Blog » Iterative Digital Photo Workflow

Digital workflow can be daunting for photographers. I know it took a long time for me to get into a routine that allows me to sort and process the thousands of pictures I sometimes wind up with after a trip or event. I recently shot on the photography crew for the Winnipeg Folk Festival. After the five day event, I had over 4000 images to deal with. This gave me a pressing opportunity to reflect on and refine my workflow.


This step might not be possible if you are facing the pressure of a tight deadline, but it’s a great idea to leave the photos alone for a couple of weeks. Darwin Wigget has a great piece on this, and I completely agree with his reasoning. I find that there is a lot of value in assessing and working with photos after the emotional intensity of the moment has passed. Ultimately what makes a good image is its aesthetics, its internal structure, and one simple way to make that more apparent is to let the passage of time attenuate the emotional memories of the event, the place, or the people captured in the photos.

Quick Sort

My first sweep through a shoot involves has two primary goals: to delete obviously bad images and to flag obviously (or possibly) good images. Avoid doing too much touching up or cropping at this point: when you have many hundreds or thousands of files to process, your first goal should be to sort out just the most promising images to spend more quality time on. If an image is likely a keeper, or shows some kind of promise that I want to explore in editing, I flag it as “maybe_good” for consideration in the second sweep. If it’s clearly excellent, I flag it as “good”. I also delete images that are obviously bad: out of focus, unflattering, blurred, and do on. I deal with borderline-bad images that I’m reluctant to delete for some reason (maybe they’re out of focus, but in a kind of cool way) by flagging them “maybe_bad” and move on. The goal in this first sorting is to quickly winnow down the number if images to spend more time and effort on. I try to balance a critical eye with a sense of possibility: any images that show promise, but might need exposure or colour tweaks or significant cropping, get flagged as “maybe”s to be brought into the second round. It’s worth including marginal images in your maybe_good queue, if only because you will next be comparing them against genuinely good images, which makes it easier to make up your mind. This first sweep can cut down the number of images by eighty or ninety percent, which is your goal. This is especially the case of candid event photography, where there are so many ways for photos to go wrong. For the Folk Fest shoot, my first sweep through my 4000 images left me with about 500 to pay more attention to.

Assess the Maybes

The second sweep through the much-reduced collection of images is intended to make decisions about what images are genuinely good or that will become good with post-processing. I usually classify images as “maybe_good” if they could be improved cropping or if I’m not sure about their exposure, colour, or sharpness. During this phase, I do just enough tweaking to confirm or deny my initial impressions: I might play with the cropping tool; adjust the tone curve or apply automatic exposure correction; adjust white balance or colour balance; or even all three. Don’t spend too much time on post-processing during this stage. Once again, the goal is gatekeeping, not yet producing a final product. If you can confirm that the image shows real promise, save your changes in a new file and move along. If relatively quick experiments don’t yield an image you are much happier with, untag it and move along. Once this phase is done, you should be left with a collection of images you are confident require either little or no further processing (tagged “good”) or that will become good with a bit of work (”maybe_good”).

Improve the Maybes

I continue to focus on the maybe_good photos at this point. Each one is edited until it is the image I want it to be. My goal at this point is overall composition, exposure, and colour. If I’m unable to salvage or sufficiently improve an image, it gets tossed back and I move along. If the image works out, it gets promoted from maybe_good to good. This is also where I force myself to choose between similarly good and similarly framed images, usually using a digital light table to compare them side by side. Once you have finished editing the maybe_good images, you should have a slightly smaller queue of images that you are genuinely happy with.

Quality control

This final stage involves making sure each remaining image file is as good as it can be. This is when I finally deal with any polishing, in terms of noise reduction, sharpening, removing sensor spots, and so on. This final pass should be a matter of refinement, and how long it takes will depend on your own sense of perfectionism and the intended format of the final product. Images destined for the web can be dusted and cleaned pretty casually; a 20”x30” gallery print needs painstaking attention.

End result

This iterative process should maximize the number of useful images you end up with and minimize the amount of effort it takes to get them whipped into shape. I find that the “tag and move along” approach cuts through a lot of the indecision that is usually part and parcel of assessing my own work, and that working my way through smaller and smaller sets of better and better images is the best way to draw distinctions between better and worse images, and to refine your sense of what a successful image means for you.