Stars vs flags for culling
Somewhere in your catalog there is a shoot where half the frames are 3 stars, a quarter are 4, and you no longer remember what the difference meant. That is a system problem rather than a discipline problem, and the fix is simple: stop grading and start sorting.
- Should I use stars or flags for culling?
- Flags for the cull, stars after it. The first pass through a shoot is a binary decision, deliver or don't, and flags answer exactly that question with one keystroke.
- Why is a 1 to 5 star scale slower for culling?
- Because it forces an absolute judgment on every frame, is this a 3 or a 4?, and humans are measurably worse at absolute grading than at simple comparative calls.
- What do the star levels mean in a hybrid system?
- Stars mark purpose, not quality.
Five stars is five decisions per frame
The 1 to 5 scale feels professional: nuanced, granular, serious. But watch what it makes your brain do on every frame: place this image on an absolute quality ladder, calibrated against every photo you have ever taken. Is this a 3 or a 4? You will deliberate that boundary hundreds of times per shoot, and the boundary itself drifts as you go. Your 3 at frame 200 is not your 3 at frame 2,000.
Perception research points at why. Humans are much better at comparative judgments (this one, of these two) than at absolute ones (this is a 7 out of 10). A star scale demands absolute grading thousands of times in a row. A flag asks one comparative question, does this frame make the gallery or not, and you can answer it in a beat.
There is a second cost. Research on decision fatigue suggests choice quality degrades as decisions stack up. A five-way choice fatigues faster than a two-way one, which is partly why star-based culls end in mid-catalog mush: everything is a 3 because 3 is what a tired brain presses. More on the psychology in why picking photos is so hard.
The deeper issue is that during a cull, the star information is unused. The client gallery does not care whether a selected photo was a 3 or a 4. It ships either way. You are paying extra seconds per frame to record a distinction that has no downstream consumer.
Flags ask one question
The flag system asks the only question the cull needs answered: in or out. In Lightroom that is P for pick, X for reject, and with auto-advance on, your hands never leave the keyboard. The frame appears, you answer, the next frame appears. No deliberation about degrees. Setup details are in the Lightroom culling workflow, and the full key-by-key flow is in the keyboard-only cull.
Two rules make binary culling fast:
- Default to reject. A frame must earn its flag. If you hesitate more than a couple of seconds, that hesitation is the answer. A frame you have to talk yourself into is not one the client will miss.
- Sort, do not grade. Resist the urge to mark how good a pick is on this pass. You are a sorting machine right now. Taste comes later, applied only to the survivors.
This is also the shape AI culling has converged on: tools make an in/out recommendation, you approve or overrule. Nobody built a tool to argue about 3 versus 4 stars, because the distinction never carried weight.
Stars are for purpose
None of this means stars are useless. They belong after the cull, applied only to picks, where each level maps to something you will do with the image. Grading 3,000 frames is a waste of an evening. Tagging a few dozen standouts among your picks takes minutes and pays off at delivery time.
| Marking | Means | Downstream use |
|---|---|---|
| Reject flag (X) | Out | Filtered away; deleted per your retention policy |
| Pick flag (P) | Delivers | The edit set and client gallery |
| 3 stars | Highlight | Extra editing attention |
| 4 stars | Sneak peek | Same-week preview, blog, socials |
| 5 stars | Portfolio-grade | Album spreads, website, competitions |
Every level now has a consumer. The filter bar builds your sneak-peek set in one click, album pre-designs start from the 5s, and your portfolio refresh is a saved search instead of a dig through old folders.
One system, everywhere, forever
Whatever scheme you pick, the compounding win is consistency. Write your marking key down in one sentence and use it on every shoot. Five years from now, "show me 5 stars from 2026" should mean the same thing as it does today. Star ratings live in standard XMP metadata and travel between apps. Flags were app-specific for years and, outside Lightroom Classic 13.2 and later, may still not survive a move between programs. That is a practical reason to let flags do the temporary work of the cull and let stars carry the meaning that needs to persist.
The whole system in four steps
- Sweep pass: reject the obvious kills, blinks, blur, misfires, at speed.
- Pick pass: flag your selects, default-to-reject, auto-advance on. In or out, nothing else.
- Star pass (picks only): a fast scroll through survivors tagging 3s, 4s, and 5s by purpose.
- Filter and go: picks to editing, 4+ to the sneak peek, 5s to the album shortlist.
The star pass touches a fraction of the shoot, so it barely registers on the clock, while the binary passes run at full keyboard speed. It is the marking backbone of every fast method we teach, from the two-hour wedding cull to the complete culling guide. And it removes the mistake that wrecks most culls, rating everything 1 to 5, which sits at number four in our culling mistakes list for a reason.
FAQ
Should I use stars or flags for culling?
Flags for the cull, stars after it. The first pass through a shoot is a binary decision, deliver or don't, and flags answer exactly that question with one keystroke. Stars come in later, applied only to the picks, to mark purpose: sneak peek, album candidate, portfolio.
Why is a 1 to 5 star scale slower for culling?
Because it forces an absolute judgment on every frame, is this a 3 or a 4?, and humans are measurably worse at absolute grading than at simple comparative calls. The boundary between adjacent star levels is fuzzy, so you burn seconds deliberating distinctions that don't change what gets delivered.
What do the star levels mean in a hybrid system?
Stars mark purpose, not quality. A common scheme after a flag-based cull: no stars = deliver as normal, 3 = highlight worth extra editing attention, 4 = sneak-peek and blog candidates, 5 = portfolio and album spreads. Because stars only touch the picks, the star pass is short.
Do flags and stars transfer between apps?
Star ratings are part of the XMP metadata standard and travel well between Lightroom, Bridge, Capture One and most culling tools. Flags used to be app-specific; Lightroom Classic has written them to XMP since version 13.2, but other programs may still ignore them, one more reason to treat flags as a working tool for the cull and let stars carry the meaning that needs to persist.
Never cull alone at 1 AM again.
Kepla for Mac clears the obvious misses from a card, names the reason on every frame it sets aside, and leaves the choosing to you. Nothing is ever deleted, moved or renamed. Free through the private preview. The first hundred photographers keep it at $99 a year.
- P for pick, X for rejecthelpx.adobe.com