Image Editing Overview
One of the characteristics of modern photography is that image editing has become a central part of the process. There are two basic ways that images can be adjusted. Pixel editing works at the pixel level and requires altering the original image. Parametric Image Editing works by saving instruction sets that change the appearance of images without actually changing the original image data. The differences between these two methods have a lot of implications for how you approach image editing.
Basic image editing best practice
Batch image editing
Optimized image editing
The master file concept
Image editing types
Parametric Image Editing (PIE)
Parametric Image Editing for rendered files
High Dynamic Range (HDR) imaging
Just as we recommend that you use raw file capture whenever possible, dpBestflow® recommends that you use parametric image editing (PIE) tools for as much image editing as possible. For many aspects of image editing, especially the basic ones that adjust white balance, exposure, brightness and tone, PIE tools tend to be easier and more intuitive to use than Photoshop. Raw files, of course, give you the most adjustability in PIEware, but you may find that even rendered files respond well to parametric edits.
Parametric image editing is non-destructive and you can create multiple versions of files just by creating extra XMP instruction sets – which takes up much less space than several versions of rendered files, or even multi-layered rendered files. For these reasons, we always start in PIEware, and move on to pixel editing only to go beyond the current capabilities of parametric image editing applications.
Although batch image editing can be done in pixel editing applications like Photoshop, it is really much more feasible to do batch editing in PIEware. Pixel editing is really designed to be done as a one-at-a-time process. Although image edits can be done via Photoshop actions, the same action must be applied to all the images in a folder, which are "acted on" one-by-one.
PIEware, on the other hand, works by creating tiny instruction sets that are associated with each image file. These instruction sets can be applied either to individual files or to any number of "selected" image files. These instruction sets can either be executed as you work, or saved up and applied as a batch.
The difference between batches in Photoshop via actions and batches in PIEware is that many discrete instruction sets can be applied, making some images darker, others lighter, some more contrasty and others less, and so on, in an almost infinite number of combinations.
The level of perfection with PIEware batch image editing depends entirely on how much time you want to spend on each image. This is where ratings come in – they can guide you to spend more time with highly rated images, and less time on the less successful images.
Optimized image editing is reserved for highly rated images, and especially those selected by you or your client. Batch image editing is often done with the intent of making images good enough for proofing on a web gallery, or perhaps good enough to make batch processed high res delivery files.
Optimized image editing takes image adjustment to a higher level of perfection. Although, increasingly, images can be optimized to near perfection in PIEware, there are still many images that will require the one-at-a-time process of Photoshop. In many cases, this is because special techniques such as high dynamic range (HDR) processing, panoramic stitching, compositing, combining of type or other forms of vector graphics need to be integrated into the image.
Even without these special needs, there are many things that Photoshop can do that PIEware can't do at all, or can't do as well or as quickly.
Optimized image editing results in the creation of what we call a masterfile. A master file is defined as a large version of the image (at least as large as the camera’s native file size) in a wide gamut color space. Unlike batch-processed capture files, a master file is processed with greater attention to detail, tone and color. Even if the project specifies Web use only, creating a master file is a necessary first step.
Currently, we are in an interesting nexus. File optimization can occur in PIEware as well as in raster image editing software, such as Photoshop. This gives the option of having both a traditional masterfile created in Photoshop, and a virtual masterfile created in PIEware.
The master file is a high-value image file, since it is usually highly rated by either the photographer, the client, or both, but also because it usually has a large time investment in either the PIEware, Photoshop, or both.
As a high-value file, be sure that these files are backed-up, as Working files, and eventually as Archived files. The back-up strategy will vary depending on whether the masterfiles are virtual, created and stored in PIEware, or are in a demosaiced standard format such as TIFF, PSD or, more rarely, one of the JPEG variants such as JPEG 2000 or JPEG XR.
Much of the information on this website is directly related to the class of image editor that is being used. The workflow, file handling, backup and data validation for images adjusted with a pixel editor such as Photoshop is very different from the workflow for images adjusted in a PIEware program like Lightroom or Capture One Pro.
At its core, Photoshop is a pixel-pushing, one-at-a-time image editor, which is quite different from a parametric image editing tool which can work on multiple files just as easily as it can on one file.
As we have discussed above, PIEware is much more tailored for batch processing than Photoshop, and while PIEware can now do amazing things in a one-file-at-a-time approach, Photoshop still reigns for most masterfile work.
In the most basic kind of image editing, you must actually alter the image pixels to make a change. So, to make a JPEG more blue, you increase the amount of blue in the pixels, compared to those of red and green. When you save the file, you've made a "destructive" change to the image, since you've replaced the original color information with new color information. We call this destructive because the original image no longer exists (unless you made a copy), even though you might have improved the image.
Over the years, Photoshop has changed, and it's possible to use layers to work on images non-destructively. The original image can be saved as a layer, for instance, and all the "pixel pushing" can be done to duplicates of this layer. Fundamentally, however, Photoshop must alter the file at some level in order to adjust the image.
Moreover, Photoshop is fundamentally a one-at-a-time program. While you can run batch processes on a folder of images, Photoshop must work on them sequentially. Open, change, close. Open, change, close.
A parametric image editor, by contrast, generally cannot change the original pixels. It makes changes by creating instructions or parameters for interpreting the file. If you want to make a picture more blue, you create an instruction that tells the software to make the pixel more blue. The original pixels – the Source Image – never get changed, only reinterpreted. Parametric image editors are fully non-destructive.
Because parametric image editing software (PIEware) works by means of text-based instructions, it's very well suited to many-at-a-time workflow. You can take the instructions that are made for one image and apply them to any number of additional images. Of course this makes it ideal for digital photography because there is often a need to adjust multiple images in the same way.
While parametric image editing has been around for several decades, and has been the structure of choice for video and graphic design software, the explosion of PIEware choices was fueled by raw file photography. Digital camera raw files simply can't be edited with traditional pixel pushing. As digital photography – and raw file capture in particular – has exploded in popularity, there has also been an explosion in the use of PIEware.
Originally, parametric image editing was a work-around for image files that were proprietary, and could not be rewritten safely. As the tools have matured, many people who don't shoot raw have requested the same kind of many-at-a-time, non-destructive image editing that is possible with raw files. Most PIEware will now work with rendered files as well as raw files.
Of course, rendered files don't have the same bit depth and color information that raw files do, so working on JPEG files won't provide the same quality of results. What does remain the same, however, is that the original JPEG image data is untouched and can be used over and over again to generate multiple different variations and versions.
After color and tone, proper sharpening has the greatest impact on perceived image quality. As discussed in the camera section, most DSLR cameras require capture sharpening to overcome the blurring effects of the Bayer Array and the anti-aliasing filter it often requires to prevent moiré. Capture sharpening is easy to apply as a batch process in PIEware. It can also be applied in Photoshop as an individual or batch action using Unsharp Mask, Smart Sharpen filters, or sharpening plug-ins such as PhotoKit Sharpener, Nik Sharpener Pro, Focal Blade or other specialized tools.
A second round of sharpening, which we refer to as "creative" or "process" sharpening is usually done at the pixel editing stage during masterfile creation. A good example is when we might sharpen only some features in an image – such as eyes and hair, leaving skin unsharpened (or even blurred a bit).
A third round of sharpening, which we refer to as output sharpening, is required by all image files regardless of reproduction method. This targeted output sharpening occurs when the image is at the final size for the specific display or printing device and substrate. Sometimes the only piece of missing information preventing delivery of a completely finished image file is the exact reproduction size.
There are two fundamental ways to sharpen digital images: sharpening the pixels that make up an image, and/or edge sharpening, which is sharpening the edges of shapes within an image.
While sharpening pixels works well in most cases, and especially for capture and creative sharpening, when it comes to sharpening for output, we prefer edge sharpening. Edge sharpening avoids emphasizing noise or other artifacts. In addition, edge sharpening holds up better if an image is resized slightly — which is often an unavoidable occurrence when image files are put into page layouts.
Image compositing still requires the pixel editing tools available in Photoshop and other raster image editors. This kind of work can take days and can use many images and layers. There are entire books on the subject, notably Photoshop® Masking & Compositing by Katrin Eismann (New Riders Press, 2008). Sometimes these techniques are referred to as "photoimaging" instead of photography.
High Dynamic Range imaging can far surpass the dynamic range restrictions of traditional chemical and digital photography, and can be used to recreate a scene more accurately in terms of how human vision works. Specialized software (referred to as tone mapping operators or TMOs) is used to compress the greater tonal range of HDR images into a tonal range that can be displayed on conventional monitors or printed.
Shooting HDR scenes with conventional digital cameras generally requires photographing a sequence of exposure-bracketed Low Dynamic Range (LDR) captures, using software to merge them into an HDR file format, and then tone mapping the files back into an LDR format that can be displayed and printed. Raw files and raw file processors can also be used to extend the dynamic range of images, although not as dramatically as blending separate exposures. Creating HDR images from a single raw file processed multiple times and then blended does allow the use of this technique on moving subjects, however.
Stitching multiple images together is an increasingly popular image editing technique. It can be used to create panoramas, virtual reality (or virtual tours), and super high resolution images of mosaiced imagery. Automated software has been developed that maps out the connection points and blending of pixels required when two or more images are stitched together.