AI Photo Culling for Event Photographers
BetaAI does the first pass, you decide.
AI culling is in beta: the engine is built, and the panel screen that starts and reviews a cull is not open to studios yet. Before delivery, an automatic first pass runs over your frames: weak frames are flagged and bursts are grouped. No frame is rejected without your approval.
This feature is in beta: parts of it are not open to every studio yet. The documentation says exactly what works today.
What it is
AI culling measures each professional frame of an event: sharpness, exposure, faces, how open the eyes are, and frames that are almost identical. From those measurements each frame gets one of three results: keep, suggestion or reject. Suggestions are kept unless you reject them.
Culling is preparation, not a verdict. When you change the strictness the results are worked out again, and on any frame you can replace the rule's decision with your own. When you approve, the kept frames are transferred; rejected frames are not transferred, but they are not deleted.
How it works
Upload your frames
Upload the event's professional photos. Before you start culling you see how many frames there are and how many credits it will use.
Choose the strictness
“Light” flags only obvious faults. “Medium” is the recommended level. “Strict” uses the tightest thresholds and keeps only the best frame of a burst.
Get notified when analysis is done
The frames are measured in the background; when it finishes, a notification in the panel tells you the cull is ready. You can do other work in the meantime.
Review and correct
Look at the rejected and suggested frames, each with its reason. Bring back or reject any frame you like. “Target frames” lets you bring the remaining set down to a number.
Approve
On approval the kept frames are transferred. Rejected frames are not transferred and remain in storage.
What guests see
- Guests never see this screen; culling is the studio's work.
- Meets fewer blurry and eyes-closed frames in the delivered gallery.
- Sees the best frames of a burst instead of ten copies of the same pose.
- Gets selfie search results from a smaller, cleaner set of frames.
What the studio controls
- Chooses the strictness, light, medium or strict; changing it recalculates the results.
- Accepts the rule's decision for each frame or overrides it.
- Can reject all suggestions in one action.
- Sets a target for the number of frames left; the lowest-scoring frames beyond the target are rejected.
- Approves or cancels the cull.
Limits and details
- It costs 1 credit for every 100 frames started, with a minimum of 10 credits. If no frame can be measured, the credits are refunded.
- Rejected: empty or black frames, frames where the flash misfired, blurry frames and test or grey card shots. Closed or half-closed eyes, very dark or very bright exposure, a face cut by the frame edge and an open mouth are flagged as suggestions.
- In a burst, consecutive frames that are almost identical are grouped; the best 3 are kept on light, the best 2 on medium and only the best one on strict.
- Each frame gets a score from 0 to 100 based on sharpness, open eyes and exposure.
- No frame is deleted. The face data used to find the main people is deleted when the analysis ends.
Which events use it
It saves the most time at weddings, marathons and corporate events, where thousands of frames are shot in a day. For studios that use selfie search it also saves credits: only the frames you will deliver are made searchable.
Planned additionsPlanned
- Highlights: the best 50 frames of the event
Frequently asked questions
Does the AI delete my photos by itself?
No. No frame is deleted. Culling only flags; approval is yours, and rejected frames also remain in storage.
Which faults are flagged?
Empty frames, flash faults, blurry frames, a main person with closed or half-closed eyes, very dark or very bright exposure, a face cut by the frame edge and surplus copies in a burst. On strict, looking away and a tilted horizon are flagged too.
How many credits does AI culling cost?
1 credit for every 100 frames started, with a minimum of 10 credits. For example, 3,000 frames cost 30 credits.
Are frames I shot dark on purpose rejected?
Dark frames are not rejected; they are flagged as suggestions and kept unless you reject them. On light, a frame that is dark but sharp is not flagged at all.
Can I change the cull's decision?
Yes. You can keep or reject every frame individually, and your decision takes precedence over the rule.
Does culling reduce the cost of selfie search?
Yes. Selfie search on professional photos costs 1 credit per photo; if you bring 3,000 frames down to 500, you pay for 500.
Related features
Designs from DesignAsist
QR table cards, posters, event covers, slideshow themes and share frames are designed in DesignAsist, so the event looks like one piece.
See the featureYour brand in front
Guests see your studio's logo and colour on the cover, in the gallery and on shared photos. CollectAsist stays in the background.
See the featureMessages to the studio
A question about an order or a missing photo? Guests write to you from the event page and you reply from your panel.
See the feature
Put the QR on the table this weekend.
Open your account, set up your first event with the welcome credits, and see how your guests use it.
40 welcome credits when your account is approved. No subscription.