=Paper=
{{Paper
|id=Vol-1328/GSR2_Molyneux
|storemode=property
|title=Developing a Conceptual Design Framework for Multi-Format Map Publishing
|pdfUrl=https://ceur-ws.org/Vol-1328/GSR2_Molyneux.pdf
|volume=Vol-1328
|dblpUrl=https://dblp.org/rec/conf/gsr/MolyneuxC12
}}
==Developing a Conceptual Design Framework for Multi-Format Map Publishing==
https://ceur-ws.org/Vol-1328/GSR2_Molyneux.pdf
Developing a conceptual design framework for multi-format map publishing
Craig Molyneux and William Cartwright
School of Mathematical and Geospatial Sciences,
RMIT University, Melbourne, Victoria, Australia
Craig Molyneux s8303468@student.rmit.edu.au
Key words: design: multi-publishing: multi-scale mapping: cartographic design: design operator
Author Biographies
C. Molyneux: Craig is studying his Master of Land Information at RMIT University Australia. His area of research is developing
a multi-publishing system for student atlases. He has recently completed production of the Jacaranda Atlas 7th Edition and World
History Atlas for John Wiley & Sons.
W. Cartwright: William Cartwright is Professor of Cartography and Geographical Visualization in the School of Mathematical
and Geospatial Sciences at RMIT University, Australia. He is Chair of the Joint Board of Geospatial Information Societies and
Immediate Past President of the International Cartographic Association. His major research interest is the application of integrated
media to cartography and the exploration of different metaphorical approaches to the depiction of geographical information.
Introduction
Map design for print does not necessarily translate to the screen. Today, screens vary from the large computer monitor and
wide screen television, to the hand held smart phone. The utility of maps hasn’t changed, merely the method of delivery..
Good cartographic design is critical to the exchange of information between the cartographer and the map user. Traditionally
cartographers have designed maps for particular uses and users, such as topographic maps for bushwalking, road atlases for car
touring, or student atlases for children, to name a few. Each user has their own particular requirements and the particular
cartographic design takes these into account, such as larger text and vibrant basic colours for student atlases; or clean white
backgrounds and clear class distinctions between roads for a road atlas—all decisions taken by a cartographer (or cartographers)
when designing a specific map for a specific user.
As we enter the digital age however, designing maps for web sites or mobile devices has presented the cartographer with
interesting challenges. How best to convey user specific information on a screen that comes in various sizes and resolutions, in
conditions that the cartographer cannot predict (indoor, at a desk, in a shopping mall, or outdoor, in the sun or rain, or in-car)? The
transition from purely printed map products to purely digital is moving apace now, with publishers struggling to keep up with the
pace of change. Large cartographic libraries of work need to be moved across to the digital platform without sacrificing the design
integrity of the initial map design. This is not simply a matter of converting the digital file to a raster format, rather careful
consideration needs to be taken into account of modifying each design feature of the map, such as font sizes, colours, line weights,
symbols sizes, to work on the device they are intended to be displayed upon.
This research will propose a design framework whereby cartographers can modify each component of their print map designs
to successfully‘work’ on the digital device they intend to display it upon, whilst still retaining the integrity of the design—a
‘design once, publish many’ model. The research has reviewed existing research done by experts in their field and applied it to a
‘conceptual’ framework. Whilst we have created a framework that does ‘work’, it is conceptual in the fact that this framework
needs to be built into an existing software application, or created as a web page where design parameters can be input and
applicable outputs are presented to the user to apply to their cartographic design. To date, the framework has been created in the a
spreadsheet application to display its functionality. An example output using this design framework is also provided as a proof of
concept.
Mapping: print, web and mobile
For the purpose of this paper, multi-format map publishing is defined as taking GIS data comprising of multiple themes and
producing maps for print, web and mobile media from this data. The data may be stored in a form of database, where geographic
features are represented as point, line and polygon objects and are stored with a spatial component. Typically spatial databases are
a method of storing data and allowing complex querying to be done to that data, however more recently this spatial data is being
used as the foundation for map production.
For the cartographer, designing for one media is a skill developed over time. Understanding the media involves understanding
the choice of colours, selection of fonts, styling of lines and design of symbols. For instance print mapping takes map artwork and
separates this artwork into four colours, cyan, magenta, yellow and black (CMYK), and recombines these colours on a printing
press to recreate the map artwork. Map features, such as lines, points, areas and text are made up of these four basic colours.
Selecting colours for print requires understanding of the process colour system, what percentages of cyan, magenta, yellow and
black make a particular colour.
Web mapping involves creating digital versions of a map or maps that can be static or dynamic and be displayed on a monitor.
Monitor colours are represented by coloured pixels created using combinations of wavelengths of red, green and blue (RGB) light.
The RGB display of colours involves 256 wavelengths or red, green and blue, or 256 x 256 x 256 = over 16.7 million colours
available on screen. A vast palette for use on a map, however not all RGB colours are the same as CMYK. Converting a design
between the two colour spaces does not necessarily give exactly the same result.
Mobile mapping displays raster or vector maps on hand-held devices, such as smart phones, tablet devices and e-readers. Since
this technology is in its infancy the most popular mapping application has been Google Maps, which has been available for the
major mobile operating system platforms—Apple’s iOS and Google’s own Android. Google Maps uses a Tile Map Service (TMS)
to serve to the user small raster tiles of a map area at various scales. The TMS allows the user to zoom in to see more detail and
pan around a geographic area. For the purposes of this research a TMS is assumed for use on mobile devices. The main difference
between a web map and a mobile map, for the purpose of our definition, is the screen size that the map is delivered on and the
resolution of the screens.
----
There has been a great deal of research into cartographic design over the last 70 years or so, with researchers such as
Robinson,(1952), Petchenik, (1976), Keates, (1993), Bertin, (1983), MacEachren, (1995) and many others contributing
significantly to our understanding of cartographic communication and map design for print. This foundation work has covered
areas such as correct fonts styles to use, font sizes, symbol sizes, legibility, proportional symbols, correct line weights and styles,
visual variables and many other nuances of cartographic design. This foundation work in cartographic cognition has spilt over to
studies in designing maps for children, atlas publishing, symbol design, colour, and map projections to name a few examples.
This print map research has stood the test of time and is a staple of cartographic training that all new cartographers accept as a
point of truth in map making. This paper doesn’t question these long-held beliefs and accepts them as truisms for the profession,
however, later research has shown that they do need to be modified when designing maps for screen display.
In 2008 Jenny, et al, compiled a set of guidelines for designing maps for the internet. At the time, maps on the internet were
becoming mainstream. GoogleMaps had brought web mapping front and centre into the consciousness of the general public and
visualising location, finding and calculating routes and always having a map at your fingertips was now a reality. Jenny et al’s
guidelines took into account the existing technology of the time— “...limited screen resolution and anti-aliasing, minimum
dimensions and distances for map features, the generalisation of information density and geometry, screen typography, colour
rendition…”. In four short years since we have seen the introduction of smart phones (2008) and tablet devices (2010), each with
their own unique issues and design constraints.
This individual research gave us some guidelines about what minimum standards apply to design elements for particular media
technology, albeit technology that is now aging As Meng (2003) stated, “The development of cartographic theories and methods
lags far behind the technical evolutions.” The proposed multi-format publishing model requires a set of design guidelines that can
be applied across a range of outputs (media). Whilst design elements and technologies will vary, the design framework should be
flexible enough to create minimum design specifications for multiple outputs, saving time and money in the process—a ‘design
once, publish many’ model.
Towards a design framework
Publishers may often take the quickest and cheapest option when going down the multi-published route. The cheapest
workflow often involves creating artwork for print products (Figure 1), either a map or book of maps, converting this artwork for
use on a website, often Flash-based, and re-purposing the Flash-based map artwork for use on tablet devices (Figure 2). Whilst
there are cost benefits to this approach, the resulting variety of file formats and poor reproduction quality reflect back to a
compromised production approach, or, at the very least, one that was hastily adapted to meet emerging technological or consumer
demands.
The optimum approach is to carefully assess the final delivery methods required, what technology is best able to deliver a high
quality result, create a design specification that will create a product that looks good in print, on the web and on mobile devices,
regardless of software platform, hardware used and delivery technology utilised. Whilst map designs will be universally different
depending on the cartographer, being able to port a design to a digital platform with confidence is critical in the design process.
The framework proposed in this paper endeavours to take the guesswork out of what may or may not ‘work’ on screen displays
based on existing research done on legibility, as well as present the relationship of map and data extents required for multiple
production platforms.
Figure 1: Jacaranda Atlas 7th edition, print atlas example. (Ramsdale, J. et al., 2010)
Figure 2: myWorldAtlas 1st edition, iPad example. (myWorldAtlas 1st ed., 2012)
Before print map production begins the cartographer will ask questions such as:
• what will the geographical extents of the map be?
• what physical size will the finished map be? and
• what is the most suitable map projection and scale?
In the multi-format publishing environment, the answer to these questions will often be different for each media. A print map
will be limited by the sheet or page size, which is often dictated by the press size used by the printer. A web and mobile map, on
the other hand, are limited by the screen display and resolution of the device being viewed upon, however the geographic extents
can potentially be global. Scale is fixed for a print map, whereas web and mobile can have multiple scales, if the user has the
ability to zoom the map. Map projection in a TMS is defined in the Web Mercator projection, something not often used in print
products. The Web Mercator projection contains high levels of distortion towards the poles, making digital map extents physically
larger than a similar scaled print map using a more appropriate projection.
In multi-format publishing the variables of scale(s), projection, extent and screen resolution, need to be defined early in the
design phase to better quantify project parameters. As will be seen later, these values can also help in the production process.
Design considerations for print maps
For maps to be printed they need to be produced in the CMYK colour space. Each colour on the map is specified as a
percentage of Cyan, Magenta, Yellow and black (K stands for Knockout). In instances of black type placed over colour this needs
to be further specified as an overprint, and consideration needs to be made for colour trapping. All necessary components of
preparing artwork for printing on a multi-colour printing press.
Minimum line weights on a paper map have traditionally been around 0.1 mm or 0.25 pt. In part this is due to what the eye can
perceive, but also due to the photomechanical method of plate making and the tolerances it can work within, and the tolerances of
printing presses in registering such fine lines when they are made up of percentages of cyan, magenta, yellow or black.
There are no hard and fast rules when it comes to symbol size on print maps, and they will vary depending upon the type of
symbol being displayed and its use on the map. Due to the high resolution of the finished artwork—around 300 dots per inch (dpi)
for modern printing presses—symbol shapes can be distinguished down to small sizes. For example a circle can be distinguished
from a square when it is displayed at a size of 0.8 mm, (Jenny, et al 2008).
Fonts come in all shapes and sizes, however when used on a map there are certain cartographic conventions that hold true
based on the research of Robinson (1952) and others that sans serif fonts are easier to read on maps. Font sizes and styles may
vary, however the minimum legible point size on a print map is around 5 to 6 pt, depending on the font used (font heights vary
from font to font).
These minimum standards in print provide a baseline from which the cartographer knows the print map design can be
reproduced successfully and be legible to the user. These minimum standards create the platform for the conceptual design
framework to build upon.
Design considerations for web maps
Preparing a map for screen display on the other hand has a different set of requirements. The colours displayed on screen will
be in the RGB colour space (sometimes expressed in HEX [hexadecimal] notation). The amount of different colours available in
the CMYK colour model is less than the RGB colour model, which limits the choice of colours available when looking at a multi-
published design. Whilst there is no way to precisely convert a CMYK colour to an RGB colour, colour tools from graphics
software applications give a good approximation of what the user will see both on screen and in print. It is notable that some
colours that are viewable in one system cannot be replicated in another and older monitor models cannot display the full spectrum
of 16 million + colours that are available on present-day (24 bit) monitors
Screen sizes and resolutions vary in computer monitors. Jenny, et al’s 2008 research highlighted the variations in an attempt to
determine minimum specifications for design web maps. Results of the monitor research are shown in the Table 1.
Table 1: Size and resolution of commonly used liquid crystal displays (LCD) (Jenny, et al 2008)
Display Size Number of pixels Visible area Pixel size Resolution dpi
17’’ 1280 x 1024 338 x 270 mm 0.264 mm
96
19’’ 1280 x 1024 376 x 301 mm 0.294 mm 86
20’’ 1400 x 1050 408 x 306 mm 0.292 mm
87
20’’ 1600 x 1200 408 x 306 mm 0.255 mm 100
Symbols for web and mobile maps will be represented by pixels, so issues regarding pixelation and anti-aliasing may occur
around the edges of the symbol, which will affect small features. Shapes of symbols need to be quite different from one another to
be instantly distinguishable, yet simple enough to be understood at various sizes. Symbol designs have to be familiar and
meaningful and used in a relevant context, and should be designed at the size that they are being used, (Hofmann 2011).
Brown et al. (2001) suggested a series of conventions to be followed when labelling features on maps for the web, such as
keeping minimum point sizes to 10 point for legibility, increasing inter-letter spacing, using simple sans-serif fonts, avoiding italic
text where possible, especially on non-horizontal text, adding a thin white line (halo) around text to increase legibility and keeping
label density to a minimum to improve readability. Jenny et al. (2008) also concluded that sans serif fonts were optimal for screen
reading, with the use of sans serif fonts only to be used for larger screen type, and specifically using sans serif fonts designed for
screen display.
The work presented here by Jenny and Brown et al. provides a basis for calculating minimum standards acceptable for screen
display of map elements. The scale factors calculated form the ‘design operators’ used in the conceptual design framework.
Design considerations for mobile maps
Mobile devices use the similar screen technology as computers—but with smaller screens. Colours appear in the RGB colour
space, therefore the colour constraints mentioned in the previous section are still relevant to designing in the mobile space. All
major models of tablet devices have 24 bit displays.
Screen sizes and resolutions vary from computer monitors to mobile devices (smart phones and tablets). The latest Apple iPad
and iPhone have screen resolutions of 264 ppi (pixels per inch). To update Jenny, et al’s 2008 research Table 2 shows the latest
screen resolutions of the two market-leading tablet devices.
Table 2: Size and resolution of the market-leading tablet devices (Apple, 2011; Samsung 2012)
Model Display size Number of pixels Visible area Pixel size Resolution ppi
Samsung Galaxy 2
10.1” 1280 x 800 218 x 136 mm 0.17 mm 149
10.1
Samsung Galaxy
7” 1024 x 600 154 x 90 mm 0.15 mm 169
Tab
iPad and iPad 2 9.7” 1024 x 768 197 x 148 mm 0.192 mm 132
new iPad 9.7” 2048 x 1536 197 x 148 mm 0.096 mm 264
As can be seen from Tables 1 and 2, in the space of four years display technology has become vastly improved with the
introduction of more pixels. Images are sharper and the visible appearance of pixels on screen has become diminished.
Jenny, et al (2008) estimated that computer monitors were viewed from a distance of around 60 cm which assumed a minimum
viewing line width of 0.17 mm. Paper maps, on the other hand, had a viewing distance of 30 cm, which had a minimum viewing
line width of 0.09 mm, hence a long held belief that the minimum line width for paper maps should be 0.1 mm.
Due to their hand-held nature, mobile devices (tablets and smart phones) would have a different viewing distance to computer
monitors. Whilst this viewing distance will vary from person to person, the resolution of devices affect the quality of the line
widths displayed. Using Jenny’s calculations, we can see (Table 2) that the new iPad almost exactly replicates the ideal minimum
line width to be displayed on paper-based maps. The earlier iPad models and the Samsung Galaxy tablets have the same screen
resolutions as a computer monitor, but when viewed at a distance of 30 cm the lines become pixelated and not rendered crisply.
In the area of multimedia cartography, scale is one level of abstraction that affects the user’s understanding of the map
displayed, (Cartwright and Peterson, 2006). When working with a static map, scale has a purpose. The representative fraction is
generally rounded to the nearest thousand, ten thousand, or million to allow for easy distance calculation and measurement. In
multimedia cartography the need to measure and calculate has in most instances been removed, as the application can do this for
the user. In a TMS, scales are based around the standard 256 x 256 tile size and Web Mercator projection and as such displaying
static scale bars becomes less necessary. As screen resolution changes, so too does the physical size of the TMS tile. For example,
a 256 x 256 pixel tile on a device with 132 ppi resolution the tile will be 1.93 x 1.93 inches or 49.02 x 49.02 mm. On a device with
264 ppi resolution the tile will be 0.96 x 0.96 inches or 24.63 x 24.63 mm in size—half the size and one quarter of the area as that
displayed on the standard screen. Screen resolution impacts considerably on the way tiles are displayed, compressing the tile size
and the content on the tile changing the scale of the map displayed.
A conceptual design framework
When considering a design framework for multi format publishing we are trying to draw together the minimum standards of
the various media formats to deliver an outcome that allows us to develop one master design that we are confident will work
across the various media—a ‘design once, publish many’ model.
At the beginning of a mapping project the design may be unknown or not specific, as design is a fluid concept that evolves
through an iterative process, refining design components to achieve an overall product that communicates clearly and concisely.
The conceptual design framework proposed here takes the final print map design and modifies this by a series of ‘design
operators’. These ‘design operators’ are applied to the various design elements—colours, lines, symbols—and converting the print
design to one that will work on the web. By incorporating the screen resolution of the mobile device the map will be displayed
upon, units will be calculated to modify the design elements for display on the required device. By using this design framework a
map design will be modified for display on multiple media. The user of the map products should have the same design experience
across all media the map is delivered on.
During the initial design phase care should be taken to include colours that display well across the various media, symbols are
designed so that present well and are easily distinguishable and understood, and choice of fonts include styles that are legible and
easily read across all media. If these basics are adhered to, the resulting design for the various media will appear cohesive and
seamless.
GIS software and illustration software allow the cartographer to create templates or libraries of styles to record individual
design elements. Perhaps the closest to this is the eXtensible Markup Language (XML), which comes in various flavours
depending upon its use, such as Geographic Markup Language (GML), used in the spatial industry.
A derivation of XML is XSL (eXtensible Stylesheet Language) (Berglund, 2006), that, through written expressions, allows us
to specify design components to be used in web design mainly, though it can be used in any software application that can read
xml. The XSL language has been around since 1999, though there are very few examples of this being used widely. XSL appears
to have been overtaken by CSS (Cascading Style Sheets) (Celik, 2008) as the language of choice for web designers, as this is not
as complex and is more intuitive for non-coders. CSS has now been introduced to cartography through the software application
TileMill (2012), which uses a derivation of CSS known as Carto, that enables styles to be created as code for map objects, and
rendering these in a TMS.
Significant work has been done by Roth, et. al (2011) in the development of a design schema, named ScaleMaster, to assist
cartographers when creating seamless multi-scale maps. In the words of Roth, ScaleMaster is “a conceptual schematic for
organizing, maintaining, and sharing the scale-dependent design specifications of a multi-scale mapping project.” Defined in a
series of spreadsheet tables, ScaleMaster uses unique codes to describe actions done to data, such as turning data layers on or off,
filtering data, reducing line weights or font sizes, and many others. In this way clear rules are established and map designs can be
replicated on other projects across a series. Figure 3 displays the unique codes and the actions they describe to be done to data.
Figure 3: An example of Scale Master’s design schema. (Brewer and Hatchett, 2010)
ScaleMaster differs from this proposed conceptual design framework in that our framework considers minimum standards
required for all media and provides conversion factors necessary to achieve optimal results. Our framework could be used in
conjunction with Scale Master to work across multiple media, whilst ScaleMaster deal with design in a TMS.
In the ideal map design world, a stylesheet or template with design elements should be able to produced for use in multiple
software packages used in creating maps for print, web and mobile. A product or schema such as this doesn’t exist, however with
a it is envisaged that having a design template that could be used across multiple packages and includes a series of ‘design
operators’ to translate styles dependent upon the delivery method of the design, would be the ideal tool for the cartographer and
publisher.
How will the conceptual framework operate?
In a ‘design once, publish many’ world we want the design from one media source to be reflected across other media, to
provide a consistent user experience for the user. To achieve this, design elements need to be broken down into their individual
components and turned into data. Individual design elements such as colour, line weight, area patterns and font size are all
elements that can have a value that can be stored, retrieved and manipulated.
The conceptual design framework allows the user to input known parameters for the print map and readily calculate the
necessary parameters for web and mobile maps. As an example, standard line weights in millimetres can be input, and the most
suitable line weight size in pixels can be calculated based on the screen resolution of the monitor the map will be displayed on.
This is a simple mathematical equation based on a number of known values: the monitor resolution (in pixels per inch); and the
line weight (in millimetres), with the equation looking like this:
X =Y x 72/Z
where X = line weight in pixels; Y =line weight in points; and Z = monitor resolution in ppi
This overarching design framework we are calling a ‘design operator’ that can translate known design elements from a print
map to optimal values for screen display. The user has control over the screen parameters by adjusting the monitor resolution,
which in turn adjusts individual design elements as described above.
Conceptually, a design framework for multi-format publishing would consist of the following elements: a spatial database; a
comprehensive set of design assets (consider this a design template or stylesheet); an application to create the map or maps with;
and a ‘design operator’ that translates the design to various media formats (Figure 4).
Figure 4: Conceptual design framework which includes a ‘design operator’ to translate designs to various media formats.
To date, the ‘design operators’ have been developed in a spreadsheet format, incorporating the mathematical equations
required for converting design elements and testing to ensure that each design element translates correctly. The spreadsheet
template also contains map parameters that cannot be readily translated at this stage without the appropriate software, such as print
map projection and sizes translated to TMS Web Mercator projection. It is envisaged that if this design framework was adopted,
these parameters would be calculated through appropriate GIS or web mapping software.
What is the structure of the framework?
Figure 5 is a view of the spreadsheet application containing the design elements (indicated in bold) and the resulting values
(indicated in red) of the ‘design operators’. The design framework template contains the basic information required when planning
to create a multi-published map. In detail they are the following:
• Map size of the print map
• Projection of the print map
• Scale of the print map
Figure 5: Conceptual design framework containing ‘design elements’ which are modified by ‘design operators’, giving resulting
values to assist in creating designs for a multi-published map. The conceptual design framework has been created in spreadsheet
applications for both Mac and Windows.
These parameters are not affected by ‘operators’ as such, but are included for reference and planning, however it is envisaged
that if this conceptual framework was adopted more broadly and automated within a GIS application, factors such as change in
projection from print to TMS could be calculated and extents for the new Web Mercator projection could be created.
When determining a colour palette for print, web and mobile maps the design framework allows the user to input the print
CMYK values and automatically calculate the RGB and Hex values suitable for use on web and mobile map design. The RGB and
Hex values have been calculated by a complex workflow of translations done in the spreadsheet.
The formula for the translations was sourced from open-source information found on easyrgb.com (2012). Conversion
formulae from RGB to HEX values were derived from Cardillo (1996).
Determining the screen resolution of web and mobile devices provides a ‘design operator’ to select appropriate line weights
and symbol sizes. This ‘design operator’ is essential to ensure line weights are correctly calculated when designing for the various
screen resolutions available in modern computers and mobile devices. As stated earlier, resolution will affect the size of tiles
displayed using a TMS, which in turn affects the line weights displayed on screen.
Line weights are variously calculated by simple unit conversions from millimetres to inches, picas and points—all units used
in the printing industry. Pixel widths have been calculated by multiplying the inches value by the monitor resolution, as monitor
resolutions are determined in pixels per inch. A rounding has been applied to the resulting pixel line weight to ensure that an
integer is created, as pixels can only ever be a whole value.
When reading type on screen, the minimum point size of a font is important. This may be larger than the point size used for
display on a printed map, which typically has a minimum of 5 or 6 points, for screen it should be 10 point, (Brown, et al. 2001).
(Note: 1 point = 1/72 inch.) Jenny et al. (2008) suggested that type should be set at a 12 point minimum for screen display, but
suggested that sans serif fonts could be displayed as small as 10 point. For the purposes of this design framework we have
approximated 10 point as a minimum. To allow for this, the ‘design operator’ multiplies print font sizes by 1.75, making 6 point
text 10.5 points in size, 12 point text would become 21 point, and so on. These larger font sizes are to be used as a guide, as
cartographers will want to create their own font hierarchies, however, as Brown suggests, it is recommended to not use font sizes
less than 10 point for type on screen.
The last major feature of the conceptual ‘design framework’ is a utility to determine what scale datasets to use for which TMS
zoom level This is useful for using datasets that come from one data provider, such as Geoscience Australia, that provides various
scale topographic datasets which have been generalised for the particular display scale. The design framework also uses this zoom
level information, in conjunction with the print map geographic extents, to determine the maximum geographic extent for the
uppermost zoom level. This is important for clipping data to the optimal geographic extent when preparing maps for the TMS, as
geographic extents of the print map may fall at a point within a tile’s extents rather than neatly finish at the edge of a tile, (Figure
6).
Figure 6: Difference between print map geographic extents and TMS tile geographic extents.
To calculate geographic extents for the maximum zoom level tiles the uppermost zoom level number have been selected from
the TMS Zoom Scales table (Figure 4), that is the lowest selected number whose value = TRUE, as zoom level selection is
controlled by a Boolean operator. Calculations for converting latitude and longitude extents to tiles and converting tiles to latitude
and longitude were sourced from OpenStreetMap wiki (2012).
Testing of concept
To verify that the conceptual design framework and the ‘design operators’ created do work, a map was produced by the
authors of Wilsons Promontory National Park. The print map design was created with the final map produced in Adobe Illustrator,
using the MAPublisher plug-in to import and export data. Data was styled using existing design specifications which were stored
in a graphic style library within Adobe Illustrator. These design styles were created with a print map specifically in mind.
Character styles were created exclusively for a print map. Figure 7 shows or portion of the map design for print. Of note are the
font sizes and used on the print map, ranging from 6 to 7 point in the sample shown, whilst the web map font sizes range from
10.5 to 14 for the web map. As much as possible colours and styles have been retained.
Figure 7: Left, portion of Wilsons Promontory print map (Spatial Vision, 2012). Right, portion of web map (Spatial Vision, 2012)
using ‘design operators’ to calculate line weights, font sizes, symbols and colours. (Note that the print map is at 1:50,000 scale,
whilst the web map is displayed at zoom level 14, approximately 1:36,000)
A tile set was also created for a mobile device, in particular an Apple iPhone with a retina display with a screen resolution of
264 ppi. As the pixel resolution is double that of the standard screen display, tiles from the TMS get reduced 50%. A scaling
factor of 2 was applied to all design elements, line weights, fonts and symbols, to allow for the halving in size of the tiles (Figure
8).
The ‘design operators’ created for the design framework on visual inspection appear to retain the look and feel of the print
map’s design. The scaling factors for line weights give the user of all maps the impression that they are looking at a consistent
design across different media. The fonts and symbol ‘design operators’ bring a consistency to web maps, however when scaled for
high resolution displays a straight 200% enlargement appears visually too much and will need to be tested further achieve a more
harmonious and consistent result.
Figure 8: Left, portion of Wilsons Promontory map for retina display (Spatial Vision, 2012), with features scaled 200%. Right,
portion of map shown at display size. (Note images shown at at zoom level 14, approximately 1:36,000)
Future development
Naturally there is more to map design than colours, fonts, symbols and lines. More research is required in the area of pattern
representation in web maps, such as how certain patterns display when rendered in pixels. Understanding what conversion factors
are required to scale these patterns needs to be tested further. Likewise, the use of textures, as an alternative to patterns, is an area
with little research. How this could be measured empirically and be controlled by ‘design operators’ would require further testing.
Calculating the correct scale factor for fonts and symbols for high resolution displays will further improve the quality of output for
mobile devices.
The concept of ‘design operators’ in a multi-publishing environment has been tested here in a simplified manner, using a
spreadsheet application to apply mathematical ‘operators’ to design elements. In the perfect world we would like to see these
‘operators’ applied to stylesheets or style libraries from one media to create new stylesheets or style libraries for another media.
To illustrate this concept further, a map style library from Esri’s ArcGIS could have ‘design operators’ applied to it for creation of
a CSS version for use in TileMill—not only would colours, line weights, fonts and symbols be exported correctly, but a readability
factor would be built in to enhance the design for screen display.
Conclusion
Multi-publishing is an evolving area, not just in spatial science but the publishing industry in general. In cartography it draws
on many areas of existing research for print and web map design. Research has been done in areas of generalisation and schema
development, but less so in areas of design.
Designing for multiple media is different for designing for a single media. Consideration needs to be given to colours that can
be displayed in different colour spaces, fonts need to be legible on print and screen, at various resolutions, symbol sizes and
designs need to easily understood at small and large sizes, and line weights and styles will be variable across the media. It is
essential to determine the geographic extents of the map(s) early in the design process to ensure a smooth workflow and save time
and money in the production process. The framework presented here attempts to put some certainty into designs when moving
across media platforms.
The conceptual design framework presented here is very much in its infancy and is a stepping stone to further research in this
area, working towards a multi-publishing workflow for student atlases. The simple spreadsheet application should be further
enhanced to be presented as a web page or built in to an existing mapping application.
The use of a ‘design operator’ will enable the cartographer to convert the design elements of a print map to design elements
suitable for display on screens of varying resolutions, such as those found in the multitude of mobile tablet devices in use today.
References
Apple, 2011. Apple - iPad 2 - View the technical specifications for iPad 2. apple.com. Available at:
http://www.apple.com/ipad/specs/ [Accessed September 30, 2011].
Berglund, A. ed., 2006. Extensible Stylesheet Language (XSL) Version 1.1. w3.org. Available at: http://www.w3.org/TR/xsl11/
[Accessed October 13, 2012].
Bertin, J., 1983. Semiology of Graphics, University of Wisconsin Press.
Brewer, C.A. & Hanchett, C., 2010. ScaleMaster for Proposed Design for Multiscale Topographic Mapping from USGS National
Map Data.
Brown, A., Emmer, N. & van den Worm, J., 2001. Cartographic Design and Production in the Internet Era: The Example of
Tourist Web Maps. The Cartographic Journal, 38(1).
Cardillo, D., 1996. RGB - Hexadecimal Color Conversion. gristle.tripod.com. Available at: http://gristle.tripod.com/hexconv.html
[Accessed October 14, 2012].
Cartwright, W. & Peterson, M.P., 2006. Multimedia Cartography. In W. Cartwright, M. P. Peterson, & G. Gartner, eds.
Multimedia Cartography. Springer, pp. 1–10.
Celik, T. et al. eds., 2008. CSS Color Module Level 3. w3.org. Available at: http://www.w3.org/TR/2008/WD-css3-color-
20080721/ [Accessed May 27, 2012].
EasyRGB, EasyRGB - The inimitable RGB and COLOR search engine! easyrgb.com. Available at:
http://www.easyrgb.com/index.php?X=MATH [Accessed October 14, 2012].
Hofmann, P., 2011. Why design matters. NacIS, pp.1–125.
Jenny, B., Jenny, H. & Räber, S., 2008. Map Design for the Internet M. P. Peterson, ed., Springer.
Keates, J.S., 1993. Some reflections on cartographic design.
Meng, L., 2003. Missing Theories and Methods in Digital Cartography. In International Cartographic Conference. Durban, South
Africa, pp. 1887–1893.
MacEachren, A. M., 1995. How Maps Work.
myWorldAtlas 1st ed., 2012, John Wiley & Sons, Australia.
OpenStreetMap Wiki, Slippy map tilenames. wiki.openstreetmap.org. Available at:
http://wiki.openstreetmap.org/wiki/Slippy_map_tilenames [Accessed October 23, 2012].
Ramsdale, J. et al., 2010. Jacaranda Atlas 7th ed. A. Rossimel & C. Wardle, eds., John Wiley & Sons, Australia.
Robinson, A.H., 1952. The Look of Maps: An Examination of Cartographic Design 1st ed, The University of Wisconsin Press.
Robinson, A.H & Petchenik, B.B., 1976. The Nature of Maps, Chicago, Illinois: University of Chicago Press.
Roth, R.E., Brewer, C.A. & Stryker, M.S., 2011. A typology of operators for maintaining legible map designs at multiple scales.
Cartographic Perspectives, pp.29–64.
Samsung, Galaxy 10.1 - It’s time to tab it with Samsung’s Galaxy 10.1 Tab - OVERVIEW. samsung.com. Available at:
http://www.samsung.com/au/consumer/mobile-phone/mobile-phone/archive/GT-P7500FKDVAU [Accessed October 15,
2012].
Spatial Vision, 2012. Wilsons Promontory 1000 x 810 mm(1:50,000). Available at: http://www.svmaps.com.au.
TileMill | Fast and beautiful maps. mapbox.com. Available at: http://mapbox.com/tilemill/ [Accessed May 27, 2012].