Tools
Mira Brunner8 May 202623 min read

Photo Selection Tool for Photographers: AI Culling & Client Picks

A model in a denim shirt and green skirt striding across a coral paper backdrop, lit by a large softbox on the right and a reflector on the left, light stands and cables visible on both sides.
Photo: Patricia Palma / Unsplash

You have culled 2,000 frames down to 200. Now what? For most photographers, the next step is a quiet bet: pick the 50 you think the client or model wants, edit them, and hope you guessed right. When you guess wrong, the revision round begins — a long email thread, a few WhatsApp screenshots, a model asking for "the third one from the second look," and a week you will never get back.

The selection step is the least-tooled part of the photography workflow. Culling has Aftershoot. Editing has Lightroom. Delivery has Pixieset and Pic-Time. But selection — the part where the client or model actually picks which images get edited — is still happening over messaging apps, drive folders, and email replies. It is the workflow gap nobody has properly filled, and it is where most of the friction in a shoot lives.

This guide covers both halves of the problem: the culling software that handles the mechanical first pass, and the collaborative selection step that comes after it — what a selection tool needs to do, how photographers handle it today, and how to set up a workflow that eliminates revision rounds.

What this guide covers:

  • The difference between culling and selection, and why it matters
  • Photo culling software compared: Aftershoot, Narrative Select, FilterPixel, Excire Foto and Photo Mechanic
  • What AI culling cannot decide for you
  • What a good photo selection tool needs, and the five ways photographers handle selection today
  • A step-by-step workflow from AI cull to agreed final selection
  • Common mistakes, how to choose a tool for your shoot type, and an FAQ

What Is Photo Selection (and How Is It Different from Culling)?

Culling and selection are often used interchangeably, but they are two different jobs with different goals.

Culling is the photographer's first pass. You sit down with the full shoot — every frame from the card — and remove the obvious failures. Closed eyes, soft focus, blown highlights, near-duplicates. AI culling software speeds this up dramatically, but the principle is the same: cut the unusable. Culling is technical, fast, and almost always done by the photographer alone.

Selection is the second pass. From the keepers, you choose which images actually get edited and delivered. This is a creative decision, not a technical one. And here is the part most photographers skip: the client or model often has stronger opinions about this stage than you do.

A model knows which expression feels authentic to them. A wedding client knows which guest reaction matters. A brand client knows which frame fits the campaign. The photographer sees composition and light; the subject sees themselves. Both perspectives belong in the selection step, and combining them produces a final set everyone is proud of.

When photographers fold selection into culling — picking the final 50 alone, in Lightroom, before anyone else has seen the shoot — they short-circuit the most collaborative part of the work. The result is more revision rounds, more re-edits, and more "can we swap this one out?" requests.

Culling Software: The Solo First Pass

Culling used to be the least enjoyable, least automatable hour of the job. That changed fast. A 2025 Professional Photographers of America survey found 78% of professionals now treat AI culling as an essential part of their workflow. The tools genuinely work, and if you shoot high volume and are not using one, you are spending hours a week on something a machine does in minutes.

What Culling Software Actually Does

Every AI culling tool does roughly the same three things:

  1. Rejects technical failures. Out of focus, motion blurred, badly exposed, eyes closed. This is the easy, reliable part, and all of them are good at it.
  2. Groups near-duplicates. Fourteen frames of the same pose become one stack with a suggested pick. This is where most of the time saving comes from.
  3. Scores and ranks what is left. Sharpness, face detection, expression analysis. This is the part where tools differ most, and where the results are most debatable.

What it produces is a shortlist of technically competent frames. That is genuinely valuable, and it is not the same as a selection.

The Main Culling Tools

Aftershoot is the most complete package — culling, editing, and retouching in one application, running entirely offline. It won the TIPA World Award for Best AI Workflow Software in 2025 and has a strong following among wedding photographers, particularly for face grouping and culling across multiple shooters on the same event. The trade-off: it wants to own your editing too. If you have a Lightroom process you like, you are paying for half a product.

Narrative Select is the speed king for a first pass where you stay in control of the decisions. It shows a large preview with a focus-point check and close-up on the eyes, which is exactly what you want to see. It is macOS only, and deliberately less automated than the rest — a feature or a limitation depending on what you want.

FilterPixel was fastest in recent benchmark testing, and it is the only tool with genre-specific models: you tell it whether this is a wedding, a portrait session, a concert or a conference, and it loads a model trained for that. Processing is cloud-based, which means uploading your RAWs — a consideration for sensitive work.

Excire Foto is the cheapest serious option by a wide margin, on a lifetime licence rather than a subscription, with strong search and keyword features alongside the culling. It suits photographers who resent subscriptions or need to search a large back catalogue, but it is less specialised at pure culling than the tools built only for that.

Photo Mechanic is not AI. It is still the fastest manual culling experience around, because it reads embedded previews rather than rendering RAWs, and working photojournalists still swear by it. It does no thinking for you — which is the entire point.

ToolApproachPlatformPricing modelProcessing
AftershootAI cull + editMac, WindowsSubscriptionOffline
Narrative SelectAssisted manualMac onlySubscriptionOffline
FilterPixelAI, genre-specificMac, WindowsSubscriptionCloud
Excire FotoAI + searchMac, WindowsLifetime licenceOffline
Photo MechanicManual, fastMac, WindowsLicenceOffline

If you shoot RAW at volume and do not currently use any of these, start with a free trial of Narrative Select or Aftershoot on one recent shoot. The time difference on a single job is usually enough to make the decision for you.

What AI Culling Cannot Decide

AI culling answers: which of these frames are technically good? It is very reliable at this, because sharpness, exposure and open eyes are measurable properties of an image file.

It does not answer: which of these good frames do we want?

Picture the realistic end state of an AI cull on a portrait shoot. You had 900 frames. The software gives you back 120 keepers, all sharp, well exposed, eyes open. Now you have to get from 120 to the 15 you will actually edit and deliver. Those 105 rejections are not technical decisions. They are judgements about which expression is right, which frame looks like the person rather than like a person, which of two almost identical smiles is the real one. No amount of face-detection scoring gets there, because the criterion is not in the file.

And on any shoot with a person in it, you are not the only one with an opinion about the answer. You pick your favourite frame, show the model, and they immediately go to the one beside it — the one you nearly deleted. They are not wrong; they are answering a different question. You are looking at light and at whether the hand reads. They are looking at whether that is how they see themselves.

What usually happens next is bad. You export 120 JPEGs, put them in a folder, send a link, and wait. The reply comes back as a WhatsApp message saying "I like 47, 52, and the one where I'm laughing." You go looking for the one where they are laughing. There are nine.

The failure mode is not that clients give bad feedback. It is that a download link gives them no structured way to give good feedback. Filenames are not a language for talking about pictures.

That remaining decision is what a photo selection tool is for.

What a Good Photo Selection Tool Needs

Most tools sold as "selection tools" are actually delivery galleries with a favourite button bolted on. A real selection tool needs a specific set of capabilities.

  • Thumbnails large enough to evaluate. A model cannot judge their own expression from a 200px square. Selection grids should default to a comfortable size with one click to fullscreen.
  • A favourite or star system the client can actually use. No login walls, no app downloads, no five-step instructions. One tap to mark a pick.
  • Comments on individual photos. "I like this one but can you crop tighter?" should attach to the frame, not arrive in a separate email.
  • Two-sided visibility. The photographer should see the client's picks and vice versa — without anyone needing to copy file numbers between apps.
  • Side-by-side comparison. When two frames are nearly identical — the kind that survive an AI cull — you need to see them next to each other to choose.
  • Mobile-first design. Models review between shoots, on the train, in bed. If the gallery is unusable on a phone, it will not get used.
  • Access control. The client sees the curated set you want them to see — not the raw dump, not images from other shoots.

Most tools nail two or three of these. Few cover all of them, which is why the selection step still feels broken even when you are paying for a "professional" platform.

A useful test: send the gallery link to your own phone, in incognito mode, with no app installed. If marking ten favourites takes more than two minutes, your model or client will not finish the task — they will give up and reply with "I like the ones with the green dress" instead.

5 Ways Photographers Handle Photo Selection Today

Photographers handle selection in five main ways. Each has tradeoffs.

1. Google Drive or Dropbox Folders

The default for photographers who have not adopted a dedicated tool. You upload web-resolution JPEGs to a shared folder, send the link, and ask the client to "reply with the file numbers you want edited."

Pros: Free. Familiar. Handles any file size.

Cons: No favourite system, no comments tied to photos, no way to see two frames side-by-side. Clients have to type file numbers, which they get wrong roughly half the time. Scrolling through 200 thumbnails on a phone in Drive is genuinely painful.

2. WhatsApp or Messaging Screenshots

For casual shoots, photographers sometimes drop selects directly into WhatsApp and ask the model to react with thumbs-up emojis or reply with names.

Pros: Zero setup. Models always respond because they live in the app already.

Cons: Compression destroys image quality, replies are scattered across hundreds of messages, and there is no record of which frame the model meant when they typed "the third one." It works for two friends doing a TFP shoot. It does not scale.

3. Lightroom Web Galleries

Lightroom's web sharing feature lets you publish a collection and invite clients to flag favourites. The flags sync back into your Lightroom catalogue.

Pros: Tight integration with the photographer's existing tool. Selections appear directly in the catalogue you will edit from.

Cons: The client experience is clunky for non-photographers. There is no comment system, no comparison view, and no notion of two-sided selection — only the client picks, the photographer sees the result. Adobe documents the feature on helpx.adobe.com if you want to try it, but most photographers find the friction too high to use regularly.

4. Dedicated Gallery Tools (Pixieset, Pic-Time, picdrop, ShootProof)

These are the platforms most professional photographers know. They are built for delivery — sending finished, edited galleries to clients — but most include a "favourite" feature that gets used for selection too.

Pros: Polished client-facing galleries. Mobile-friendly. Brand customisation. Often the same tool you already use for delivery.

Cons: They are designed around a one-way model: photographer publishes, client favourites, photographer reads the list. There is no shared workspace where both parties select side-by-side, no real comparison view for nearly-identical frames, and no comment threads tied to specific photos. They work for "pick your top 30 from these 100 edits" but not for "let us select the best frames together before I edit."

For deeper comparisons of these tools, see our breakdowns of Picdrop vs Pixieset, Pic-Time alternatives, and ShootProof vs Pixieset.

5. Collaborative Selection Platforms

A newer category designed specifically for the selection step rather than delivery. Both photographer and model mark picks in the same view, with comments, comparison, and a single shared timeline. Cullengo sits here, alongside a small number of tools focused on review rather than handover.

Pros: Two-sided selection by default. Comments per photo. Side-by-side comparison built in. Mobile-first. Designed around the assumption that selection is a conversation, not a one-way poll.

Cons: Smaller ecosystems than the established gallery platforms. You are adopting a tool for one specific phase of your workflow, which means deciding how it slots in alongside your existing culling and delivery setup.

Setting Up a Photo Selection Workflow That Works

Whichever tools you choose, the workflow that surrounds them matters more than the platform. Here is the structure most portrait and editorial photographers settle on after a few shoots.

Step 1 — Back Up, Then Cull Without the Client

Ingest and back up before anything else. Then do the technical pass alone with Aftershoot, Narrative, FilterPixel or Excire — or manually in Photo Mechanic or Lightroom, which takes longer but gives you tighter control. The output should be a set of "keepers": every frame that is technically usable, regardless of whether you personally would choose it.

A typical ratio: 2,000 raw frames → 200 keepers on a long shoot, or 900 → 120 on a portrait session.

Step 2 — Make Your Own Pass to a Shortlist

Cut the keepers down to a shortlist you would be happy to deliver — usually 40 to 60 for a portrait session. This is a taste decision and it is yours. Do not hand the whole AI shortlist to the client: 120 images is not a generous choice, it is an unanswerable question.

Step 3 — Export Web-Resolution Selects

Export the shortlist as web-resolution JPEGs (around 2048px on the long edge is plenty for selection). Full-resolution files slow down loading and make mobile review painful; the client only needs enough resolution to judge expression and composition.

Step 4 — Upload and Share with Clear Instructions

Upload to your chosen tool and send the link with one specific ask. "Please mark your top 15 favourites for editing" is a hundred times more useful than "let me know which ones you like." Set a deadline. Tell them what happens next.

If you are working with a model on a TFP shoot, agree the count up front — both of you should mark roughly the same number, so when your picks overlap you have a natural shortlist.

The single most useful instruction you can include: "Please mark anything you definitely do not want delivered, even if I have favourited it." Models often have boundaries around specific frames — a particular angle, a wardrobe moment that did not work — and giving them an explicit veto saves awkward conversations later.

Step 5 — Review Selections Together

Once both sides have marked their picks, look at the result side-by-side. Where you agree, those frames are locked in. Where one of you picked a frame the other did not, have a quick conversation — in comments on the photo, not over WhatsApp. Often the answer is "include both."

This conversation is the actual point of collaborative selection. Done well, it takes ten minutes and replaces a week of revision emails.

Step 6 — Edit Only the Final Selection, Then Deliver

Once both parties have agreed the final set — typically 15 to 20 frames on a portrait shoot — edit those images and nothing else. The selection is the contract; the edit is the delivery.

The saving from the AI cull in step 1 is time. The saving from the shared review in step 5 is rework: you are not editing 30 frames and finding out afterwards that the client wanted different ones. Our post-shoot workflow guide covers the backup, editing and delivery steps around this in more detail.

How Cullengo Handles Photo Selection

Cullengo is not a culling tool and does not compete with the software above. Run your AI cull first — that is the right tool for the technical pass. Cullengo is built around the selection step that comes after it.

Upload the shortlist to a shoot, and the model or client is a participant in that shoot rather than the recipient of a link. The photo board shows every frame in a grid you can resize, click into for detail, or open fullscreen. Both the photographer's favourites and the model's "suggested" picks appear in the same view, with clear visual markers showing where they agree and where they diverge.

A few specifics that come up often:

  • Two-sided selection is the default, not an extra. You can filter to "selected by both," "selected by photographer only," or "selected by model only" — which makes the conversation in step five take ten minutes instead of an hour.
  • Comments thread directly on each photo. With @mentions and threaded replies, "the one where I'm laughing" becomes a comment on that specific photograph, not a riddle in a separate chat.
  • Side-by-side comparison is built in. When two frames are near-duplicates — the two smiles you cannot choose between — open them next to each other in one click. Similar and duplicate photo detection catches the repeats that slipped through the cull, so nobody is asked to choose between two copies of the same thing.
  • Per-gallery visibility means clients see only what they should. Private galleries keep raw culling work separate from the curated selection set. The model never sees the rejects.
  • Invite link signup gets the model in fast. No app download, no password reset, no "I forgot which email I used." A signed link, a name, and they are in.
  • Delivery mode separates selects from finals. Once you have edited the selection, switch the gallery to delivery mode. The model sees the final edited set, with single or batch ZIP downloads, while the original selection grid is preserved as a record — attached to the shoot it belongs to, alongside the moodboard it was planned from.

The aim is not to replace Lightroom or Aftershoot. Cullengo fills the specific gap between culling and editing, where most photographers currently rely on a patchwork of email, drive folders, and messaging apps.

For a comparison with delivery-focused platforms, see our guide to the client photo gallery for photographers.

Selection is a conversation, not a form

Both sides mark picks, comments attach to individual frames, and a side-by-side view settles the two shots you cannot choose between.

Common Mistakes from Cull to Selection

Culling the day of the shoot. You are too close to it. A day's distance produces better decisions, and AI culling makes the delay cheap because the mechanical pass no longer blocks you.

Trusting the AI's top pick without looking. It is ranking measurable properties. On a portrait, the best frame is frequently not the sharpest one.

Sharing every keeper without curation. Dumping 120 or 200 frames in front of a model and asking them to pick 15 is overwhelming. Pre-curate to your own shortlist first. A focused set produces faster, better selections.

Skipping the conversation step. Collecting selections and disappearing to edit means the model never gets to flag the "please don't deliver this" frames or explain why they preferred one expression over another. Build in a short review call or async comment exchange.

Editing before selection is locked. The most expensive mistake on this list, and the most common. Starting edits on your own favourites before the model has weighed in means re-doing work when their picks differ.

When the finals are ready, our guide on how to send edited photos to clients covers the handover.

What to Look for When Choosing a Tool

If you are weighing options, the criteria that matter most depend on your shoot type.

For wedding and event photographers, volume matters at both ends. An AI culling tool earns its keep on the first pass — Aftershoot is popular here for face grouping and multi-shooter events — and Pixieset, Pic-Time, and ShootProof are built around handing off hundreds of finished images to one client. Selection is secondary.

For portrait, fashion, and editorial photographers, the selection step matters more than the delivery layer. You are picking 15 to 20 from a shortlist of 40 to 60, and getting that right is the difference between a great gallery and a re-edit cycle. A tool with two-sided selection and per-photo comments is worth more than fancy delivery branding.

For TFP and collaborative shoots with models, two-sided selection is non-negotiable. The model has equal stake in the outcome and equal expertise in how they should be presented. Tools that only let one side mark favourites are working against the spirit of the arrangement.

For commercial and brand work, comment threads and version control matter. Brand clients want to discuss specific frames, often with input from multiple stakeholders. Threaded replies on individual photos save the photographer from being a courier between three email chains.

For a deeper dive into the review step itself — terminology, process, and how to get useful feedback — see the online photo proofing guide.

FAQ

What is the difference between photo culling and photo selection?

Culling is the photographer's first pass — removing technically flawed images like out-of-focus frames, blinks, and duplicates. It is usually done alone, increasingly with AI culling software, and produces a set of "keepers." Selection is the second pass — choosing which keepers actually get edited and delivered. Selection is where client or model input belongs, because the choice between two technically excellent frames often comes down to expression, posture, or personal preference rather than craft.

Is AI culling software accurate enough to trust?

For technical rejection — focus, exposure, closed eyes — yes, and it is more consistent than a tired human at frame 900. For deciding which good frame is the right one, no, and it is not trying to.

What is the fastest photo culling software?

FilterPixel benchmarked fastest in recent testing (around 3 minutes for 3,000 photos). Narrative Select is the fastest assisted-manual experience. Photo Mechanic remains the fastest purely manual browser.

Do I still need Lightroom if I use culling software?

Yes, for editing. Most culling tools hand off to Lightroom or Capture One. Aftershoot is the exception in also attempting the edit.

Does Cullengo replace my culling software?

No. It handles the selection stage after the technical cull — the part where the photographer and the subject have to agree.

How many photos should clients select?

Tie the count to your package — if you deliver 15 edited images as standard, ask the client to pick 15. On a portrait session that typically means 15 to 20 picks from a shortlist of 40 to 60; larger editorial or event shoots scale up from there. Setting a clear number prevents over-selection and keeps the editing scope predictable.

Should I show clients all my photos or only my favourites?

Show your curated shortlist, not the raw dump — and fewer than instinct suggests. For a portrait session, 40 to 60 frames is plenty; above roughly 100, people stop choosing and start scrolling. The exception is if the client has explicitly asked to see everything, which is rare and usually a sign of a trust issue worth addressing directly.

What is the best photo selection tool for photographers?

The best tool depends on what you are optimising for. Pixieset, Pic-Time, and ShootProof are strong for delivery-focused workflows where the client picks favourites from finished images. Lightroom web galleries work if your client is comfortable with the Lightroom interface. For collaborative selection — where photographer and model mark picks in the same view before editing begins — Cullengo is built specifically for that step.

Can clients select photos on their phone?

Yes, on any tool with a mobile-friendly interface. Models and clients almost always review on their phones, often between other commitments. Tools that require pinch-zooming, tiny tap targets, or a desktop browser will not get used. Test the mobile experience before you commit.

What happens if the client and photographer disagree on selections?

Usually you include both frames. Editing one extra image is cheaper than the conversation cost of arguing. When the disagreement is real — a frame the client loves but you cannot deliver due to a technical issue, or vice versa — talk it through on a short call. A shared selection tool with comment threads makes this conversation specific and quick rather than abstract and drawn out.


AI culling is a genuine improvement and worth adopting — just be clear about what it solved. It removed the mechanical hour. The decision about which pictures are the right pictures is still a conversation between the person who took them and the person in them. The right selection tool turns that conversation into a ten-minute task instead of a week-long email chain.

If you are looking for a platform built specifically around the selection step rather than the delivery handover, explore what Cullengo offers for photographers and models working together.

Let them choose properly

Free to start. Marked picks on both sides, per-photo comments, side-by-side compare.

Sources: Aftershoot — best culling software, FilterPixel — AI culling benchmarked, Excire — best culling software 2026, Imagen — AI culling tools

Mira Brunner

Mira Brunner

Writes about the craft side of a shoot — proofing, culling, selection, and the software photographers actually keep open.