The following section contains explanations of some important terms you will encounter in the course of this book: non-destructive processing, non-modal editing, selective correction, all-in-one programs, and image attributes.
This term refers to image editing and processing steps that are not immediately saved to the actual image data, but that are saved separately as a set of parameters that can be applied to the image. The correction concerned will be applied visually to the preview image but is saved separately, either in a database, in a special segment of the image data, or in a separate file. Some programs use a combination of these techniques, such as Adobe Lightroom. All RAW editors are non-destructive, and some image processing software allows you to work non-destructively with JPEG, TIFF, or PSD image files, too.
The major advantage of this concept is that individual image processing steps can be altered or deleted without affecting the quality of the final image. Rounding errors during resampling are also reduced using this type of process. Non-destructive editing methods also produce much smaller image files due to the fact that processing steps are saved as simple protocol lists and not as memory-intensive image layers. Multiple versions of a single image also use much less disk space, as you only have to save the original image once, along with the much smaller processing parameter files. This type of memory management is becoming increasingly important with increasingly powerful image sensors and the larger image files they create.
Individual corrections can also be saved for later application to multiple images, which can speed up batch processing enormously – especially if you use your preset corrections as a basis for further image fine-tuning. Your corrections are only actually embedded in the image data when you export your image or convert it to a different format.
But non-destructive processing does have its limits, especially when an image is subjected to multiple, complex corrections such as adjustments to perspective or lens distortion. Selective editing steps made in specific image areas can only be saved in simple form if the parameter files are to remain small enough to be practical. Non-destructive processing requires a lot of processing power if it is to remain useful. Slower systems tend to redraw preview images slowly, making it all too easy to over-correct images based on “old” preview information.
Ever-increasing computer power combined with increasing availability of cheap memory and better editing algorithms is sure to give us better and better processing functionality in the coming months and years.
The usual method for applying a correction to a digital image is to use either menu items or keystrokes to call up a dialog. The next correction step is then only possible once you explicitly end the last one, usually with a click on an OK button.
Because images are often corrected using multiple processing steps, and because it is often necessary to return to earlier image states during processing, this type of “modal” processing can be complicated and time-consuming.
Non-modal editing does not require you to explicitly close one dialog before starting another. The open dialog is either automatically closed (and the correction applied) or it simply remains open and active during the other process. Most contemporary RAW editors and all-in-one image processing programs work on this principle.
Global corrections (i.e., those which affect the entire image) usually belong to the initial processing steps, while later fine-tuning often requires us to make local or selective corrections to certain tonal values, colors, or forms within an image. These types of corrections are equivalent to the adjustments made using masks, dodging, or burning in a traditional darkroom.
The spatial limitations of such corrections can be selected manually or by using explicit and implicit masks for:
Geometric shapes (with or without transitions)
Specific areas of color, brightness, or saturation
Pixel masks made with brush tools and techniques
Control points (à la Viveza)
These selections can be adjusted later if you are working non-destructively, as described above.
Photoshop still uses its versatile selection tools and layer masks to edit selectively, but creating masks can be a complex (and destructive) process. Photoshop masks are also difficult to alter once they have been created.
Nikon Capture NX  introduced the new U-Point control point selection technique. This process allows very selective editing, but it is difficult to judge accurately – which is why Capture NX still includes traditional masking tools such as brushes and gradients. Nik Software  also offers U-Point technology as a plug-in (called Viveza) for Photoshop and Lightroom.
LightZone  uses geometric techniques to select so-called Regions within an image. These are easy to edit, and the precise image areas they affect are easily identifiable.
In its infancy, the digital photo workflow consisted of separate programs for downloading, browsing, RAW editing, image processing, image management, and output management. The first major evolutionary step saw the integration of download modules and image browsers in RAW editors, but this still left us having to switch from program to program with all the interface and data incompatibilities implied.
Apple Aperture and Adobe Lightroom were the first all-in-one programs that combined all the above-mentioned functionality in a single package. Bibble  has also used an all-in-one concept since version 5. These applications allow the user to switch quickly and smoothly from function to function and phase to phase of the workflow, and also provide a unified user interface. The all-in-one approach also helps to avoid data incompatibility and simplifies the workflow while reducing the software licensing costs involved in purchasing multiple programs. We will look more closely at all-in-one packages in Chapter 6.
The term attributes is probably already familiar to readers with an IT background. Attributes are additional data that describe other data. In the case of digital image data files, attributes are the formal and technical details known as metadata, which we described in detail earlier in Metadata.
In the analog photographic world, recorded image attributes were kept to a minimum, but digital technology makes them easy and often extremely practical to use.