Build your culling criteria: a worked example

Most photographers already have culling criteria. They are just not written down, so they move: stricter on a Monday morning, softer at midnight in October. One page fixes that. Here is how to build that page, layer by layer, with a full worked example you can copy and adapt to your own work in an afternoon.

Why two galleries from the same photographer look different

You pick a January gallery carefully. You are fresh, you have three days, and you look at every frame twice. You pick an October gallery at eleven at night with two more shoots waiting, and you are generous, because being generous is faster than deciding.

Both galleries go out under your name. They do not look like the same photographer's work. The shooting was consistent, the standard was not.

This is not a discipline problem. Judgment made from memory shifts with fatigue, with the order you happen to view frames in, and with how much of the shoot you have already seen. Picking is a real skill with its own failure modes, which we go into in why culling is a skill, not a chore. The fix is not trying harder. The fix is moving the standard out of your head and onto a page, so that the tired version of you is applying the same test as the rested version.

It is also worth the time. A UK survey of more than 300 wedding photographers, reported by PetaPixel, put culling at 11 percent of working time and editing at 55 percent. That was 2020, before the current crop of tools, but it tells you the stage is large enough to be worth a page of writing.

What a criteria set actually is

It is one page. Not a style guide, not a shot list, not a manifesto about your artistic values. It is a short list of tests you can apply to a single frame in about two seconds, arranged in three layers.

• The floor. What any frame must clear before it can be considered at all. Technical, boring, mostly objective.
• The frame. Which one of the six nearly identical shots of the same moment goes in. This is where your taste actually lives.
• The gallery. How many photos, covering what, in what balance. The layer almost nobody writes down and the one clients feel most.

One rule governs the whole page: write tests, not adjectives. "Sharp enough" is not a criterion, it is a feeling with a word attached. "At 100 percent, the eye nearest the camera holds detail in the lashes" is a criterion, because two people can apply it and agree. If you cannot check a line in two seconds at normal viewing size, rewrite it or cut it.

Keep the whole thing under one screen. A page you will not reread is the same as no page.

The floor: what any frame has to clear

Start here because it is the easy half, and because it is the half you can hand to software. Write four to six lines, each one a pass or fail.

Be specific about focus rather than describing it. Name the thing that has to be sharp, name the viewing size, and name the exception. Do the same for eyes, for motion blur, for exposure headroom, and for framing errors like a cut hand at the edge. Anything you find yourself arguing with belongs in layer two instead.

The reason to write this layer even though it feels obvious: it is the layer machines now handle. Adobe's Assisted Culling in Lightroom Classic, labeled Early Access on that page, scores subject focus, eye focus and whether eyes are open, and can set aside exposure problems, misfires, and documents and receipts. Dedicated tools have done versions of this for years, and we cover what the detection actually sees in how blur and blink detection really work.

If your floor is written as tests, you can configure a tool to approximate it and then check its work against your own words rather than against a vague sense that something looks off. If it is written as adjectives, you cannot.

The frame: choosing between near identical shots

This is the layer that makes a gallery feel like you, and it is the layer you disagree with yourself about most, because it is a ranking problem rather than a pass or fail one.

So write it as a ranking. List your tie-breakers in priority order and commit to the order, not just the list. Almost everyone's list contains expression, sharpness, hands, background and timing. What separates two photographers is which one wins when they conflict, and if you never write the order down you will resolve it differently at eight in the evening than you did at ten in the morning.

Most photographers, once they actually look at their own delivered work, find they rank expression above technical perfection, which is the argument in why your best photo is rarely your sharpest. Write that down if it is true for you. Write the opposite down if it is not.

Add one arbitrary rule for genuine ties: keep the earlier frame, or keep the wider one. It sounds unserious. It saves you the twenty seconds you would otherwise spend on a choice that has no consequence, and across a wedding those twenty seconds happen a few hundred times.

The gallery: how many, of what

A gallery can be full of good individual photos and still be wrong: forty frames of the first dance and none of the room, or a portrait session where nine of the twelve delivered images are the same pose.

So the last layer is about balance and count. It answers a small set of questions in advance, once, so you are not answering them per gallery at midnight.

Typical ranges help you sanity check your own numbers: a full wedding day is commonly 2,000 to 4,000 RAW frames captured and 400 to 800 images delivered. If your page says 1,400 for eight hours, that is a decision you should make deliberately rather than by accident, and how many wedding photos to deliver works through the tradeoff.

A worked example you can copy

Here is a complete criteria page for a full-day wedding photographer. It is deliberately opinionated, because a page that tries to suit everyone decides nothing. Copy it, then change every number and every ranking that is not true of your work.

To adapt this: a newborn session swaps the family group rule for a caregiver and safety rule and drops the cap on setups, since the setups are the session. Corporate headshots collapse layer three to a fixed number per person. Sports and events push the ranking so that peak action outranks expression, which flips the whole page. The point is that the structure survives the change of genre, and only the lines inside it move.

Test the page against work you already delivered

A criteria page written in the abstract will be wrong, and you will find out on a live gallery, which is the expensive way to find out. Test it instead against a shoot you have already sent.

Take a gallery from six months ago that you were happy with. Open the full take, not the delivered set, and apply the page as written. Then compare the two selections and count the disagreements. Each one falls into exactly three buckets.

1. The rule is wrong. The page rejects a frame you are certain about. Change the rule, and be honest that you are changing it because of the photo, not the other way around.
2. The rule is missing. Past you applied a standard that is not on the page at all. This is the most valuable bucket. Write the missing line.
3. It is an exception. A real one, once. If the same exception appears three times in one gallery, it is not an exception, it is a rule you have not admitted to.

Run this on two galleries and the page usually stabilizes. Run it again at the end of the season, because your work moves and a page written for the photographer you were two years ago will quietly hold you there. Consistency across a body of work is its own skill, which is the subject of culling for a cohesive portfolio.

Where the page lives once it exists

A criteria page that lives in a note app you never open has no effect. Put a copy in your folder template so it is next to the RAWs, and pin it wherever you actually make the picks.

It also becomes the thing you hand to anyone else who touches the take. Most second shooter selections are not bad, they are blind, and the fix is the page rather than more feedback. That handover has its own considerations, covered in second shooter culling standards. The same page is what you use to set up whatever software makes your first pass, and it gives you something concrete to test that software against on a shoot you have already picked by hand.

One thing your page should never do is send a frame to a bin. A photo that fails section 1 is simply not in the gallery, and it stays in the folder with everything else, which matters more often than people expect: clients ask for the frame you passed over, and albums need the wide shot you did not pick. What to do with the rest of the take is covered in what to do with the photos you do not pick.

Kepla is being built to work this way: it picks a shortlist against criteria like these and leaves every original exactly where it is, so nothing you did not pick is ever out of reach. The picking app for iPhone, iPad and Mac is still in development. The booking page is live and free today.

FAQ