Keepers or crud

Koala, fauna, mammal, marsupial

Processing your photos typically involves lots of steps. You need to decide which images to use, how to develop them (including cropping), and which ones to delete. We each find our own process of managing this, but I am aware that many people struggle with this and can get overwhelmed with the backlog. In this article I’m going to suggest a few steps which might help you develop a routine that works for you.

What are YOU looking at?When you’re generating many thousands of photos a year (sometimes per day!) it’s important to develop some tools and processes to help you avoid being overwhelmed with a huge backlog of images. Even if you’re not being that prolific, it’s almost guaranteed that you’ll have a backlog at some point, and you need to find a way of working through it and catching up. Using keywords, star ratings, labels, collections, and various other metadata in tools such as Lightroom gives us lots of flexibility. But sometimes the choices can leave us wondering where to start.

At a very high level there are two ways of selecting images: picking the “winners”, and picking the “losers”. Of course you also need to deal with all the ones in between those extremes, but narrowing down the field like that can help make things manageable. They’re both useful techniques.


Most of us start off with the technique of picking some images we like. I do it myself soon after a shoot (sometimes even that night while sitting around a campfire on safari). It makes us feel good that we’ve achieved something, and it gives us images we can show to others (or send to a client). Star ratings provide an easy tool to mark some initial favourites. We can scan through the images and when we find an interesting one we can mark it (on a fresh shoot I often use a rating of 2 stars for the initial standouts) and keep moving on. At the end we can use a filter to just show us these “winners”. Of course, along the way if we notice an image that should be deleted we can mark it as a reject (and then review/cleanup those rejects as a final step).

But selecting images this way can be a trap. Once you’ve picked out some winners, it’s easy to keep going back to those and ignoring the rest. If we’ve been in an environment where we’ve shot lots of images there can be a thousand images in the system and even if we find 50 great images, what about the other 950? Maybe they’re all duds, but hopefully there are more useful images in there. Sometimes they’re images which are good but only slightly different to the ones we initially picked, sometimes there are more winners hiding in there: only waiting for the right crop or Develop adjustments to expose them.



Having fallen into the above routine for a while, I had built up a large catalog of images of which I was only using a small proportion: I needed to thin it out. You might find yourself in the same situation.

Depending on the subject it can be hard to mark an image as a reject. If it’s a set of photos of your child or pet, it’s tempting to keep everything. Actually throwing anything away is hard. If you find two very similar pictures but one has a little more blur, it makes that decision easier. And sometimes that’s the start you have to go with. Any way of deciding “out of these two images, this one is worse” is a start. If it means that you’re never going to choose that image for anything, you may as well delete it. It will save space, and mean that you never have to bother with looking at it again.

If you’ve been shooting action (such as some of the work I do photographing birds flying past me while I’m balancing on a small boat floating on the ocean) you may have thousands of photos from a single day. This provides an easier target for an image purge than family photos. Many images will have great focus where for example the bird’s eye is perfectly sharp. But just next to one of those there might be some where the AF system locked onto part of the wing instead and the image isn’t up to scratch. Or there might be a fraction too much motion blur. The next image might be good, the one after that OK, but the one after that great and thus you can go back one and mark the “OK” one as a loser. Even if you only quickly pick out the obviously-flawed ones it can give you more confidence in the quality of the remaining images.

Black-browed Albatross Black-browed Albatross

I’ve developed a process where I can go through hundreds of images very quickly, checking 1:1 and whole-image views and deleting images, usually with an average time per image of 3-4 seconds. This is the average across both the keepers and losers, and includes the review process described below. I’ll describe in a later article some of the hardware and software I’ve put in place to facilitate this.

Sometimes you want to keep an image because although it’s not quite in focus, it’s at least a record of a particular species or event. Of course that does complicate the decision about which images to throw away, so you might need to refine the process further.

Using “soft” deletes

When I decide to delete an image, I don’t immediately delete it: I mark it for later deletion by marking it as a reject. In Lightroom if you’re working in a Smart Collection you can’t delete an image, you can only change the image somehow so it’s excluded by the Smart Collection’s filters (which I usually have set to exclude rejects). In a normal Collection if you delete the image it will disappear from the collection. In both cases the image will still be sitting in your library and on disk.

By first marking it as a reject, you can later select all the rejects in the catalog, review them to convince yourself that yes they should go, check to see they’re not being used in any other collections or Publish Services (maybe it’s one you’re using for a project that illustrates image flaws and you actually want to keep it around), and then delete from the library (and usually from the disk at the same time). This has parallels in the Trash and Recycle Bin functions our operating systems provide when deleting files.

Incidentally, before we had global reject flags in Lightroom we were still able to perform the same function by assigning one of the “labels” (associated with colours) to a to-be-deleted status, and using similar filtering methods. Such methods can also be used when interfacing with other software such as Bridge.

I prefer to clean out the rejects every day or so. If I made the decision to reject an image recently I’ll probably remember it and this can simplify the review process. The longer you leave images as rejects the harder the review process will be, and you probably won’t actually get to the step of freeing up space!

And then the rest of them…

If you’ve picked out some winners, and trashed the losers, what about the images in between? Star ratings are one tool to help with this. Usually these are used as some indicator of quality: the more stars the better the image.

To avoid reviewing the same images over and over each time you sit down to work on your library, filters and Smart Collections can be a help. For example you can use an “Unrated” Smart Collection which shows you any image with a rating of 0. You can use metadata filters within that collection to narrow the list down to particular date ranges or locations (or lots of other choices) if that helps. Determine your rating system so you know how to decide what rating to give, and then you should be able to fly through this Smart Collection assigning initial ratings. As soon as you press a number key from 1 to 5, that rating is applied to the image, it will disappear from the collection, and you’ll be presented with the next. Sometimes an Attribute Filter in the Library is a more-flexible tool, allowing you to change it on the fly and also glance at the rated images.

There are many ways to use the ratings, but a start might be something like this:

1. Probably not good, but I can’t quite decide to throw it out yet.
2. Worth consideration.
3. Definitely worth using.
4. Great images.
5. My best images.

You can then come back and look at the 1-star images to pick out more losers, or the 2-star images to consider upgrading their ratings (or the images with at least 2 stars, etc). You will probably refine your rating system over time, but the above is probably a reasonable starting point.

But however you do this, if you’ve managed to throw out half the images as “losers” before you start then you’ve at least halved the amount of work you need to do. Sometimes the binary decision of whether to trash an image is quicker than picking a rating for it.

Review old images

Your standards will probably change over time. When I was first shooting with a DSLR in 2000 I spent a lot of time at a local park practising on the ducks. I made a lot of mistakes, but at the same time I came away with some images I was proud of, and a lot of experience as to how my camera’s AF and exposure systems behaved. I threw away lots of images which were out of focus, but I had lots of things like the back of a duck’s head which were technically OK but not interesting. I remember I kept a lot of them because they showed things such as feather detail and I thought that maybe I’d use them some day.

Years later when trawling through those folders again I was able to convince myself that I hadn’t used them by now, I’d taken better photos since, and it really was time to purge them.

Sometimes you can’t make good deletion decisions immediately. At one stage of 2014 I had a combination of a faulty lens (in the end it turned out to have at least one mis-aligned element) and a new camera model, and for a while I had lots of images I wasn’t really happy about. But I didn’t delete them straight away: were they my fault or the camera’s fault? Once I got the technical faults fixed (and changed some technique things to do with the new camera) it was much easier to look at those images from a month or so prior and make bold yes/no decisions.

Yellow-nosed albatross

Hopefully this article has given you some ideas to consider in handling your own image libraries. Image backlogs are the bane of many photographers. I use combinations of all of these methods (and others) interchangeably: it’s good to have many ways of attacking a backlog!

I will be expanding on some of these ideas in future articles.

Leave a Reply