Card straight to selects on an iPad: the mechanics

An iPad is a good review screen and a poor archive. So the useful question is not whether you can cull on one, it is whether you can get from a full card to a named set of picks without pouring the whole shoot into a tablet you bought for apps. Here is the mechanical version: the two routes off the card, what each one actually records, and where each one breaks.

What "straight from the card" is really asking for

The general case for tablet culling, which apps read RAW, what a tablet does better than a laptop and where it hits a ceiling, is already written up in how to cull photos on an iPad. This piece is narrower and more mechanical. It is about one specific route: card goes into a reader, iPad does a pass, and the whole shoot never lands in tablet storage.

Start with why you would avoid the import. A wedding is 2,000 to 4,000 RAW frames. Even a partial card of those is more than most iPads have spare, and an import that runs out of space partway is worse than no import at all, because you now have a half copy you cannot trust.

So set the target properly. A card-side pass on an iPad should produce three things: a small sneak peek set you can send tonight, a candidate set that tells your desk where to start, and a card that is still completely untouched. If you try to finish the wedding on a tablet in a hotel room you will do it slowly and then do it again.

Two routes off the card, and what each one costs

There are really only two ways to do this. One goes through Lightroom, one goes through the Files app, and they differ in a way that matters: where the decision gets written down.

Route one, a partial Lightroom import. Adobe documents direct import on iPhone and iPad plainly: it "lets you transfer photos and raw images from camera memory cards, cameras, or USB storage devices straight into Adobe Lightroom on iOS without using your computer." The lever that keeps this honest is the selection step. Per Adobe's own help page, you can "select individual photos by tapping them, or select the device name or a date heading to select all photos from that source or date," filter the card view by RAW or by JPEG, and choose the destination album on the way in. You are importing a slice, not a card. The catch is stated on the same page: if your plan includes cloud storage, "the imported photos automatically sync across your devices," so every frame you accept is also an upload.

Route two, browse in place and copy only the picks. Plug the reader in, open the card as a volume in Files, and work through the folders. Nothing is imported at all. When a frame is worth keeping you copy that one file into a SELECTS folder on an attached SSD. Storage cost is close to zero and there is no upload, but you get no flags, no stars and no sidecars. The folder is the rating.

There is a third option that sidesteps the whole argument: shoot RAW plus JPEG and let the iPad see only the JPEGs, which are small enough that the storage question stops existing. That is a workflow of its own and we cover it in the RAW plus JPEG rush workflow. It is the right answer more often than photographers expect.

The order that keeps you out of trouble

The route works or fails on preparation, not on cleverness. This is the sequence, and it is worth doing once at home with a real card before you do it in a hotel at midnight.

1. Set the folder convention before you travel. On the SSD, make YYYY-MM-DD_Client with RAW, SELECTS and EXPORTS inside it. Do this before the job, not during it, because in the field you will use whatever already exists.
2. Copy the full card to the SSD first, then browse. The pass should read from the SSD copy, not from the card, so the card is never the volume you are poking at with tired hands. This is also your second copy, which is the whole point of the backup workflow.
3. Verify the copy by count, not by feeling. Compare file counts between source and destination, then open three or four RAW files from the copy. Thumbnails appearing is not verification, because a thumbnail can be an embedded preview from a file that copied badly.
4. Do the pass. Whichever route you chose, work in capture order and keep moving.
5. Leave the card alone. The card gets formatted at your desk, after a third copy exists, and never because the iPad showed you pictures.

Two hardware notes that are boring right up until they cost you an evening. A reader that is slower than your card turns a ten minute copy into an hour, and you will blame the iPad. And a bus-powered SSD plus a reader on one port can ask for more power than a passive hub will pass through, so test your exact combination of hub, cable, reader and drive at home. The failure mode in the field is a drive that mounts, copies for six minutes and then drops off, which is the single most annoying way to lose an hour on location. Our full field kit list is in the on location culling workflow.

The five things that break, roughly in this order

1. Previews for a new camera. A card-side pass is only as good as the previews the tablet can draw. Support for a brand new sensor's RAW format arrives in system updates and app updates on their own schedule, so a body you bought last month may browse beautifully or may show you gray rectangles. Test with your actual camera before an event, not after.

2. Filename collisions. Two bodies both counting up from DSC_0001 will produce identical filenames in the same SELECTS folder, and the second copy will either overwrite or arrive with a number bolted on. Set a distinct file prefix per body in camera, which takes a minute and saves an ugly afternoon.

3. Storage filling in the background. On the Lightroom route the frames you accept are copied into the app and queued for upload. On a weak hotel connection that queue sits there holding space while you assume it drained. Check that the upload finished before you decide the tablet has spare room.

4. Long copies and a locked screen. Large transfers on a tablet do not love being interrupted. Keep the copy in the foreground, keep the iPad plugged in, and let it finish before you start swiping.

5. No checksum. This is the honest limitation of the whole approach. A desktop ingest tool can verify a copy byte for byte. A tablet file manager compares counts and sizes and that is it. So the iPad copy is a working copy and a convenience, never the copy that lets you format a card. Treat it that way and the route is safe. Treat it as a verified backup and eventually it will cost you a job.

Which decisions a card-side pass can actually make

You are usually looking at an embedded preview rather than a rendered RAW file, so calibrate what you ask of yourself accordingly. Three passes, each one fast.

Pass one, the obvious failures. Black frames, the shot of your own feet, a fired flash that did not, eyes fully closed, a frame where the subject is not in it. These are visible at any size and they are most of the volume.

Pass two, bursts as groups. Move through a burst and take one. Not the best one, one. If two frames are genuinely fighting, take both and settle it at the desk where you can put them side by side at full size.

Pass three, the sneak peek. Fifteen to thirty frames with variety in them: the establishing shot, one detail, one portrait, one moment with real emotion, one wide of the room full of people. This is the set that goes out tonight, and it is the single highest value output of the entire tablet pass.

Keep the vocabulary small. On the Lightroom route: flag equals candidate, three stars equals likely edit, four stars equals sneak peek, and nothing else. On the Files route you have exactly one signal, which is whether the file is in the SELECTS folder, so use subfolders if you need two levels and stop there. Every color label system invented in a hotel room is forgotten by Tuesday. What stays hard on a tablet is the comparison work: near-identical frames, focus at 100 percent, and any color call. That split, volume on the tablet and difficulty at the desk, is the argument in phone versus computer culling.

Making the picks survive the trip home

A field pass only pays if the desk inherits it. If you open the shoot on your computer and your picks are not sitting on the original files, you did not save an hour, you spent one twice. We took that apart properly in syncing picks between devices, which covers what XMP carries, what Adobe's sync carries and what neither of them carries.

For this particular route, the short version. If you used Lightroom with sync on, the flags and stars are metadata and they travel, so the desk job is to check they arrived before you start over. If you used the Files route, your decision is not metadata at all, it is a list of filenames sitting in a folder, and that is more portable than it sounds. Back at the desk, ingest the full card to your working drive as usual, then use the SELECTS folder as a filter: sort by filename, or paste the names into your ingest tool's search, and rate the matches in one sweep. Ninety seconds of clerical work, and the field pass keeps all of its value.

Then finish properly. Copy the card structure to the main working drive, compare counts, spot check files from each body, and only then reuse cards. The full sequence is in the after the shoot ingest checklist. And keep the sneak peek set separate from the real cull. They are different jobs: one is twenty frames chosen for reach, the other is the whole gallery chosen for the story.

The version of this that should not need a folder convention

Read the routes above honestly and the picture is clear enough. The tooling for a card-side pass on a tablet is workable, but every piece of it is something you assemble yourself: your naming convention, your verification habit, your filter trick at the desk. The apps that could do a first pass for you are desktop programs, so on the drive home they are not available at all, and the app that can flag on the tablet has no opinion about which frames are worth flagging. That gap is the reason the first pass so often never happens, and it is why the best culling app for iPhone and iPad is still an open question rather than a settled one.

Kepla is a native app for iPhone, iPad and Mac, and it is still being built. The idea is the one this article keeps circling: it checks every frame against clear criteria, sharp focus, open eyes, expression, near-identical frames, and hands back a shortlist you approve on the tablet, then the same shoot opens on the Mac with every frame it set aside still sitting there for the side by side pass. Nothing is deleted, moved or renamed at any point, which is the same instinct behind never erasing on the card.

The booking half of Kepla is live and free today: a booking page that holds a date and takes a deposit while the picking app is finished.

FAQ