Guides · Wedding Photography · 9 min read · April 28, 2026

Cull 3,000 wedding photos in 90 minutes.

The keyboard-first workflow that gets you from 3,247 raw frames to a 387-photo client gallery in 90 minutes flat — no burnout, no second-guessing, no mouse.

3,247
Frames in
90 min
Cull time
387
Final gallery
0
Mouse clicks

The 4-round funnel

Start
3,247frames
↓ Round 1 — reject obvious garbage · 12 min · −487
After R1
2,760surviving
↓ Round 2 — mark keepers · 38 min · pick ~40%
After R2
1,124picks
↓ Round 3 — hero per moment · 28 min · collapse bursts
After R3
412heroes
↓ Round 4 — final pass · 12 min · story flow
Final
387delivered
TL;DR — the 4-round recipe
  1. Round 1 (12 min): reject obvious garbage. X on closed eyes, motion blur, accidents.
  2. Round 2 (38 min): mark the keepers. P on every shot you'd be proud to deliver.
  3. Round 3 (28 min): pick the hero from each moment. Bursts collapse to one winner.
  4. Round 4 (12 min): final pass — exposure, story flow, sneak-peek picks.
  5. Export: XMP sidecars to Lightroom. Editing starts in seconds.

Why culling breaks wedding photographers

Shooting takes 8 hours. Editing 400 photos takes 6. Culling the 3,000 frames between them? That's the part that destroys photographers. Most spend 4–6 hours on culling alone, then start editing already exhausted — second-guessing every selection because they're tired.

Faster culling isn't about clicking faster. It's about removing decisions you shouldn't be making at this stage. Color correction is not a culling decision. Crop is not a culling decision. "Is this my best work?" is not a culling decision.

The only question that matters: would I deliver this to the client? Yes or no. Move on.

And the tooling makes it worse. Lightroom renders the full RAW for every preview — painfully slow on a 3,000-frame shoot. Photo Mechanic is fast but locks you into a $150/yr subscription. Cloud AI culling apps ask you to upload a couple's wedding to someone else's servers. Finder + Quick Look collapses past a few hundred files. There's never been a free, fast, offline, Lightroom-compatible culling app built for macOS — which is why most wedding photographers either pay forever or suffer.

Setup before you sit down

Three boring things before you open the photos. Skip them and you'll waste 15 minutes mid-cull rebuilding context.

1. Ingest into a single folder

Pull both card slots into one folder named YYYY-MM-DD_couple-name. One folder per wedding. Multiple folders fragments your culling brain and you'll triple-handle frames.

2. Back it up before you cull

Two copies, two locations, before any culling. Backblaze plus an external SSD works. The reason isn't paranoia — it's psychology. When culling moves fast, you'll start hesitating before reject, worried you'll lose something. Knowing the originals are bulletproof lets you reject without flinching.

3. Pick your app — and stay on the keyboard

ShotSelect is built for this. So is Photo Mechanic ($150/yr). Lightroom can do it but the RAW preview generation is slow and the mouse-driven UI breaks rhythm. Whatever you pick, the rule is the same: both hands on the keyboard. The mouse is a trap.

Round 1 — Cut the obvious garbage (12 minutes)

Open the folder. Move through the entire shoot at full speed. to advance. The only key you press otherwise is X (reject) on:

Do not reject because the colors are off. Do not reject because the crop is awkward. You're not editing yet. Single rule: is this an unrecoverable mistake? Yes → X. Maybe → skip.

Expect to cut 400–600 here. Round 1 should land in 10–14 minutes. If you're slower, you're thinking too hard.

Round 2 — Mark the keepers (38 minutes)

Back to the start. This time you're answering one question per frame: would I be happy delivering this?

If yes, P (pick). If no, . Don't agonize. The mental rule that helps: if I'd hesitate to put this in their gallery, it's a no.

For a typical wedding, expect ~40% of remaining frames marked as keepers. So from ~2,700 surviving photos, ~1,000–1,200 picks. That's normal. If you're picking 70%, you're being too generous and Round 3 will hurt.

Tip: Don't zoom to 100% during Round 2. The pixel-peep is for after culling, when you've narrowed to ~400. Zooming on every frame doubles your time and almost never changes the verdict.

Round 3 — Pick the hero from each moment (28 minutes)

Most photographers skip this round. It's where you save your editor self the most time.

Filter to your P picks. Group them by moment — first kiss, ring exchange, bouquet toss, first dance. Wedding photographers naturally shoot in bursts of 3–10 frames per moment. Pick one. Maybe two if the second is a genuinely different expression.

Spot bursts by sequential filenames and matching scenes. From each burst, demote all but the hero — remove the P from the runners-up. Your remaining "picks" set is the gallery.

From ~1,100 picks, expect to land at ~380–420 heroes. That's your client gallery.

Round 4 — Final pass (12 minutes)

Filter to your hero set. Walk through one more time, looking for:

Optionally rate the absolute heroes 5 stars — your sneak-peek shots, the ones for Instagram and the album cover. About 15–25 per wedding.

Color labels turn culling into a delivery dashboard

Star ratings answer "how good?". Color labels answer "what's its state?". The combo is what separates a pile of picks from a deliverable gallery.

A workflow that scales for working photographers:

Tap a number key during Round 4. When you start editing in Lightroom the next morning, filter to Green only — that's your fast-path gallery. Filter Red when you have time and energy for the rescue work.

Color labels travel through XMP. Lightroom and Capture One read them natively, so this segmentation survives the handoff. Your culling app becomes a delivery project tracker — without ever leaving the keyboard.

Exporting to your editor

Don't move files. Export XMP sidecars — small text files Lightroom and Capture One read alongside your RAW. Ratings, picks, rejects all transfer in seconds.

In ShotSelect: ⌘E writes XMPs in place. Open Lightroom, sync the folder, and the gallery you picked is already filtered. Editing starts immediately.

New to XMP sidecars? See our guide on how XMP sidecars work — it's the unlock for treating culling and editing as separate apps.

Three pitfalls that double your culling time

1. Color-correcting mid-cull. You're not editing. Reset white balance to camera default, leave it, move on. Editing while culling triples your time and burns you out.

2. Showing the bride mid-cull. Don't. Cull alone, deliver clean. Otherwise every "oh keep that one" balloons the gallery and your edit time.

3. Re-doing rounds. Once Round 2 is done, don't go back. Trust your earlier self. The 5% of shots you'd retroactively change wouldn't make the final gallery anyway.

The numbers from a real shoot

Last weekend's full-day wedding: 3,247 frames in. Culling timer: 89 minutes 41 seconds. Final gallery: 387 photos. Round breakdown:

Editing those 387 takes ~6 hours instead of the 11–12 it would after a sloppy cull, because every frame in the set actually deserves the work.

Why ShotSelect for wedding culling

  • Free, forever. Photo Mechanic is $150/yr — that's a wedding's worth of profit gone.
  • On the free tier, 100% offline. Your couples' photos stay on your Mac. No cloud upload, no analytics, no AI training on their wedding.
  • Native Apple Silicon. Cull a 50MP RAW set at full speed on M1+. No spinning beach balls.
  • Lightroom-ready. Writes XMP sidecars natively — your picks and ratings show up in Lightroom in seconds.
  • Keyboard-first. The whole 4-round workflow above is mappable to P / X / 1–5 — no mouse hunting.

Try this workflow with ShotSelect

Free, keyboard-first. RAW + XMP sidecars. Built for exactly this.

Download for macOS


Keep reading