Lower is better
The signal looks for softness patterns associated with camera movement or subject movement. It can indicate a pattern in the archive, but it cannot identify which movement caused it.
04
Exposure-problem likelihood
The likelihood of an under- or over-exposure pattern.
Lower is better
LensAtlas evaluates the luminance distribution for evidence that important parts of the frame are pushed too far toward either tonal limit. It is not a judgement of your creative exposure choices.
05
Noise
Texture variation in shadow and midtone regions.
Lower is better
Noise is interpreted with ISO context because a high-ISO photograph naturally has a different baseline from a low-ISO photograph. The result is a comparative signal, not a sensor laboratory test.
06
Highlight clipping
The proportion of pixels near the bright tonal limit.
Lower is better
The indicator estimates how much highlight information may be lost at the upper end of the luminance range. Bright areas can be intentional, so use it alongside the photograph and its context.
07
Shadow clipping
The proportion of pixels near the dark tonal limit.
Lower is better
The indicator estimates how much shadow information may be lost at the lower end of the luminance range. Dark backgrounds and deliberate silhouettes can make a high value meaningful rather than problematic.
Context is part of the measurement.
The profile is designed to help you notice patterns, compare groups, and decide what to look at more closely.
0–100 indicators
Values are normalized within the analysed archive. They are most useful when comparing similar photos, lenses, bodies, or capture conditions.
Sample counts matter
Each gauge shows how many photographs were analysed. A small group can be interesting, but it should not outweigh a broad, well-covered pattern.
Per-lens profiles
Lens detail pages aggregate the quality indicators from all analysed photographs taken with that lens across your saved scans.
Missing data stays visible
If a source has no pixel analysis or a value cannot be measured, LensAtlas shows that limitation instead of treating it as a poor result.
Signals, not verdicts.
LensAtlas cannot establish an objective lens defect, autofocus failure, or controlled optical score from a finite personal archive. Results can reflect subject movement, light, technique, framing, processing, and the way a lens was used. Use the profile to find patterns worth investigating, then make the final call from the photographs themselves.