Cull 1,000 photos in an hour: the timed sprint

This is not your normal workflow. It is what you run when the client wants images tonight and you have one hour of usable attention left. Four timed blocks, two rules, and a long list of things you deliberately do not do. Set a real timer where you can see it. The clock is not motivation here, it is the mechanism.

Sixty minutes, a thousand frames, and a timer you can see

Do the arithmetic before you argue with the idea. Sixty minutes across 1,000 frames is 3.6 seconds a frame, and that has to include setup, a break, and getting the selection out the other end. So the sprint is built around one hard constraint: you cannot look at any frame twice. Every rule below exists to protect that.

Which means this is a deadline protocol, not a way of life. Your normal method has room for a second opinion and a walk around the block, and it should: that version, with a target time on each pass, is in how to cull photos fast. The sprint is for the Sunday night when a venue needs images for a Monday post, or when the alternative is not doing it at all, which is the real competitor here. A shortlist picked in an hour beats a perfect one picked in three weeks, every time.

One thing to be clear about up front, because it is what makes the speed safe. Sprinting is a marking job. You are not throwing anything away, you are drawing a line around the frames that go forward tonight. Every other frame stays on the drive, untouched, and you can revisit any of it tomorrow with a clear head. That is the whole reason you are allowed to decide this quickly.

The five minutes that decide whether the hour works

Almost every failed sprint fails in setup. Get these six things right and the rest is mechanical.

• The copy is finished and backed up. Never sprint against a card that is still transferring. You will spend the hour watching a progress bar.
• Previews are built. Standard previews or smart previews, built and done, before the timer starts. Rendering time is not culling time and it will eat a third of your hour.
• Your number is written down. On paper, next to the keyboard. Not a feeling, a range. A shortlist without a target has no finish line.
• One window, full screen, loupe view. No grid, no second monitor with email on it, no browser.
• Phone in another room. Not face down. Another room.
• A visible timer. Kitchen timer, phone alarm on the shelf, anything with block alarms. Watching a clock in the corner of your screen does not work, because you will stop watching it.

Then learn or confirm the two keys you will use, and only those two. Fumbling for a shortcut at 1.5 seconds a frame is the difference between finishing and not: keyboard shortcuts for culling covers the short list worth memorizing. If your software needs configuring to move at this speed, do that on a quiet day rather than during the hour.

The sprint, block by block

Each block ends when the block ends, finished or not, and the next one starts anyway. That rule is the entire method. Without it the technical sweep quietly expands to fill the hour.

The skim (0 to 5). Move through the shoot at speed without pressing anything. You are setting your standard for the night: how good was the light, how much of this actually worked. Five minutes of looking makes the next seventeen far more consistent, because you are no longer inventing the bar frame by frame.

The technical sweep (5 to 22). Seventeen minutes over 1,000 frames is almost exactly one second each. That is one look and one key press, so the only question is whether a frame could ever be delivered. Missed focus, unintended blur, shut eyes where the eyes are the point, the frame of the floor. Everything else stays in.

The group pass (27 to 50). Now compare, but only within moments. If your sweep left roughly 400 frames sitting in something like 120 bursts, 23 minutes is about eleven seconds a group, which is enough to look at five near identical frames and take one. Run your own numbers on the way in, because knowing the pace per group is what stops you drifting.

The gap check (50 to 57). Put the shortlist in time order and read it as a sequence. Is anybody missing, is there an establishing shot, is every item on the brief covered. Seven minutes, looking for holes rather than admiring the picks.

What you deliberately do not do during the hour

The sprint is defined by its omissions more than its blocks. Every one of these is a reasonable thing to do on a normal week and a hole in the boat tonight.

• Zooming to 100 percent. The biggest single time sink, and it rarely changes the outcome. A frame you have to inspect at full magnification to defend is not making tonight's gallery.
• Editing anything. Not one exposure slider, not one crop, not to see whether it could work. Culling and editing are different brain states and the switch costs you minutes each way.
• Star ratings. One key, one meaning, binary. Grading things one to five is three extra decisions per frame you do not have time for.
• Renaming, keywording, folder tidying. All of it can happen tomorrow, none of it affects what you deliver.
• Deciding what happens to the frames you did not pick. Nothing happens to them. They stay where they are.
• Album and print selection. A different job with a different standard. Do not let it into the hour.
• The second shooter's card. One card, one hour. Merging two photographers' coverage doubles the group pass and it is the classic way this overruns.
• Comparing across the shoot. Only compare frames from the same moment. The question is never whether this is better than something from two hours earlier.
• Going back. No returning to a block that has closed. If you are second guessing block two during block four, you are tired, not careful.

The binary point is worth dwelling on, because it is where most people lose ten minutes without noticing. Flags force one decision, stars invite a negotiation with yourself about whether this is a three or a four, and nobody has ever delivered a gallery based on that distinction. The longer case is in star ratings versus flags.

Two rules that keep the clock honest

Rule one: in doubt, it stays in. During the sweep, hesitation is expensive and being generous is cheap. A frame you were unsure about costs you a couple of seconds in the group pass later. A frame you agonize over costs you thirty seconds now, and there are eighty of them. Keep it in and let the comparison sort it out.

Rule two: five seconds, then move. Any single decision that takes longer than five seconds is a decision where the options are equivalent. That is what being torn means. Take one, move on, and trust that a client who was not in the room cannot tell which of two near identical frames you picked.

Both rules push the same direction: keep momentum, because the cost of stopping is not the seconds, it is the restart. Once you break rhythm you re-read the last few frames, and the pace never quite comes back.

Why there is a five minute break in a sixty minute plan

Giving up eight percent of the hour to standing in the kitchen feels wrong, and it is the part people cut first. It is also the part that protects the second half.

Research on decision fatigue suggests that repeated choices degrade the quality of the ones that follow. Parts of that literature are debated, so hold it loosely, but the related and better established finding is choice overload: in Iyengar and Lepper's 2000 jam study, people faced with more options were less likely to choose at all and less satisfied with what they chose. A thousand near identical frames is that experiment run on you, at speed, at the end of a long day.

The practical effect is predictable and everyone recognizes it. The last third of an unbroken cull is harsher than the first third, so frames get judged by a tireder, meaner version of you, and the tail of the gallery ends up thinner than the front for no reason connected to the photographs. Stand up, look at something more than two feet away, drink water, do not look at your phone. Then start block three. More on why this part of the job takes so much out of you in decision fatigue for photographers.

What the sprint costs, and the ten minutes tomorrow

Be honest about the trade. An hour buys you a solid, safe, technically clean selection. It does not reliably find the strange frame that is nobody's obvious pick and turns out to be the best photograph of the day. Those come from a slower second look, and you will not have one tonight.

So build in the cheap insurance. Tomorrow morning, before editing starts, spend ten minutes going back through the frames you did not flag, at speed, looking only for that kind of picture. You will usually add two or three and change nothing else. It works because nothing was lost: everything you did not pick is still sitting exactly where it was, which is the point of keeping the photos you do not pick.

Then get the selection out cleanly, because a shortlist stuck in a culling app helps nobody. Export the picks as a collection, a folder of copies, or a filtered set your editing software can read, and check it actually opened on the other side before you close the laptop: exporting selections to Lightroom covers the ways that hand off goes wrong.

When an hour is the wrong plan

The sprint is a tool with a shape, and it is worth naming the jobs it does not fit.

Album and portfolio selection. Both are about sequence and how images sit next to each other, which is exactly the judgment the clock removes. Different job, different evening.

Commercial work with a contractual deliverable. When the brief specifies items, angles and counts, the checking is the work and it does not compress. Cull against the document, slowly, then worry about taste.

Anything much over about 1,500 frames. Four thousand frames is not four sprints back to back, it is four sprints across two or three sittings, and the day divides naturally into blocks that cull like separate shoots. That is a different plan with a different pace: culling a wedding gallery in one evening lays it out.

And the honest structural point. The reason an hour is tight is that the technical sweep, the least creative seventeen minutes of the whole process, is the part a machine is genuinely good at. That is what Kepla is being built to do: the first pass on your iPhone or iPad while you are still out, then the same shoot open on your Mac for the comparisons a phone screen cannot settle. It picks and hands you a shortlist. It never deletes, moves or renames a file, so every frame you did not pick is one tap away. The picking app is still in build. The booking page is live now and free while we build it.

FAQ