Everyone who takes photographs has a workflow, whether they realise it or not. A workflow is simply the series of steps that take your images from the camera to their final destination, which might be a computer, a print, or the internet. For some, the workflow involves no more than pointing and shooting on fully automatic settings and then showing their friends on the camera’s LCD display. For professionals and amateurs who shoot a lot of images, a more detailed and structured approach is needed to catalogue and process their shots efficiently. The age of digital photography has made it easier than ever to take huge numbers of photos, and without a good workflow it’s easy to waste a lot of time processing images that really aren’t worthy of attention, or to end up with a such a confused mess of images on your computer that you can’t find the particular one you’re looking for.
So here’s a rundown of my workflow. It's a work in progress and I’m looking to refine my approach all the time, but overall it works pretty well for me. One area that I know my workflow is lacking at this stage is electronic cataloguing using metadata in a program like Adobe Lightroom. I use a simple manual system that I need to update, so I'd suggest readers look into this for themselves.
I won’t go into detail about the pros and cons of shooting in different file formats – there is a lot of information available on the internet – but I should mention that I shoot in full quality RAW format to capture the most detailed information I can in each image. The files are large but allow a high level of control over factors like colour temperature and exposure during post-capture processing.
People who suggest that post-capture processing is somehow dishonest or artificial compared to an ‘unprocessed’ image straight from the camera should be aware that shots taken as JPGs (the format most commonly used by amateur photographers) have been processed in the camera before they are viewed on the LCD screen. The camera applies default settings for contrast, colour balance, saturation, sharpness and sometimes other factors. The difference between that and a photograph processed by computer software is that the choices about how to process the JPG image are made by technicians in the camera factory, rather than by the photographer. I've never seen a Canon technician lurking around taking notes while I've been out taking photos, so I'd rather shoot in RAW and rely on my own memory of the scene to process the image.
In the field
It’s not unusual after a session of photographing landscapes to end up with several hundred images on your camera, particularly if like me you often bracket your shots to cover the full dynamic range of the scene. My first task after each shoot is to dump the contents of my memory card(s) onto a computer. I take a small netbook with me when I go away on a photo camping trip – it’s compact and light and the battery lasts for days if I’m careful. The downside with a lot of smaller laptops is that they struggle to run photo-editing software like Photoshop, and if they can cope, a session of editing will drain the battery quite quickly. If you don’t have access to power for recharging, it’s wise to limit your computer’s on-time to downloading and backing up images, and leave the editing until you get home.
I set up a folder under My Pictures (Windows) for the whole trip, with a separate sub-folder for each session, named according to the date and location. Images from each session are then dragged and copied from the camera to the relevant folder. As soon as all the images are downloaded to the computer, I back them up on two separate pocket-sized 320Gb external hard drives.
There are alternatives to a taking a laptop in the field, like the photo storage units made by Epson, Digital Foci, Vosonic, Memory Kick and others. These are smaller than a laptop and good for multi-day hikes, but also quite expensive and don’t provide some of the other benefits of a computer like word-processing and internet access.
Once I’ve checked that I have three downloaded copies of all the images, the memory card is cleared ready for the next session.
After a week out shooting, I might come home with well over a thousand image files from ten or more different sessions. Some of these are duplicates, some might have focus or exposure problems, and others will be lacking in regard to composition or content. There’s no point clogging up your computer with files that are no use to you, so a process of culling and cataloguing is needed. I do this work in SilkyPix Pro and Microsoft Windows. There are various software programs specifically designed to manage and catalogue your photographs and do nothing else, or there is the well known Adobe Lightroom that performs this function as well as others.
The Silkypix Pro interface.
Back in the office, the folder containing all of the images from my trip are dragged and copied from one of the external hard drives onto my desktop computer. My main photo directory is divided up by year and month so the already named folders from each photo session of the trip can just be dropped into the relevant month folder.
The next step of browsing through all the images can take some time, but it is one of the most enjoyable parts of the process - it’s usually the first time I get to see my shots on a large screen. I open a folder in SilkyPix and then browse from image to image looking for irretrievable problems with focus, exposure or composition. It’s handy to turn on the highlight/shadow clipping warning while you browse so any exposure problems are obvious. It gets easier to quickly assess an image with practice, and those that I don’t want to keep are tagged with a delete mark – this is a one key operation and is quicker than individually deleting images as I go. I usually make two visual passes through each folder, the second one concentrating more on picking out the best option among duplicated shots, and getting rid of bracketed shots that won’t add any useful exposure details to my final image. I also individually colour tag images that show the most promise for processing so they can be selected as a group later. After dumping all the files tagged with a delete mark, I select those remaining and rename them with a location_date format – most good browsers will let you do this on a batch of files rather than having to select and rename them individually.
Now that we’re left with only the best shots, it’s a matter of methodically working through them – if you’ve tagged the best looking images during the culling process you can easily go back to these as your first priority for further processing.
RAW processors have default settings for things like contrast, saturation and sharpening that will automatically be applied to all your files. You can choose to turn these off and start from scratch – I leave them on in SilkyPix as the settings are quite conservative and a good starting point for further processing. You can easily alter them individually as you go if you don’t like the effect they’ve produced. There are a couple of additional settings that I usually apply to the images in a folder as a group – a subtle S-curve to boost colour and contrast, and a standard setting to remove chromatic aberration for the lens I most often use. Both of these will often be adjusted later on individual images.
Looking at the images one at a time, my first tasks are to check for a straight horizon and rotate if necessary, and decide it I want to crop the image. I always try to get these things right in the camera before I press the shutter button, but reviewing an image on the computer can sometimes lead to a change of heart about where the edge of the frame should be, or which part of the horizon or other element in the image should be used as a reference to give the best balance. Sometimes when shooting creekscapes and scenes that don’t include a horizon, what appears straight on location might look tilted on the computer screen. This problem improves with practice and experience, but it’s something I certainly don’t always get right. It’s worth making these compositional adjustments early in your workflow so you know exactly what you’re dealing with, and not letting parts of an image that might later be cropped out affect the way you process the shot.
Any dust on your camera’s sensor will show up as blurry little blobs in featureless light parts of the image, especially the sky. I use the SilkyPix spot removal tool to get rid of these, but it can be equally well done later in Photoshop or other software.
The next step is to set the white balance. I leave my camera on auto white balance which usually gives a reasonable rendition of the scene, but the camera can get confused when a scene includes areas of deep shadow and bright light. The ability to make significant corrections to white balance is one of the big advantages of shooting RAW files – the scope for change is much less with a JPG file. RAW converters have a white balance eyedropper tool so you can click on a part of the image that should be neutral in tone - a grey sky or dead branch often work well – to set the colour temperature and remove colour casts. It might take several clicks with the eyedropper to find the white balance that looks right to you, and further manual adjustments can be made with the colour temperature sliders. The key point here is to find the setting that looks right to you, or gives the effect that you are after – there is no one correct answer to where the white balance should be set for any given image. You were the one who took the photograph, so you will have the best idea of what the scene looked and felt like at the time and therefore how you want the final image to look and feel.
Another advantage of RAW files over JPGS is the ability to alter their exposure setting. Parts of an image that might look completely black or white and lacking in any detail will often be found to contain plenty of detail as you move the exposure slider up and down in the RAW converter. Unfortunately digital cameras cannot record nearly as wide a range of brightness levels as our own eyes can perceive – for example, standing in the deep shade of a rainforest, we will clearly see twigs and leaves beneath a fallen log, and with a quick glance upwards, will also see a deep blue sky with white puffy clouds through the canopy. A camera sensor, as clever a piece of machinery as it is, will likely record a deep black shadow under the log and a burnt out or overexposed sky. This is where high dynamic range (HDR) techniques come in to try and make up for the shortcomings of our cameras. Specialised HDR software like Photomatix is used by a lot of people and produces some excellent results – others like myself use simple manual techniques in Photoshop to blend together different exposures of the same scene to try and reproduce the full brightness range of the scene as I saw it. (See my article on Exposure Blending)
But I’ve wandered off the track here – my point is that after setting the white balance, I set the exposure level of the image using the exposure slider. If I intend to blend two images together to increase the dynamic range of the final image, I will expose one image to make sure that there is detail in shadows, and the other to show detail in the brightest part of the shot.
Because I shoot at low ISO settings as much as possible, I don't often have to contend with noisy images. The exception is in long exposures where noise will build up and become particularly noticeable if I've increased exposure of the RAW file to open up shadow areas. In those cases, I apply some noise reduction to the RAW file.
That is most of the work done. I rarely change the default saturation, contrast or sharpening levels set by SilkyPix, but will sometimes add some midtone brightness to lift the image a little. By following the process described above, my RAW files are pretty close to how I want the final image to look in terms of overall composition, colour, saturation and contrast. Once I’m happy with the RAW file settings, the image is saved as an uncompressed 16-bit TIFF ready to be worked on in Photoshop.
There are three main tasks I work on in Photoshop: blending different exposures of the same image using layers and masks to give a composite image that covers the dynamic range of the scene, making subtle local and/or global adjustments to colour and contrast using layers and masks, and resizing and sharpening for final output. My method for blending exposures is described in another article.
Most digital images will benefit from adjustments using levels and curves. Once I have an image looking right in terms of exposure with detail in the shadows and the highlights, my next step is to add a levels adjustment layer. The levels histogram will show the spread of brightness values in the image, which should ideally extend pretty much from one end of the horizontal axis to the other (0-255 on the x-axis scale) for most images without being too bunched up at either end. If the histogram curve (the black area) falls short of either end of the horizontal axis, grab the little triangular pointers in turn (black on the left, white on the right) and move them inwards until they just touch the curve. If you slide the black pointer too far to the right, darker parts of your image will turn black – if you slide the white pointer too far to the left, light areas will lose detail and turn bright white. It is sometimes okay to slide the pointers in so they just clip into the curve, but be aware that this will mean small parts of your image will be pure black or white. To see which areas are being affected like this, you can hold down the ALT key on your keyboard while you move the pointers – if the effect is not what you want, move the sliders back out to the edge of the curve.
Often after you’ve applied a global levels layer, there will still be parts of the image that could benefit from a little more contrast, lightening or darkening. One way to achieve this is to add a second levels layer and move the pointers in until the target area looks right, then grab the brush tool and paint black into the layer mask over the areas where you don’t want the second adjustment to apply. That way the levels adjustment will only show on your target area where you haven’t painted black.
I’ll often add several levels or curves adjustment layers to target specific parts of the image. You can achieve similar results using either levels or curves, but a curves layer offers a lot more subtle control than a levels layer over the shadows, highlights and all areas in between. I often add an auto curves layer next to see what effect it has – sometimes good, sometimes not so good. If I like the effect but it is too strong, then I’ll keep it but reduce its opacity until the effect looks right.
Apart from painting into the layer mask as described above, there are several ways to selectively adjust the brightness, hue and saturation of specific colours in your image. Both the selective colour and hue saturation adjustment layers let you target and adjust specific colours with minimal effect on the rest of the image. I use these very sparingly, but find them particularly useful in rainforest scenes where lush green foliage can sometimes come out looking a little dull or yellowish, and in blue skies that look washed out. The key to all of these adjustments is to tread lightly – too much adjustment will result in a very unnatural looking final result.
For some images, I will finish by adding some subtle dodging and burning or a vignette to focus attention on the subject. One good way of creating a vignette is to add a curves adjustment layer, grab the diagonal line somewhere around the middle and drag it downwards to darken the image, and then paint black into the middle of the layer mask so the darkening only shows through around the edge of the frame. You can do the same sort of thing to create dodging and burning effects – create a new curves layer for each, darken one and brighten the other, and then selectively paint into the masks as required. In this case, it works best to fill the masks with black, and then selectively paint in white to reveal the areas you want to lighten or darken.
There’s no end of other clever things you can do in Photoshop but in most cases that’s as far as I go at this stage. I save the image as a Photoshop file (.psd) so the layers are intact and I can go back and make adjustments at any time. That becomes my master file and is backed up on two external hard drives. Then if I want to get a file ready for printing or display on the internet, I’ll reopen the Photoshop file, flatten the layers, resize to the dimensions and resolution required, and add an appropriate amount of sharpening if necessary.
That probably sounds like a heck of a lot of work to someone who hasn’t dabbled in digital processing before, but I see it as simply the modern version of film processing. A lot of the best film photographers were and still are clever darkroom technicians and would spend hours preparing their images for display. The alternative was to rely on a lab to develop their negatives, slides and prints, which could also give good results, but put the final outcome in someone else’s hands – a little like shooting JPGs on auto settings on a digital camera.
If you want to keep full control over how your photographs come out, then it’s worth putting some thought into your workflow. My approach is only one of many possibilities - you’ll develop your own flow and shortcuts the more time you put in.