The photo culling checklist: pick or set aside

This is a working checklist for picking photos after a shoot, split into three passes so each one asks a single question. One column is what you pick. The other is what you set aside. Set aside means marked and left exactly where it is, because at no point in this checklist does a photo get deleted. Print it, or keep it open on your phone while you work.

Two columns, and neither of them is a bin

Most culling checklists you will find are written as two columns, keep and bin, and that framing quietly makes the job harder than it is. If the second column is a bin, every decision carries a consequence, so you slow down, you hesitate on the ones you are unsure about, and at midnight you either keep far too much or make a call you regret in March.

Change the second column and the whole job gets faster. This checklist has a pick column and a set aside column. Set aside means the frame is marked as not picked and stays exactly where it is, in the same folder, at full resolution, ready to be looked at again in five minutes or five years. Nothing is destroyed at any pass. That single change is what lets you move quickly, because a decision you can walk back does not deserve thirty seconds of thought.

The three passes each ask one question, in this order: is this frame usable, is this the best frame of this moment, and is anything missing. Run them separately. Mixing them is the most common reason a cull takes all night, and it is on the list in our guide to the culling mistakes that take years to show up. If you want the wider theory behind the passes rather than the checklist itself, start with the complete guide to photo culling.

The setup that makes the rest fast

Five minutes here saves an hour later. Do all of it before you look at a single photo.

• Back up first, in two places. Cards to a working drive, and to a second copy somewhere else, before you start judging anything. Never pick photos while a single copy exists. Our backup workflow guide covers the order to do this in.
• Build previews and let them finish. Waiting three seconds per frame for a preview to render is what turns a two hour job into four.
• Decide how you will mark a pick. A flag, a color label, whatever your software uses. One key for pick, one key for set aside, both reachable with the same hand.
• Do not work on shoot day. The morning after is faster, because you are judging the photo on screen rather than the memory of taking it.
• Set a stop time. Two blocks of forty minutes will beat one block of three hours, every time.

Pass one: is this frame usable

One question, asked at fit to screen, answered in about a second. Do not zoom. Do not compare. Do not rank anything yet. You are drawing a floor, not choosing favorites.

Pick when

• Focus landed on the subject, at the point you meant it to land
• The eyes are open on the people the frame is actually about
• Exposure is recoverable in raw, even if it looks wrong right now
• Something readable is happening: an expression, a gesture, a moment, a detail you were asked for

Set aside when

• Focus landed behind or in front of the subject and no crop saves it
• Motion blur you did not intend, on the part of the frame that matters
• The only subject in the frame blinked
• Highlights are gone with no detail left on a face
• Test frames, misfires, lens caps, the floor, the inside of your bag

Two habits protect this pass. Go forward only: no scrolling back to reconsider, because the set aside column is not permanent and reviewing twice doubles the job. And stay off the zoom key. Checking eyelash sharpness on a frame you have not decided you like yet is the single most expensive habit in culling, and it changes almost no decisions.

Pass two: one frame per moment

Everything that survived pass one is usable. Now you are choosing between frames that are all fine, which is a different and slower kind of thinking, so group first and decide inside the group. Six frames of the same hug is one decision, not six.

Pick the frame where

• The most faces are open eyed, prioritizing whoever the photo is about
• The expression is peaking rather than arriving or fading. A smile on the way up beats the same smile on the way down
• Hands and feet are doing something explicable
• The background is cleanest, or at least nothing is growing out of anyone's head
• If two frames are still tied, the sharper one wins. Sharpness is a tiebreaker, never a ranking

Set aside

• The rest of that burst, once you have your one
• Second and third versions of the same static setup where nothing changed but your framing

One warning. Frames that look near identical in a grid are sometimes not near identical at all: two people reacting to the same joke half a second apart are two different photographs. Group by moment, not by thumbnail similarity. The tiebreakers for genuinely close calls are in choosing the best frame from a burst.

Pass three: the pass that adds photos back

The first two passes only ever subtract. This one exists to catch what that leaves out, and it is the reason nothing gets deleted along the way: pass three regularly needs a frame you set aside an hour ago.

Work from the day rather than from the folder. Read down a coverage list and check that each item has at least one picked frame:

• Getting ready, both sides
• Details you were asked for: rings, dress, shoes, stationery, the room before anyone is in it
• Ceremony wide, ceremony tight, and the faces watching it
• Every group on the family list, checked name by name against the list itself
• Couple portraits, and the venue as it looked that day
• Speeches, cake, first dance, and the room during each
• Every guest at least once, and every vendor who worked the day
• The exit, and whatever happened after it that nobody planned

Where an item has no picks, go back into the set aside pile and take the best frame that exists, even if it fails the bar you were holding in pass one. A slightly soft photo of something that happened beats a perfect photo of nothing, and the grandmother who appears in exactly one usable frame all day is not a candidate for your standards. This is also where you sanity check volume against the length of the day, which we go through in how many wedding photos to deliver.

The printable version

This is the part worth printing or keeping on your phone. Everything above, condensed into the two columns and the pace each pass should run at.

The final read is the only time you look at your picks as a group rather than as individual photos. Three frames that each earned their place can still be three frames of the same thing when they land next to each other in a gallery. If you find yourself repeatedly unsure at this stage, the problem is usually upstream: your criteria are not written down anywhere, and building explicit culling criteria is a better fix than a fourth pass.

Where the set aside photos actually live

They live where they already were. Mark them with a flag or a color label and leave the files in the shoot folder at full size. Do not move them into a rejects folder, do not rename them, and do not clear them out at the end of the job.

The reasons are practical, and every one of them has happened to a working photographer. A client emails in March asking for anything at all with her father in it, because he died in February. A magazine wants a horizontal version of a frame you picked vertical. A family member was in the background of forty frames and in exactly none of your picks. You look at a wedding two years later with a changed eye and want back three photos you passed on. In each case, the answer is either one search away or gone forever, and which one depends on a decision you made at midnight while tired. There is more on managing that pile in what to do with the photos you don't pick.

If storage is the real worry, the answer is archiving rather than erasing. Move the finished shoot, picks and set aside frames together, onto slower and cheaper storage once the gallery is delivered. Drives are cheap compared to the one email you cannot answer.

Running the checklist without sitting at a desk

Pass one is the pass a machine should be doing. It is a fixed test applied identically to every frame, it is where most of your evening goes, and it is the only pass with objectively right answers. Pass two is partly automatable, since grouping bursts is something software is genuinely good at. Pass three is entirely yours, because no tool knows there was a family list or who flew in from three time zones away.

Plenty of good software does the first part. Aftershoot and Narrative Select are desktop apps that run the mechanical pass locally, Photo Mechanic makes you faster without deciding anything for you, and Imagen offers picking as an add-on to its editing service in the cloud. If your workflow already starts at a desk, any of those will shorten this checklist.

Kepla is being built on the assumption that pass one should not require the desk at all. The picking app for iPhone, iPad and Mac is still in development, and it is built on the same promise this checklist is: it only marks what it picked. No file is deleted, moved or renamed, so the set aside column stays one tap away on every pass and on every screen. The part that is finished today is the booking page, which is live and free while we build the rest.

FAQ