What you don't want to see after ordering a large photo printout are gigantic pixels staring back at you in place of portrait pupils. Printing large posters often requires people to do something they are urged to avoid: enlarge (rather than reduce) a photo, with the expectation of improved (rather than degraded) detail. Blemishes and artifacts are magnified. Obtaining the best reproduction will require some sort of photo retouching - pixel manipulation if you will.
Start with the highest resolution available - likely in excess of 3 megapixels. Professionally shot super high megapixel cameras may not require each of the following steps unless a small area is cropped for enlargement. Keep the image in its default RGB color space until modifications are complete.
Reactive Images strives to offer an Adobe RGB (1998) target in most instances though paper selection affects gamut so other targets may be required. Within Adobe Photoshop it is possible to select View/Proof Setup/Custom and load an ICC Profile (calibrated for the intended paper). To adjust for a specific paper type, perform all adjustments as described below. Then select all layers and right click (or Control-click) to select "Covert to Smart Object." Choose the appropriate ICC Profile and then add adjustment layers to compensate for paper.
Workflow to Improve Quality of Enlargements
Preface - Using Adobe Bridge, some prefer to open RGB JPEG or TIFF images as Camera Raw (Command-R on Mac, Control-R on PC). There are advanced color adjustments available like clarity, vibrance, sharpening, color balance, and noise reduction all at once with the option to open the resulting image as a Smart Object. To those comfortable with these settings, this offers a good method of beginning with a reasonably good color balanced image. At another time, we may visit the Adobe Bridge options in depth. The steps below assume the photo will be enlarged beyond the recommended 120%. Therefore individual pixels are enhanced before and after performing color changes.
Step 1: Crop image - Without getting carried away, isolate the area you plan to use (with a bit extra room) and crop (not scale). This will speed filter operations and minimize amount of computer RAM required. Save this as TIFF or PSD and work on this copy, keeping the original in tact.
Step 2: Reduce noise - Before performing scaling or color manipulations minimize the errant color pixels. Doing so after enlargements prevent some plugins from properly profiling distinct camera patterns. Plugins and third party applications can simultaneously minimize JPEG artifacts.
Step 3: Color balance - Photos may inherit color casts from surroundings such as fluorescent lighting, a brightly colored wall, garment, or green foliage. Balancing colors after reducing noise can prevent errant pixels from being targeted. It may be necessary to mask areas of the image.
Step 4: Retouch blemishes - Spots, scars, and blemishes will only become more apparent after enlargement. Using the Photoshop clone stamp tool, these can be minimized. Selective blurring may be used for significant blemish reduction.
Step 5: Convert to Smart Object - In Photoshop CS3, it is possible to convert image layer(s) to a Smart Object before performing many modifications. (Unavailable filters like Liquify, Extract and perspective transformations are grayed in menus.) It can be cropped or have its color space altered. Smart Objects can be nested to restore or edit key image states. Subsequent actions are recorded separately. By double clicking a Smart Object layer, the image at that state is opened for editing. Close the window, saving changes, to have all operations performed on the Smart Object reapplied. Combining layers or flattening the image cancels links to prior states, effectively disabling Smart Object. A TIFF or PSD file can be saved with Smart Objects in tact for future editing after closing.
Step 6: Enlarge - Rather than double or quadruple an image size in one step, it is better to do so with successive bicubic enlargements of about 110%. In this way adjacent pixels are averaged to add new ones. These intermediate pixels are then averaged again, producing smoother results. How do you know how large to make the photo? If it is placed within a page layout application, the percentage is displayed there. For inkjet posters, you will typically desire 180 to 240 PPI at the final output size. If enlarging up from a smaller image, the higher resolution will not provide more detail so work with 180 PPI. (Option: Nest Smart Object.)
Step 7: Enhance - At the larger size, note potentially problem areas. There may be banding issues or areas that require sharpening, softening, or addition of noise. It may sound like a contradiction but flesh and most organic surfaces are not entirely smooth. Banding within gradations can occur with standard 8-bit images because of the attempt to include an infinite number of color transitions within a possible 16.7 million. Adding a light layer of noise (Artistic Film Grain) or dabbing areas with a custom brush tip can diffuse banding and add more realism. Professional cameras may support significantly larger 16-bit files providing smoother gradations with support for 281.4 trillion colors. If working on a Smart Object, painting out imperfections must be done on a separate layer. The Smart Object layer can be duplicated and rasterized to blend effects.
Step 8: Artistic Filter - Additional enhancement is optionally performed at this time. Either because of degraded detail or a desire to enhance composition, different photos may be merged or Photoshop artistic filters may be added. One favorite for large photos is "Paint Daubs." The one called "Glass" is another that adds artistic properties to a photo. Generally, the primary layer is duplicated before running the filter and adding a slight gaussian blur. Areas may be masked and layer transparency adjusted to satisfaction before merging the layers. However, merging Smart Object layers disables them. (Option: Nest Smart Object.)
Step 9: Sharpen - Sharpening is one of the last steps and it is recommended that a copy of the enhanced but unsharpened file be saved first. Because of the enlargement, edges have become soft. Sharpening exaggerates edge and highlight contrast. If overdone, it ruins a photo for further enlargement and can present an unnatural appearance when printed. Sharpen just enough for the desired output.
Step 10: Convert color space - After all color manipulations are done, a 16-bit image can be converted to 8 bits. Depending on ultimate printing requirements, color space may change from RGB to CMYK. For simplification, RGB may be preferred for large inkjet output. CMYK 4-color is used for digital laser and offset printing. Techniques for proper color space conversions depend upon black component and target colors. After conversion, some colors may require selective enhancement. A copy of the final image can be flattened and saved as TIFF for output.
This may seem like many steps for each image. Keep in mind, these are recommended for the largest of images that have the potential for close viewing. To those without the patience or software, Reactive Imaging offers two services based on these steps.
Bulk Digital Image Enhancement employs steps 1-4 with no masking, saving a set of 50 or more photos as JPEG. Alternatively, Camera RAW adjustments may be applied if deemed more appropriate.
Artistic Transformation includes steps 1-10 on a single photo that is then printed and framed.