Phase 3: Image File Conversion and Optimization – The Digital Photography Workflow Handbook

Phase 3: Image File Conversion and Optimization

The largest part of the book is dedicated to this phase. In recent years, RAW converters have become increasingly capable of performing the image optimization processes that used to be the preserve of Photoshop. Most of today’s RAW editors (such as Nikon Capture NX and LightZone) as well as the newer all-in-one workflow packages (such as Apple Aperture, Adobe Lightroom, Adobe Camera Raw, Bibble 5, and Capture One) are capable of processing not only RAW images, but also JPEG, TIFF, and even Photoshop PSD files.

Figure 2-8. The steps involved in converting and optimizing RAW image data

Most RAW editors use non-destructive processing techniques, which means that only the preview image and not the image data itself is altered during processing. The processing steps are recorded in a separate file or in a separate segment of the image data. The changes are only resampled into the image data when you convert the image to another format or when you export it to another application.

Note

The basic principles of image optimization are: “correct major errors first and minor errors later”, and “global corrections first, local fine-tuning later.”

Non-destructive processing saves disk space (compared with complex Photoshop image layers) and helps to avoid the exponential rounding errors that are produced whenever an image is resampled to include a new correction. It also allows you to correct your image multiple times and at a later date without any loss of image quality, and makes it possible to apply exactly the changes you have made to one image to other images too. This makes it much quicker and easier to correct multiple images from a single shoot, and you can still fine-tune each individual image later.

Some RAW editors also allow you to print directly from the monitor image without having to convert your data to a more printer-friendly format first. Here, the RAW editor converts the image data on the fly and sends it directly to the printer.

Which Steps are Better Performed with a RAW Editor, and Which Using Photoshop?

Your choice of processing tools will depend largely on the type of image data you are processing.

Processing RAW Image Data

If you are processing RAW images, you can make the following types of corrections more effectively using a RAW editor:

  1. White balance

  2. Exposure optimization

  3. White point and black point settings

  4. Simple dust and dirt removal

  5. Correcting chromatic aberration

  6. Correcting vignette effects

  7. Simple global color corrections

  8. Alignment, leveling, and cropping

  9. Some compensatory sharpening

Note

Some RAW editors (DxO and DPP) also include functionality for removing lens distortion.

For all other corrective steps, the choice of whether to use Photoshop or a RAW editor is up to you. Both have advantages and disadvantages which we will address specifically later.

A slightly different approach to steps 2, 3, 4, 7, and 9 above is necessary if you plan to process your image later using Photoshop. For instance, you should leave some “breathing room” at the top and bottom ends of the luminance histogram when setting your black and white points, and global color and sharpness corrections can sometimes be better left to Photoshop. It is, however, sometimes necessary to sharpen an image a little in order to be able to judge its quality at all.

Processing JPEG Image Data

If a JPEG image requires major corrections, we always convert it to TIFF format first. This helps to avoid the loss of image quality that is caused by the quantization that is part of the image saving process.[38] The resulting TIFF file is then our new master file that we use as the basis for subsequent corrections. We save the original JPEG files and don’t touch them any more. We compress our TIFF files using LZW or ZIP because both algorithms produce smaller, lossless compressed files and are supported by all Adobe applications. This means you can embed 8-bit, 16-bit, and 32-bit color profiles, you can access and process layers and channels, and you add EXIF, IPTC, and XMP metadata to your files. LZW compression produces slightly larger files than ZIP, but it is compatible with more third-party applications.

If we are working on TIFF data (from a scan, for example), we compress the file using ZIP or LZW and embed the appropriate color profile if it has not already been assigned.

We then apply all necessary correction steps for RAW files as described above and in Chapter 4, using either your RAW editor or Photoshop.

Fine-Tuning an Image Using Photoshop

Here, we perform the image optimization steps that are either not possible using our RAW editor, or that are simpler to perform (or produce better results) using a conventional image processor or specialized plug-ins. Appropriate plug-ins are PTLens [92] for correcting lens distortion, Viveza for performing non-destructive color corrections, and a range of Uwe Steinmueller’s scripts for sharpening (EasyS Plus), for improving local contrast (DOP_EasyD_Plus_DetailResolver), or for reducing the effects of burned-out highlights (DOP_HighlightResolver).

Note

Viveza is described in Selective Adjustment Using U Point Control Points and most of the other programs and plug-ins mentioned here in Chapter 8 and Chapter 12. Multishot techniques and software are described in Chapter 9.

We also use some standalone applications for specific tasks. We use Photomatix Pro [77], FDRTools [72] or HDR PhotoStudio [105] for producing and tone mapping HDRI images, PhotoAcute [81] or Helicon Focus [76] for focus stacking (extended depth of field), and Autopano Pro [57] for stitching panoramas. Photoshop CS4 and CS5 include similar functionality in all three areas, but the programs mentioned all produce better results and are easier to use.

We nevertheless recommend that you start out using just Photoshop and decide later whether you want to spend the time and money to acquire additional functionality.

Where the Philosophy Starts ...

If you shoot RAW images you will be faced with the philosphical question of whether you want to process your image as non-destructively as possible using a RAW editor, or whether you simply prefer to use Photoshop or some other image processor. This question will be answered to a certain degree by the functionality and the quality of the results your RAW editor offers. There are virtually no RAW editors currently available that can correct lens distortion or perspective errors (here, DxO and Silkypix are the exceptions that confirm the rule). Noise reduction functionality is also not up to scratch in many RAW editors.

But the software manufacturers are not playing wait-and-see, and every major release sees great improvements in functionality. Lightroom and ACR as well as Capture NX, LightZone and Apple Aperture, all include selective correction for RAW images. Bibble has included the Noise Ninja noise reduction plug-in and profile-based lens distortion correction since version 4.x. DxO has also been refining its lens distortion functionality for quite a while now.

We have already mentioned the advantages of working with RAW editors (non-destructive processing and small image files), and non-modal functionality is also a great aid to effective image processing. The major disadvantage of RAW editing is the sheer amount of computing power you need, which can lead to over-correction due to the long delays between preview refresh cycles. Always equip your computer with as much memory as possible.

If you use an all-in-one editor that has good Web presentation, slide-show, and print functionality, it can be a good idea to follow these three steps:

  1. Use your editor to perform basic optimization

  2. Convert your image to TIFF and hand it over to Photoshop (or another image processor) for fine-tuning

  3. Return your processed image in TIFF format to your all-in-one program in order to manage your images and prepare them for output “under one roof”

This can be a useful strategy for TIFF and JPEG images too, as increasing numbers of RAW editors (e.g., ACR, Capture One, Capture NX, and Bibble) are capable of processing other image formats. If we are processing TIFFs or JPEGs downloaded directly from the camera, we treat them like RAW originals and only work on copies. This way, we always have access to the original image data in the form it was captured by the camera.

Whether you save your originals and copies to the same folder or to separate locations is up to you – there are arguments for and against both approaches which also affect the overall workflow. We will address this subject in detail in Where Are My Image Files Stored?.



[38] Quantization involves setting neighboring pixels to the same or similar color values in order to save disk space when compressing image files. The process unfortunately causes image quality loss and the formation of color blocks known as “JPEG artifacts”.