RAW workflow for train photos: from the memory card to the finished image

Updated: August 2026 – After a good spotting day, three to four hundred shots quickly pile up on the card, and that’s exactly where most people fail: they start developing at image one and give up by image sixty. This RAW workflow for train photos flips the order – reduce first, develop next, export last.

What a RAW workflow does – and what it doesn’t

A RAW workflow is a fixed sequence of steps, not an editing aesthetic. It ensures that every shot is backed up, findable and editable in a reasonable amount of time – whether you’re dealing with twenty images or four hundred. At its core is a funnel: each step reduces the pile for the next.

What it doesn’t do: it doesn’t turn a bad image good. Focus errors, a pole in front of the nose, or the wrong light direction remain problems with the shot itself. The workflow only makes sure you spot those images early and don’t waste time on them.

The funnel: from 400 to 20 images

The central insight is: the most expensive step is individual editing – so it happens last and on as few images as possible. In practice, the funnel looks like this:

• All shots are imported and backed up. Nothing gets deleted here, not even the obvious failures.
• First pass: binary decision, technically usable or not. Seconds per image, no zooming, no debate.
• Second pass: among the survivors, pick the best variant per subject. Here you compare, you don’t score.
• Batch development: one base setting per lighting situation applied to the whole group.
• Fine work: only on the images that will actually be published or archived.

If you’re wondering which shots even belong in the second pass, the content criteria are in the guide to train photos and railway photography and in the detailed guide to trainspotter photos.

The process in six steps

Backup: the step with no exceptions

Of all the steps, backup is the only one whose absence can’t be fixed. A botched image look can be re-developed, a wrong rating corrected – a lost file can’t. That’s why it sits right after import and before any culling.

Three basic rules are enough for the hobby scale:

• Two copies on two different storage media. Two folders on the same hard drive aren’t a backup, they’re two copies of the same risk.
• Check first, then format. Open a random sample of files from the backup before wiping the card. Formatting then happens in the camera, not on the computer.
• Backup runs automatically or not at all. A process that demands a conscious decision every time eventually gets skipped – usually right after your best trip.

Anyone who also keeps a copy in a separate physical location is protected against theft, water damage and hardware failure. For most spotters, a removable drive that isn’t permanently plugged into the computer is enough.

What happens to the rejected images?

The honest answer is: they get to stay. Storage space is cheaper than the regret of a deleted image that, years later, turns out to be the only record of a sighting – say, a vehicle in a livery that’s since disappeared. A simple compromise is not to delete rejected shots but to move them into their own subfolder and hide them from the catalogue. They stay out of your way while working, but they’re there if you ever need them.

It’s different for shots that are technically clear rubbish: completely shaky images, misfires and shots with no subject at all can go straight away. Anything showing a recognisable vehicle, on the other hand, is potentially a record worth keeping.

How much time do you actually need?

The duration depends almost entirely on how strictly you cull – not on how many images you brought back. Work out the effort for your next spotting day:

Folder structure and naming: the underrated half

The editing is done after one evening; the filing accompanies you for years. A good structure answers three questions without any searching: when was this, where was this, what’s on it.

Level	Recommendation	Why it works
Top-level folder	Year	stays manageable no matter how many trips get added
Second-level folder	Date and location, e.g. 2026-08-19_Hagen	sorts chronologically and is instantly readable
File name	Date, running number, optionally vehicle	stays unique even after copying into other folders
Keywords	Class, operator, line, station	makes images findable without needing to know the folder
Rating	Stars or colour label	separates culling status from editing status
Exports	Separate subfolder per destination	stops derivatives getting mixed up with originals

Two details make the difference for everyday use. First: dates should be written in year-month-day order, because that’s the only way alphabetical sorting matches chronological sorting. Second: the location in the folder name should be the station or the line, not the trip – you’ll find “Hagen Hbf” again in five years’ time; you won’t find “Summer trip 3”.

Keywords are the part most people put off and later regret. A few consistently used terms are enough: class, operator, station, line where relevant. Consistency matters more than completeness – write “Class 146” once and “BR146” another time, and you’ve created two categories for the same thing.

This system pays off especially once you start uploading images to portals later – they require vehicle and location details, as the practical guide to bahnbilder.de with workflow and legal questions shows. Clean tagging is also worth its weight in gold for research purposes, for example alongside image databases as described in the guide to Drehscheibe Online Railview.

Are you set up for a proper workflow?

What to watch out for

• Back up before sorting. Delete during the first pass and then format the card, and you have no second chance.
• Don’t edit mid-pass. As soon as you start tweaking an image while culling, the funnel collapses.
• Don’t apply base development blindly. Light and dark liveries need different values, even under the same light.
• Keep fine work bounded. If an image still isn’t working after five minutes, it usually never will. The slider order for that is in the article on the five sliders for editing train photos.
• Treat night shots separately. They need an extra step, described in the guide to removing noise in night shots.
• Don’t improvise the export. Colour space, size and output sharpening belong in a preset, as described in the guide to optimising images for upload.

Conclusion

A RAW workflow isn’t a sign of perfectionism – it’s what lets you actually stay on top of things at all. Stick to the funnel – back everything up, cull hard, develop in batches, only touch the best images individually – and even after a long trip you’ll have a result in one to two hours. More importantly, you’ll find your images again years later, because structure and metadata run alongside from the start. For where those shots get taken in the first place, see the overview of locations in railway photography; for what matters in the image itself, see the guide to trainspotter photos and railway photography.