Lightroom culling limitations: what it still can't do

Most articles about Lightroom and culling are out of date, because Adobe now ships an AI selection feature inside the app. So this is the current version: what Lightroom Classic does well for picking photos, the five things it still handles badly, what each of those costs you in real hours, and who should ignore all of it and stay exactly where they are.

Start with what changed: Lightroom culls now

If you last read about this in 2023, the headline has moved. Adobe added Assisted Culling, and it is in both products. In Lightroom Classic you switch it on under Catalog Settings > Metadata > Assisted Culling, after which it appears in the Library module and in the Import window. The separate cloud app, Lightroom, has its own version inside an album.

It looks for the things you would expect. Adobe's own Assisted Culling help page lists Subject Focus, Eye Focus and Eyes Open on the select side, each with a slider you can move, and rejection rules for exposure problems, misfires, and documents or receipts. The results come back as flags, star ratings, color labels or album membership, so the output lands in metadata you already use.

Two honest caveats, both Adobe's own. It is early access, and Adobe says the current feature set is optimized for portraits and headshots. That is a fair thing for a company to say about a young feature, and it tells you exactly where to expect it to wobble: a dance floor at 1/60th is not a headshot.

Everything below describes Lightroom Classic version 15, checked in August 2026. Note that Lightroom Classic and Lightroom are different products, not two names for one app, and the culling experience is not identical in each. If you want the step by step method rather than the critique, we wrote that separately in our Lightroom culling workflow.

What Lightroom is genuinely good at

It would be easy and wrong to frame Lightroom as a failed culling tool. It is not a culling tool at all. It is a catalog and a develop engine that also lets you cull, and judged as that, the selection features are good.

The marking system is fast and complete. Per Adobe's flagging documentation, P sets a pick, X sets a reject, U clears the flag, 1 through 5 set stars, and 6 through 9 set color labels. Hold Shift with any of those and it applies the mark and moves to the next photo, so a first pass can be run entirely with your left hand. Whether you should use flags or stars is a real question, but either way the mechanics are not what is slowing you down.

Nothing you do is destructive. A reject flag is a note in a database. The file on the card is untouched, which means you can be decisive at midnight and still change your mind on Tuesday.

It is where the work continues. This is the underrated part. Every dedicated culling app in existence has to hand its answer back to Lightroom eventually, and that handoff is a step that can go wrong. Lightroom has no handoff, because the picks are already in the catalog you are about to edit from. If you use an outside tool, how reliably its selections travel back into the catalog matters more than anything on its homepage.

It only exists where your desk is

Lightroom Classic runs on Windows 10 or 11 and on macOS Sonoma or later. That is the complete list. There is no Lightroom Classic for iPad or iPhone, and there never has been. Adobe's mobile apps are the cloud-based Lightroom, a different product with a different catalog model, and Adobe's own Assisted Culling overview describes the feature in Lightroom Classic's Library and Import windows and in Lightroom desktop, not on a phone.

Here is the cost, and it is the largest one on this page. A wedding ends and you have a drive home, a Sunday morning, a Monday commute. None of that time can touch the job. The work cannot begin until you are physically at the machine, which in practice means late Sunday night, which is precisely when your judgment is worst. The bottleneck is not the app's speed. It is the app's location.

Photographers who have solved this have almost always solved it by moving the first pass to a device they already carry, and keeping the desk for the comparisons that genuinely need a big screen. We looked at where each pass actually belongs in phone versus computer culling.

You cannot look at a photo until Lightroom has drawn it

Lightroom does not show you a RAW file. It cannot: a RAW file is sensor data, not a picture. It shows you a preview it built, and building previews for 3,000 frames takes real time on any machine. The choice you make at import decides how much of that time you spend waiting versus how much you spend looking.

Those descriptions are Adobe's, from the import options documentation. There is also Library > Previews > Build Smart Previews, which makes a lightweight file based on lossy DNG so you can keep working with the originals offline. Smart Previews are excellent and underused, and we go into them properly in the Smart Previews guide.

The limit is not that these settings are bad. It is that none of them make the tax zero, and the wrong combination makes it enormous. If your Lightroom feels like treacle right now, the cause is usually in this area and it is usually fixable in ten minutes, which we go through cause by cause in why Lightroom slows down when you cull.

Comparing similar frames is still a two-photo job

This is the limit nobody mentions and everybody feels. Modern culling is not mostly about spotting the bad frame. It is about choosing between nine frames that are all fine.

Lightroom gives you two tools for that. Compare view (C) shows exactly two images, one designated the select and one the candidate, and you promote the winner and pull in the next challenger. Survey view (N) tiles two or more, which is better for choosing a mood but worse for judging focus, because each tile is now a fraction of your screen.

So a nine-frame burst becomes eight consecutive pairwise duels, or one tiled screen too small to judge sharpness on. And if the question is "which of these three has the sharper eye", you need to be at 100%, which means those frames need 1:1 previews, which brings back the previous limit at exactly the moment you were making progress.

None of this is a bug. Compare view was designed when a photographer might shoot 400 frames at a wedding, and it is still the right design for that. It just does not scale to the way mirrorless bodies shoot now. If bursts are where your evening disappears, the technique matters more than the tool, and we set out a method in choosing the best frame from a burst.

It measures sharpness, not the moment

Assisted Culling scores subject focus, eye focus and open eyes. Read that list again and notice what it is: the mechanical half of the job. It is the half a machine can do, and having it done is genuinely useful, because it is the half that is boring and the half where human attention degrades after the first hour.

What no model in this category can tell you is that the slightly softer frame is the one where the bride's mother finally cried, or that the technically perfect group shot is the one where the light went flat. Adobe does not claim otherwise, and neither do its competitors when you read their documentation rather than their homepages.

The practical consequence is that a clean Assisted Culling pass hands you a shortlist that is technically sound and creatively unsorted. That is a big time saving and a small trap, because a shortlist looks finished. It is not. It is the pile you now do the actual work on. We wrote about why the sharpest frame so often loses in why your best photo isn't the sharpest.

Culling is one job in an app that does forty

The Library module is the front end of a database. It has to serve the person searching for a headshot from 2019, the person keywording a stock archive, the person building a collection for a magazine submission, and you, tonight, trying to make three thousand small decisions quickly.

That shows up as friction that is individually trivial and collectively expensive: panels you close and that reopen, a filter bar you have to configure, a filmstrip you scroll, modules you switch between. Each one costs a second and a small amount of attention. Multiply by the number of frames in a wedding and you have the reason Photo Mechanic has kept a loyal following for twenty years despite doing far less, which we covered in Photo Mechanic versus Lightroom.

When Lightroom is still the right answer

Plenty of people should read all of the above and change nothing.

If you shoot under about 500 frames a session, the previews finish while you make coffee and the two-photo comparison problem barely exists. Adding another app to the stack would cost you more than it saves.

If your shoots are portraits or headshots, you are the exact case Adobe says Assisted Culling is currently tuned for. Switch it on, run it against a job you have already finished, and judge it on your own files.

If your real bottleneck is editing, not selecting, none of this is your problem. A UK survey of more than 300 wedding photographers, reported by PetaPixel, found 55 percent of working time going to editing against 11 percent to culling. Fix the larger number first.

The photographers who genuinely outgrow Lightroom for this one task tend to share a profile: high frame counts, heavy burst shooting, and a calendar where the gap between the shoot ending and the desk is measured in days rather than hours. If that is you, the answer is not to abandon the catalog, because the catalog is still where the editing lives. It is to do the first pass somewhere else and arrive at Lightroom with a shortlist instead of a card. That is the shape of the app we are building at Kepla: it picks, it never deletes, moves or renames a file, and everything it did not pick stays one tap away. The booking page is live and free today while the picking app is still in development.

FAQ