Susan D. Peck
ABSTRACT
There is a growing demand for GIS map products to be mass-produced, either for distribution and sale to the public, or in books or brochures that have accompanying text and information of which the GIS map is an integral part. When it is not cost-efficient or practical to make multiple copies of a digital product on a plotting device or color output device such as a color laser printer, it is then necessary to combine the technology of digital mapping and offset printing.
There are a variety of means to this end, depending on the user's software/hardware capabilities, and also on the capabilities of the printing vendors in one's area to accept and utilize electronic data, or to use mechanical separations to accomplish the finished, four color product.
The Alaska Department of Natural Resources, Division of Land, is the state agency charged with meeting the requirements of Alaska Statute 38.04.065 which requires the preparation of land use plans that provide for the use and management of state owned land. Along with books and maps for inter-agency and intra-agency uses, we publish easement atlases. The purpose of an easement atlas is to graphically depict legal access to public lands within a geographic area. The data base we develop for the atlases is very large, and contains, for departmental use, more information than we print for the public's consumption. The maps we produce for the public, however, are from these more complex data sets.
In the process of publishing first the Kenai Easement Atlas, and now the Prince William Sound Easement Atlas, we have employed a variety of trial and error methods including using ArcInfo's separator function, producing Postscript files, creating electronic files, making mechanical color separations on a Versactec electrostatic plotter, and on a high resolution laser printer. The purpose of this paper is to share these experiences, with the emphasis on which methods worked, including sharing the two finished printed products as well as other offset printed products which were produced using ArcInfo maps and GIS techniques.
In the process of publishing easement atlases for public use we have employed varieties of prepress methodologies to produce full color reproductions of the cartographic output from our ArcInfo data sets. The data sets were created in Arc 6.1.1, and we continued using this revision of ArcInfo until the completion of the project rather than converting to Rev. 7. This was to avoid plotting problems in Rev. 7 which have been addressed in subsequent releases of the product. These data include land status, hydrography, sections, range and townships, rights-of-way (i.e. trails, roads, sites), land use sites such as camp sites, anchorages, public cabins, etc.
Prince William Sound is a remote undeveloped area in Southcentral Alaska, spanning the area east of the Kenai Peninsula and west of Yakutat, containing many islands. The federal government (National Forest) and various native corporations own nearly all the uplands. Atlases illustrate public easements across native lands to publicly owned land. We will be printing between 700 and 1000 copies of the Prince William Sound Public Access Atlas. The need to publish this quantity of books has led us to explore the methods available to us to easily and economically produce high quality, offset printed copies of our digital data.
Historically, atlases were produced using hand drawn maps (ink on mylar layers), that portrayed land ownership and easements only. These were photographically composited as part of the printing process with USGS quadrangle maps, for geographic reference. With the advent of ArcInfo, we have been able to produce atlases, land use plans and management plans solely in a digital format. The Kenai Easement Atlas, our first digitally produced atlas, was published in January 1994. The following steps were used to produce four color process separations of each map:
These separations were then given to the commercial printer. Since the output device used for the separations was a Versatec Electrostatic Plotter, with 400 dots per inch (dpi) resolution, the image was coarse, with flaws in the screens (the black and white separates) that were inherent to the output device. The text in the legend was especially problematic and difficult to read so we opted to design one legend in CorelDraw, which was spliced photographically onto each individual map during the offset printing process. The result (example provided for viewing at conference) was satisfactory for a first attempt at using digitally produced mechanicals, and the best effort we could make at the time, with the commercial printing vendor we were using. This particular commercial printer did not have the capacity to accept digital files.
When faced with the task of publishing the Prince William Sound Public Access Atlas maps entirely through a digital process, I tried a variety of methods with mixed results. The process included:
I had limited success with CorelDRAW, and no success with QuarkXPress. CorelDRAW accepted the .EPS files, but lost part of the information. Adobe Illustrator would only accept files which had been created using DISPLAY 1040 3, the Adobe Illustrator output option in ARCPLOT. It would draw the file, but there were problems with the mapextent, i.e. annotation from outside both the mapextent and the original maplimits set in ArcInfo was shown. The drawing and processing time in Adobe Illustrator was extremely slow, even on a 486 PC. I attempted to print separations from that .AI file, both from within Illustrator and using Adobe Separator, but neither worked. This may have been due to the lack of a specific Postscript Printer Descriptor (PPD) file.
I was able to place the .EPS file (generated in ARCPLOT using DISPLAY 1040 2) into PageMaker (PC version) and print out a black and white composite, but not color separations. The only method which easily and reliably created mechanical separations was bringing the .EPS files up on the 486 PC in Adobe Separator, and printing out separations onto our GCC SelectPress high resolution (1200 dpi) printer. The Adobe Separator program sets the screen angle of each separation, adds registration marks, crop marks, gray scale bar, and color bar. This is similar to ARC's SEPARATOR function, but I found it easier to use. Adobe Separator also has the capability to save each different process color's separation as an individual .EPS file. One drawback with this method is that each separation file is as large as the original (composite) .EPS file, so to transport individual color separation .EPS files (four for each map) to the commercial printer was not practical. (Transporting the individual separation files would requir
After finding a commercial printing vendor that had the capability to accept digital files, we did a test run with the vendor to see which method of .EPS output created the best color end product. I took the vendor the same map in four different forms. I generated the ARCPLOT (DISPLAY 1040 2) .EPS file two ways, one using our department's standard shadeset which is based on a RGB (Red, Green, Blue) color scheme, and the same map generated using a special shadeset I developed using ARCPLOT's SHADEEDIT function, with the shades based on four color process (Cyan, Magenta, Yellow, black) formulas.
The commercial printer (Northern Printing) has a MacIntosh based system which outputs to a Linotype-Hell 630 Imagesetter, with a RIP 60. Their technician attempted to place the digital .EPS files into QuarkXPress, PageMaker, and Adobe Illustrator, each time receiving a postscript error in the Imagesetter when they tried to print directly from the program (Postscript error was: 'dictfull, offending command --def--'). This result is very similar to what I had encountered when attempting to open the file in or print the file from the programs on our 486 PC. The commercial printer did achieve success by opening the .EPS file in Adobe Separator and saving each process color separation as individual postscript files with their printer (output device) information.
Each separation took approximately 7 minutes to write to disk. (7 minutes x 4 colors = approximately one half hour computer time to download each map to the Imagesetter). Each color then took between 3 minutes, 10 seconds and 3 minutes, 30 seconds to image. Each map took approximately 43 minutes to process completely. The Imagesetter outputs the film which is then used to either create the color proof or burn the plates used in the offset printing process. In addition to providing the vendor these two digital files, I provided the same files in the form of mechanical separations I produced using our Adobe Separator program and output on our GCC SelectPress high resolution printer, using the method described previously.
Northern Printing shot the mechanical separations in the traditional manner, using special cameras, producing film negatives which were then used to produce both the color proofs and the printing plates. Northern provided us with a color laminate proof of the four different methods. (Available for viewing at conference). It was a dramatic demonstration and reflects the differences in clarity, color saturation, hue and registration between the different methods chosen to shade and output the maps.
The mechanical separates for the Prince William Sound Public Access Atlas maps created in Adobe Separator resulted in a product significantly better than the ones previously delivered for the Kenai Easement Atlas using the 400 dpi Versatec electrostatic plotter, since the resolution off the GCC SelectPress printer was so much better. But there were still visible color trapping problems. Despite the "perfect" registration provided by Adobe Separator, paper is not a perfect medium... it shrinks and expands in different conditions. Additionally, the process of offset printing from mechanical separations requires each separation to be shot by a camera. Then, each of the four layers are recombined on the press, as the paper is run through four times, once for each of the process colors. While the human error and mechanical error is very minimal with the best of commercial printers, perfect registration of the four process colors is very difficult.
The colors in the proofs of the mechanical separations are very bright and intense, and I would have to tone down the colors, by using different, less intense shades, were this the method of final reproduction used. The map with the shadeset based on CMYK values was somewhat better, as is demonstrated in two contiguous land ownership categories shaded yellow: one bright yellow, one very light yellow. These two categories were not very distinguishable in the map using the RGB shadeset, and only slightly better on the map using the CMYK shadeset, for the mechanical separation process.
The digital file of the map shaded with our standardized shadesets, based on RGB coloring schemes, was superior to either of the maps produced with the mechanical separations. By consensus, the preferred map was produced with the CMYK-based shadeset and output electronically as an .EPS file, and handled by Northern Printing (separated in Adobe Separator - process described previously). Both maps electronically produced had perfect registration, no visible color trapping problems and a very high degree of resolution (2400 dpi). The text (which was all input with a postscript compatible textfont) is extremely crisp and readable on the color laminate proof developed from the digital .EPS file.
I felt I had the most control over color specification by using the CMYK-based shades, and the results of the four method test showed that CYMK colors created the best end product. Next I spent time perfecting the CMYK based shadeset, experimenting with various CMYK formulas and outputting them on the Versatec electrostatic plotter to gauge clarity and intensity. One other technical problem we are faced with is physically transporting the final digital map .EPS files to the commercial printer. Fortunately since our atlas maps are 8.5 inches by 11 inches, most of the files are less than 1.4 MB and fit onto standard diskettes. At the time of this paper publication, we are working with the vendor to facilitate moving our larger files over to them. Most likely we will employ a method of file compression that is compatible to both our PC system and their Macintosh system. The method we ultimately use will be described in my presentation at the conference.
While there are a variety of answers to any production problem, some experimentation is required to discover the most straightforward approach to a solution, tailored to the hardware and software available to the user. In the case of the Prince William Sound Public Access Atlas, the straightest line was for us to produce .EPS files in ARCPLOT using shadesets based on the four color process colors of cyan, magenta, yellow and black, and to deliver them in digital .EPS format to a commercial printer who had the capacity to create color separation ..EPS files and output them directly to their printing device. In the process, for future referrence, we also learned that mechanical separations originating from these same digital ARCPLOT .EPS files, separated with Adobe Separator, and output on our department's high resolution laser printer result in very satisfactory final printed maps, in cases when we are limited to using traditional offset printing methods.
Regretfully, the final publication for this book, which includes ove 82 color maps and accompanying text, will not be published in time to share at this years User Conference, as I had originally hoped when the abstract for this paper was submitted. However, several color laminate proofs of the maps, as well as the aforementioned Kenai Easement Atlas, will be available for viewing, along with documentation on the processes described in this paper.
Acknowledgements: Compiled and written by Susan D. Peck, ADNR Contribution by Jean Tam (ADNR) on methods used to print the Kenai Easement Atlas. Edited by Elaine Thomas, Dennis Daigger, Sharon Rudolph, all of ADNR.