The biggest challenge in reproducing computer graphics on output devices in today’s marketplace is dealing with transparency in graphic files. This truly emphasizes the importance of WYSIWYG in proofing for the graphic communications industry. We must first emphasize that page layout graphic software is not developed for producing documents for mechanical reproduction. This software prioritizes the creation of documents for viewing on electronic media; they are created on a computer screen for viewing on a computer screen. We have reviewed some of the issues with rasterizing vector shapes consistently, and reliably representing colour from one device to another. Viewing a graphic with three-dimensional transparent elements is significantly different on an illuminated medium where the light is transmitted compared to an opaque medium where the light is reflected. It is very hard to judge how the transparent effects will translate from one to another. There is room for the same kind of collaborative research in this realm, as there was in developing OpenType font architecture and ICC profiles.

The problems in WYSIWYG production for transparency fall in two categories. The first problem is setting expectations so a designer can make a reasonable prediction of how the document will look when imaged on a given media. The second problem is the sheer proportions of the computational processes we are asking of a RIP. PostScript is a three-dimensional language that allows a creator to stack and prioritize elements on a page. The RIP can literally ‘throw away’ raster data that is knocked out by graphic elements that completely cover the elements behind. If those elements have to show through the foreground elements by 20%, the RIP must hold much more raster data in physical memory addresses. Many times, data is lost if there are not enough addresses available for the computations, and this can change from one processing of the document to the next.

Designers can employ strategies at each level of document creation to manage these problems. The first strategy is to use layers well in document creation. By isolating different effects on separate layers, it becomes easier to isolate and edit the transparent effects when they don’t produce the desired results in the final output. The layers can be included in a PDF file of the document, and this allows the possibility of relatively quick editing in PDF editing software closer to the output stage. This can be a completely different working style for some graphic artists. If we start with the premise that the computer screen representation of the document is NOT good WYSIWYG and will probably need editing, then we can justify working with layers more to isolate effects. We can organize design elements on layers after creation — when we are fine-tuning the effects. Usually, this is a good technique when creating many elements on several page dimensions. Designers can review their documents and decide if there are distinct dimensional levels, as page elements are pushed further into the background to pull other page elements forward. A simple example is a book cover for a retrospective, with pictures from four distinct decades. The photos and type from each decade can be set on distinct layers, and transparent values of 25%, 50%, 75%, and 100% can be set for each layer. The screen will render one version of the document, and the printer will render another. It is easier to fine-tune the four layer levels of transparency than to go back and set new transparency levels for dozens of individual page elements.

Another strategy that must be considered for processing multiple transparent page elements is allowing the page layout software to raster the page elements, so it sends raster data to the RIP. This technique treats the transparent elements, such as a photograph on the page, and allows the creator to choose the resolution of the raster. Care must be taken here to ensure overlapping vector elements will raster at the same resolution in the RIP. Let’s say we have a type block that crosses a photo on the page, but it is transparent to let the photo show through the type. If we rasterize the transparent type at 300 ppi — the resolution of the photo — it will be significantly different from the raster of the vector type at the RIP, which might be 3,000 lspi for some plate-setters. The letter shape will be 10 times thicker over the photo, and that will be VERY noticeable if the type crosses the photo in the middle of the glyph. The solution is to make sure to raster the transparent type at 3,000 ppi to match the plate-setter raster. This makes the PDF file very large because it contains lots of raster data. But this solution is also a disadvantage because it does not allow late-stage editing of the transparent values in the PDF file. The advantage is that the transparency elements will have better WYSIWYG, process more consistently in multiple RIPs, and use less RIP resources in processing.

It is very important to be aware of the transparent elements you are creating in a document. It is not always apparent when using effects, plug-ins, or effects filters available in page layout software. Using a bevel or emboss effect, or a simple drop shadow, makes that page element use transparent routines in the RIP. Programs like Adobe InDesign let designers view all the transparent elements on a page. Designers should examine each one to decide if it should be rasterized before RIP-ing or at the RIP. This is a good point at which to decide if transparent elements can be grouped, or organized, on common layers. It is also a good point to decide how the transparent element contributes to the design, and how critical the level of transparency, or WYSIWG value, is in the overall design. In the retrospective book cover design referred to above, WYSIWYG is very important in communicating the message of the book and getting predictable results.

Transparent elements can be rasterized at the page layout stage, the PDF creation stage, and at the RIP stage for the final output device. Adobe Acrobat also has a tool to view transparent elements in a PDF file. It is important for a designer to compare the transparent elements in the PDF to those in the page layout software. The primary concern is that the elements rasterized in the PDF are no longer editable, so it is critical that the levels are right to create the desired overall effect. It is also important for a preflight operator to view the transparent elements in a PDF file to check what the RIP will have to process and to make sure the computational resources are available. If there are processing errors in the final output, they are most likely to occur in rendering the transparent objects. Viewing the transparent elements on a page in Acrobat should provide a mental checklist for the operator when she or he views the final output.

Communication Is Key

The graphic communications industry still has collaborative work to do to make the processing of transparent elements on a page more predictable and repeatable. It is important for designers to understand the problems they can be creating for a RIP, especially for output on an extremely high-resolution device like a plate-setter for waterless lithography. It is also important for operators who are managing documents with lots of transparency to be aware of the checkpoints in a document, and to know when there is not adequate WYSIWYG for the transparent elements on a page. Good questions for all stakeholders to ask when processing a document that relies on many transparent elements are:

• Where are the transparent elements?
• Did they process correctly?
• Is anything missing in the layers that should show through the transparency?
• Are there transparency values that can be adjusted to optimize the overall effect?

Let’s review the primary tools for reproducing transparent page elements in a document. We can utilize layers in a document for setting common transparency values. We should view all transparent elements in a document before and after creating a PDF file. There are several stages to rasterizing the transparent elements. The earlier we rasterize them, the less editable the document becomes, and the more consistent the final output will be. We are creating a larger file to process when we rasterize transparent elements early. Much less computational resources are required at the RIP, and the more predictable our final output will be. When managing late-stage processing of transparency, we must be aware that what we are viewing on a computer screen is not necessarily a good representation of the final output. Graphic artists at all levels of production must pay attention to the transparent areas of a document to check for accuracy.