4th International Conference
for Orienteering Mapping

Marianské Lázne, Czechoslovakia - 16-18 August 1991

Instructors Conference Programme

 


Participants:
Ake Jorma, Finland
Binsch Karl-Heinz, Germany
Brogli Thomas, Switzerland
Bylemans Jos, Belgium
Dresen Andreas, Germany
Franco Joao Manoel, Brasil
Grühn Olaf, Germany
Háj Bohumil, Czechoslovakia
Harvey Robin, UK
Hermans Marc, Belgium
Kristiansen Tore, Norway
Krispinsky Pavol, Czechoslovakia
Lejsek Martin, Czechoslovakia
Lenhart Zdenek, Czechoslovakia
Lundsgaard Erling, Danmark
Norgaard Flemming, Danmark
Novotny Milan, Czechoslovakia
Parfianowicz Ludomir, Poland
Persson Björn, Sweden
Rasmussen Torben, Danmark
Rönning Olav Dag, Norway
Rehák Pavel, Czechoslovakia
Sartori Alfredo, Italy
Shaw Chris, UK
Szabó Gyula, Hungary
Steinegger Hans, Switzerland
Tarabocchia Cesare, Italy
Yu Sai Hong, Hong Kong
Václavík Martin, Czechoslovakia
Veselovsky Rastislav, Czechoslovakia
Zentai László, Hungary

?, New Zealand
Matthes Gisela, Germany

The time and theme schedule have been fixed as follows:
16.8. 17.00 Opening of the Conference
20.00 New Development in Computer aided mapping (Lecture)
(Flemming, Robin: Mac; Hans Steinegger: OCAD, others)
17.8. 9.00 New Standards for Orienteering Maps (Lecture: Flemming)
10.00 Survey and Interpretation of Vegetation in Continental Terrain (fieldwork in groups: Andreas)
14.00 Discussion and Summary of the Fieldwork (Andreas, other commitee members)
16.00 Information from different Coutries, open Discussion (all)
20.00 Demonstratoin: Computer aided mapping
18.8. 9.00 Models for Instruction of Beginner Map-makers (Thomas)
11.00 Final discussion, Conclusion (Thomas)
12.00 End of Conference (Thomas)

Vortrag zum Internationalen Kartenzeichnerseminar in Mariánské Lázne, CSFR

Darstellung der Vegetation in OL-Karten


Die Vegetation wird dargestellt, um dem Wettkämpfer Informationen über die Belaufbarkeit des Geländes und Hilfen zur Orientierung zu geben.

Es stehen folgende Grundfarben zur Verfügung:
Weiß: typischer gut belaufbarer Wald
Gelb: offene Flächen, Äcker, Wiesen
Grün: reduzierte Belaufbarkeit

Weitere Farben können ebenfalls Hinweise auf die Belaufbarkeit geben:
Grau: gut belaufbare offene Felsflächen
Schwarz: steiniges Gebiet, Felswände, Sandflächen
Blau: Sümpfe und Moore

Weiß zeigt den für das jeweilige Gebiet typischen gut belaufbaren Wald. Falls kein Teil des Waldes gut belaufbar ist, soll kein Weiß auf der Karte erscheinen.

Die Farbe Grün gibt hinweise auf die Belaufbarkeit, die wieder vom Waldtyp abhängt (Dichte der Bäume, Buschwerk und Untergrund wie Farnkraut, Brombeeren, Brennesseln usw.). Einschränkungen in der Belaufbarkeit, die bereits durch andere Signaturen dargestellt sind wie Sümpfe, steiniges Gebiet, Sandflächen usw. werden nicht zusätzlich noch mit Grün gekennzeichnet. Das bedeutet, dass die vom tiefen Untergrund her gegebene Einschränkung der Belaufbarkeit eines Sumpfes bereits ausreichend durch die Symbole 309-311 gekennzeichnet sind. Erst wenn zusätzlich eine übermäßig dichte Vegetation vorhanden ist, wird die Farbe Grün benutzt.

Die Belaufbarkeit des Waldes ist der Laufgeschwindigkeit entsprechend in 4 Stufen eingeteilt. Als Grundmaß (Basis) dient die Zeit, die ein Läufer für eine Normstrecke im offenen Wald benötigt. Der Grad der Behinderung wird an der reduzierten Strecke gemessen, die der Läufer bei dichter Vegetation in der gleichen Zeit laufen kann.

Beispiel:
Auf der Normstrecke von 1 km läuft ein Wettkämpfer der Klasse H21 im offenen Wald eine Zeit von 5 Minuten. In einem dichten Bestand schafft er in der gleichen Zeit nur 50% der Strecke; das entspricht 500 m oder einer Laufzeit für einen Kilometer von 10 Minuten.

Die Belaufbarkeit wird in folgende Stufen eingeteilt:

       Laufstrecke in Normzeit       Zeit/km in Minuten
gut belaufbar: 80 - 100% 5:00 - 6:15 leichte Behinderung 50 - 80% 6:15 - 10:00 starke Behinderung 10 - 50% 10:00 - 50:00 sehr starke Behinderung 0 - 10% > 50:00

Tabelle:
Die letzte Belaufbarkeitsstufe darf somit nur dann in der Karte erscheinen, wenn der H 21 Läufer mit einer Normzeit von 5 Minuten auf einen Kilometer Strecke im Dickicht mehr als 50 Minuten benötigen würde!

Vergleicht man diese Anforderung mit der Praxis in Mitteleuropa dann ist festzustellen, daß die Farbe Grün zu häufig in den OL-Karten erscheint und daß insbesondere die letzte Stufe zu großzügig benutzt wird. Manche Kartenzeichner erklären die unrichtige Darstellung mit dem Hinweis, daß sie die Dickichte bewusst etwas überbetonen, damit die Wettkämpfer diese Gebiete meiden. Da das Betreten von Dickichten in einigen Ländern grundsätzlich verboten ist, wird durch die Farbe Grün teilweise die Darstellung als "Sperrgebiet" ersetzt.

Diese Praxis muß von einem internationalen Verband abgelehnt werden. Dafür gibt es verschiedene Gründe:
  1. In der Regel kennt ein ausländischer Wettkämpfer nicht die speziellen Forstgesetze und das Betretungsrecht des jeweiligen Gastlandes. Der Veranstalter ist also verpflichtet, ihm in der Karte solche Gebiete klar und eindeutig zu bezeichnen, die er nicht belaufen darf. Dazu ist die Darstellung für "Sperrgebiet" (529) oder "Verbotenes Gebiet" (607) zu verwenden. Eine allgemeine Festlegung der dritten Grünstufe als Sperrgebiet ist dagegen untauglich, da in vielen Ländern der Erde ein großer Teil des Waldes sehr schwer passierbar ist, ohne daß diese Eigenschaft das Betreten verbieten würde.
  2. Wenn ein Wettkämpfer im Verlauf des Wettkampfes feststellt, daß die Grünstufen stark überzeichnet sind, wird er sie im weiteren Rennen kaum noch beachten. Da die dritte Grünstufe in manchen Karten bereits ab einer Reduktion auf 50% der Strecke gezeichnet wird, unterschätzt der Wettkämpfer im weiteren Verlauf des Rennes echte Dickichte und wird auch solche Gebiete durchlaufen, die er bei richtiger Darstellung gemieden hätte. Der gewünschte Effekt wird also verfehlt und es entstehen vermeidbare Auseinandersetzungen um Proteste und Disqualifikationen.
  3. Die Vegetationsdarstellung beeinflusst die Routenwahl des Wettkämpfers. Eine bewusste Fehlinformation wirkt sich regelmäßig zur Benachteiligung der Läufer aus, die mit der abweichenden Darstellungsweise des Kartenzeichners oder dem Geländetyp wenig vertrat sind. Es entsteht ein ,Heimvorteil", der als grob unsportlich bezeichnet werden muss. Auf der anderen Seite haben einheimische Wettkämpfer, die wenig im Ausland trainieren, dort erhebliche Schwierigkeiten, sich an die international gebräuchlichen Maßstäbe zu gewöhnen.
Aus den genannten Gründen ist streng darauf zu achten, daß die Belaufbarkeitsangaben in der Karte mit den Vorgaben der IOF-Darstellungsvorschriften übereinstimmen.


Erläuterung der Kartenzeichen mit Bedeutung für die Vegetationsdarstellung

401 Offenes Gebiet
Kultiviertes Land, Felder, Wiesen, Weideland usw. ohne Bäume, das leicht belaufbar ist.
Farbe: gelb.
402 Offenes Gebiet mit einzelnen Bäumen halboffen
Wiesen mit einzelnen Bäumen oder Büschen, Gras oder ähnlicher Bodenvegetation, die leicht belaufbar sind. Sehr kleine Gebiete werden wie offenes Gebiet (401) dargestellt.
Farbe: gelb 50% (16,75 Linien/cm).

403 Rauhes offenes Gebiet
Heide, Moorflächen, Kahlschläge, Neuanpflanzungen (Bäume kleiner als ca. 1 m) oder anderes überwiegend offenes Gelände mit rauher Bodenvegetation, Heidekraut oder hohem Gras. Die Signatur 403 kann mit den Signaturen 407 und 409 kombiniert werden, um reduzierte Belaufbarkeit darzustellen.
Farbe: gelb 40% (40 Linien/cm).

404 Rauhes offenes Gebiet mit einzelnen Bäumen
Wo in rauhem offenen Gebiet einzelne Bäume stehen, sollen weiße (oder grüne) Flächen im Raster erscheinen. Solche Gebiete können durch ein regelmäßiges Muster von großen weißen Punkten im gelben Raster generalisiert werden.
Farbe: gelb 50% (40 Linien/cm), weiß 50% (16,75 Linien/cm).

405 Wald
Offener Wald ohne Behinderung für den Läufer. Farbe: weiß.

406 Wald: Leichte Behinderung
Ein Gebiet mit dichten Bäumen (geringe Sichtweite), die die Laufgeschwindigkeit auf ca. 50-B0% der normalen Geschwindigkeit verringern.
Farbe: grün 20% (40 Linien/cm).

407 Bodenbewuchs: Leichte Behinderung
Ein Gebiet mit dichter Bodenvegetation jedoch sonst guter Sichtweite (hohe Brombeersträucher, Heidekraut, niedrige Büsche einschließlich abgeschlagene Äste), die die Belaufbarkeit auf ca. 50-80% der normalen Geschwindigkeit verringert. Diese Signatur kann mit 403 und 404 kombiniert werden.
Farbe: grün l7% (13 Linien/cm).

408 Wald: Starke Behinderung
Ein Gebiet mit dichten Bäumen oder Dickicht (geringe Sichtweite) die die Laufgeschwindigkeit auf ca. 10-50% der normalen Geschwindigkeit verringern.
Farbe: grün 50% (40 Linien/cm).

409 Bodenbewuchs: Starke Behinderung
Ein Gebiet mit dichter Bodenvegetation jedoch sonst guter Sichtweite (Brombeersträucher, Heidekraut, niedrige Büsche einschließlich abgeschlagene Äste), die die Belaufbarkeit auf ca. 10-50% der normalen Geschwindigkeit verringert. Diese Signatur kann mit 403 oder 404 kombiniert werden.
Farbe: grün 33% (27 Linien/cm).

410 Vegetation: Sehr starke Behinderungo unpassierbar
Ein Gebiet mit dichter Vegetation (Bäume oder Bodenbewuchs), das kaum passierbar ist. Die Laufgeschwindigkeit ist auf ca. 0-10% der normalen Geschwindigkeit verringert.
Farbe: grün 100%.

411 Wald in einer Richtung belaufbar
Wenn ein Waldgebiet in einer Richtung gut, in anderer Richtung jedoch schlecht belaufbar ist, zeigen weiße Streifen in der Rastersignatur die Richtung der guten Belaufbarkeit an.
Farbe: grün.

412 Obstgarten
Land, das mit Obstbäumen und -büschen bepflanzt ist. Farbe: gelb 100% und grün 20% (12 Linien/cm).

413 Weinberg
Die Linien können zur Angabe der Pflanzrichtung danach orientiert werden.
Farbe: gelb 100% und grün.

414 Deutliche Kulturgrenze
Die Grenze von Kulturland wird mit einem schwarzen Strich gezeichnet, wenn sie nicht durch andere Signaturen (Zaun, Wall, Pfad usw.) dargestellt sind. Eine dauerhafte Grenze zwischen verschiedenen Kultur1andarten wird ebenfalls mit dieser Signatur dargestellt.
Farbe: schwarz.

415 Kulturland
Felder, die wegen der Nutzung zeitweise Sperrgebiete sind, können mit einem schwarzen Punktraster dargestellt werden.
Farbe: gelb 100%, schwarz 5% (11 Linien/cm).

416 Deutliche Vegetationsgrenze
Ein deutlicher Waldrand oder eine sehr deutliche Vegetationsgrenze im Wald.
Farbe: schwarz.

417 Undeutliche Vegetationsgrenze
Undeutliche Grenzen zwischen Gebieten mit grüner oder gelber Färbung werden ohne Linien dargestellt. Der Rand des Gebietes wird nur durch den Wechsel der Farbe oder des Punktrasters angegeben.

418, 419 Spezielle Vegetationsobjekte
Mit den Signaturen 418 und 419 könne kleine Vegetationsobjekte dargestellt werden. Die Beschreibung der Signatur muß in der Kartenlegende angegeben werden.
Farbe: grün.


Sonstige Kartenzeichen mit Bedeutung für die Vegetationsdarstellung

211 Offene Sandfläche
Eine Fläche mit weichem Sandboden oder Kies ohne Vegetation, die langsam belaufbar ist. Wenn eine offene Sandfläche gut belaufbar ist, wird sie als offenes Gebiet dargestellt (401/402).
Farbe: schwarz 10% (24 Linien/cm) und gelb 50% (40 Linien/cm).

212 Nackte Felsplatten
Ein flaches Felsgebiet ohne Bodenbedeckung oder Vegetation wird als nackter Fels dargestellt. Mit Gras, Moos oder niedriger Vegetation bedeckte Felsplatten werden als offenes Gebiet dargestellt (401/402).
Farbe: schwarz 30% (54 Linien/cm) oder grau.

310 Sumpf
Ein passierbarer Sumpf mit deutlicher Begrenzung. Die Signatur 310 kann mit der Signatur 403 für deutliche offene Sümpfe verbunden werden. Für dicht bewachsene Sümpfe kann die Signatur 310 mit den Signaturen 406-410 kombiniert werden. Der kleinste Sumpf soll mit mindestens 2 Strichen auf der Karte gezeichnet werden.
Farbe: blau (ev. mit gelb/grün).

311 Undeutlicher Sumpf
Ein undeutlicher oder zeitweise vorhandener Sumpf, eine Übergangsform vom Sumpf zum festen Untergrund, die passierbar sind. Die Begrenzungen sind in der Regel undeutlich und die Vegetation ist ähnlich wie die des umgebenden Geländes. Die Signatur 311 kann mit der Signatur 403 für offene, jedoch undeutliche Sümpfe verbunden werden. Gelegentlich kann die Kombination der Signaturen 311 und 401 angezeigt sein. Für dicht bewachsene Sümpfe kann die Signatur 311 mit den Signaturen 406 - 410 kombiniert werden.
Farbe: blau (mit gelb/grün).


The DOLK Group
Working Group for Computerized Mapping within the Swedish Orienteering Federation

Ulf Levin, Björn Persson
SOME ECONOMIC ASPECTS ON COMPUTERIZED ORIENTEERING MAPPING
- and some words about the technology

Distributed at the IOF Instructors Conference, Marianske Lazne 1991/08/16 -18


1. Introduction
In most professions, traditional equipment today is replaced by computers. Several times this is done in a process where the whole organisation within a company etc is modified. The motivation for these changes is to increase productivity.
Since a couple of years ago even O-mappers have started to use computers in their work. People with solid interest for maps and computers are constructing systems for this specific purpose. The work involves programming of a computer in one way or another. In some cases just slight modifications are made to existing software packages (application programming). The output from these systems, no matter what kind, is a map that looks just as if it was made by hand.
Several systems for O-mapping have been demonstrated to the public, at seminars, conferences etc. In some articles they have also been described and compared technically.
The publicity given to the subject of computerized mapping have raised expectations. But will the expectations ever be fullfilled? Do O-mappers want, or need, to raise productivity? Will maps be cheaper to produce with computer support? What is good investment for the O-club?
The aim for this article is to discuss economy rather than technology.

2. Background
O-maps have been produced since over 30 years, by runners and professionals. During this time the methods of production have changed. O-mappers have not been late to start using new technology in the fields of graphic and printing. Today production of 0-maps is a process that involves a row of advanced steps, from photogrammetry to offset printing.
The time during which experiments have been made with computers involved in the process, represents a big part of the O-mapping history. For example the very first swedish attempt was made 15 years ago at the Royal Institute of Technology in Stockholm. The equipment used was very advanced for its time. Today similar experiments are made with ordinary office computers.
The experiments made until now typically involves the steps of mapping known as cleancopying, redrawing (with ink pens, typeletters etc) and photographic reproduction. In general, people who are involved in these activities have their income from other activities. Their motivation is mainly their interest for the technology itself.
In some cases, computers are used professionally by mapmakers. Until now, the new tools is not very much spread though.

3. The database and its value
In the map investors (O-clubs) point of view, the new technology doesn't seem to mean a lot of difference yet either. It hasn't made any big changes to the professional mapmakers pricelists.
Although the motivation to use the new technology seems small today, it might grow in the future. That is if we use a new approach to the technology, and to O-mapping itself, a database approach.
From the computer system, we do not only get the films for printing, but also a disk full of data. Provided that those data are more durable than a traditional map, we have made a good deal, even if the price is the same. Below, data of a map represented in digital form, is called a database.

3.1 Why database
Investing in a database, is a much different thing than to invest in a computer. It means a completely new approach to the subject of computerized mapping. The data becomes a product by itself. The output, the printed map, becomes a secondary product, made by a query in the database.
The database investor doesn't necessarily need to own a computer by himself, although that might be of great advantage. The service of digitizing his map information, to build his database, can be bought from someome else.
One major difference to the database approach, in comparison to traditional mapping, is the time horizon. If handled in the right way, digital data can never be "worn out", but it can easily be copied, modified etc. This is one characteristic of the database that makes it worthful.
The long time horizon makes it difficult to evaluate a database investment. Benefit will appear years after the first map is produced from the database. It's quit obvious that no one will pay a lot extra to get a database under these circumstances. Thus the database approach will only be relevant to O-clubs if it doesn't make mapping a lot more expensive than today.
The judgement wether it is relevant or not must be made by investors in the future. Here are only a few more aspects on the database, concerning its value.
What is called a database can be a worthless thing. That is if the system used for c rea-
ting and storing the database has its own way of accomplishing this, far from any standard. It is impossible to: tell when this system is oldfashioned and off from the scene. To be sure of having any use of the database in the future, a high degree of standardisation is needed.
It is likely that a database will be used many times in the future and each time save a lot of money. Maps don't have to: cover the same area each time and thereby the production of output maps will be more efficient. It is possible to: optimize map production with respect to: different products from a given area.
The value of a database is due to: the need for it, and its quality.
The need for a database is due to: physical qualities of the area to: cover, the forest. The optimum case is a large and beatiful forest, not to: far from cities (the market).

3.2 Database quality
The quality of a database consists of flexibility and currency. Flexibility means possibility to:
- change the look of a map (colors, symbols etc)
- choose the grade of information (which type of objects etc)
- connect areas (put two ore more maps together into one database)
- o change computer system (transferring data)

Currency means that the information always has to: be up-to-date. Each change in the "real world" must be followed by a change in the database. Changes can be made constantly or just before a map is to: be produced.

4. DOLK
In Sweden, orienteerers have discussed the database approach on O-mapping since a couple of years ago. The discussion has mainly been held within an official committee of SOFT (Swedish Orienteering Federation). This committee consists of professional people from different areas in the digital mapping arena (photogrammetrists, computer mapping experts, mapping s/w developers, O-map producers etc) who also carry a great interest for orienteering and especially O-mapping.

4.1 The goals for DOLK
The aim for the "DOLK-committee" is to: gather information and knowledge on the computer assisted O-mapping subject. This knowledge is ment to: be used to: establish different kinds of standards and recommendations for O-map production in Sweden. The results of the committees work will be distributed to: clubs and the O-mapping society in the form of written reports, as of which the first was delivered in march 1991.
Two main questions to: answer are:
- Under which conditions will the database be a resource for future mapmakers? Can actions be made to: optimize the benefit?
- How does the market for our information look? How can the database be used besides making a traditional O-map in the 1:15 000 scale?

The DOLK committee just finished a first step of work. The conditions for database production have been studied briefly. More studies are yet to: be made before standards are recommended.

4.2 The DOLK main thesis
The work that is going on in the DOLK committee is based on a main thesis. This thesis is that benefits from using computers in the O-mapping process will be optimized when two main features are combined:
- high quality on geometry
- standardized data structuring and coding
When these two features are combined, the database will be flexible and useful.

4.3 Geometric quality
The first feature means that the quality of measurements must be better. This means that the basemap must be produced with a degree of geometric quality such that it is possible to: directly put together two databases produced at different sites and at different times without getting connection problems.
DOLK has made a study on how the basemap is produced today and how this process must be improved in order to: achieve the degree of accuracy needed.
The basemap must be constructed upon a geodetic (ground) coordinate reference system. Optimal is to: use a common coordinate system that other public maps (e.g. topographic and economic maps in scales l:50 000 - 1:10 000) and databases is based upon. This means possibilities to: exchange data from the O-mapping database with public GIS-databases and vice versa.
The improved geometry and establishing of a coordinate referenced basemap has its price. DOLK expects this price to: decrease in the future as a result of further developments of new technologies in the whole field of surveying and photogrammetry (sattellite positioning, full-digital photogrammetry, satellite image processing etc).

4.4 Standardized structuring and coding
The second feature above is connected to: the computer system and its way of handling the information. Coding of objects in the map (knolls, houses etc) must be made in a standardized way. This standard will be similar to: the existing "Drawing specification" but somewhat more detailed.

5. Computer systems for database construction
In the past few years, several experiments have been made where maps are produced using computers. The systems used have been of different types and origins. In most cases general purpose grahic systems (such as CAD systems, publishing systems etc) have been modified to suite the 0-mapping task. However there exists two systems that are designed for one single purpose, 0-mapping.

5.1 Basic functions of a mapping system
A computer system to be used for mapping and database construction needs some basic functions. The user must obtain the following three things:
- high accuracy on the measurements (good s/w tools for digitizing, high resolution of database)
- standardized structure on the data (objects correctly coded etc)
- good presentation of the map on the graphic screen (quality WYSIWYG)
The first two features are similar to those discussed above concerning the database. The third feature describes the system as a tool for interactive mapping. The map will not be correct if it is not possible to get a correct picture of it on the screen, this means both geometrically (e.g. how close could two stones be mapped without interferring in printing) and the graphic definitions (e.g. combination of different types of rasters in complex vegetation areas).
For people with some experience in graphic s/w etc, these features might not seem difficult to find in exsiting products on the market. The problem though, is that only a few (if any?) offers all three features together.
As the standard for map data is not yet set, it would not be correct to state which of the existsing systems is the best. It seems though as if the "home-made" systems, made specifically for O-mapping, do have a bright future, given that they are able to communicate with the outside world via a well-known graphic standard.

6. Final comments and conclusions
Computerized mapping can under certain circumstances be of economic importance. Considering a single, short term project, the advantages seems less, but with a long run perspective and standardized data of good quality they seem to be more.
Since a year and a half a new committee, the "DOLK-group", is established in Sweden, dealing with questions on computerized mapping and database construction.
On each place, the benefit from having map information stored in a computer is different. The optimum case is a big forest (big enough for production of seveal maps) with high attractivity.
Flexibility gives the value to a database. A flexible database is achieved by combining digital technologies with geometric accuracy.
The best system to handle O-maps is one made specifically for that purpose. Two such systems are developed until now.
After setting a standard on data structures, coding etc, tools must be developed to handle the 0-map information in that specific way. Existing systems might need to be slightly modified.
Today advanced systems are too expensive to be used by amateurs. Less advanced systems will be needed in the future to handle existing databases.

The DOLK committee is a special committee for computer assisted mapping, under the Swedish Orienteering Federation Map Committee.
Ulf Levin is a member of the Working Group II of the DOLK committee, with special interest in data structuring and data transfer questions.
Björn Persson is chairman of the DOLK committee.


A Part-Guide to to Map Drawing using a Computer
Barry McCrae (AUS)


The IOF Mapping Committee apparently has plans to produce a 'Computer mapping' booklet. This is a formidable task with the continual development in computer hardware and software. Nevertheless, it is an important task as published guidelines will both encourage more activity and assist those already involved to be more productive. Enough people now have experience in the various forms of computer mapping for their cumulative wisdom to be sufficient to provide newcomers with a step-bystep guide to each method, complete with helpful hints and pitfalls to avoid. There should no longer be a need for everyone to 'reinvent the wheel'.
It is hoped that this paper will provide the framework for the development of such a guide to one approach to computer mapping, namely the `scan and draw' method in which an object-oriented drawing program is used to draw the final map from a scanned version (survey draft or fieldwork). I emphasise that this paper is only a starting point - I ask almost as many questions as I answer. I am hoping, however, that the experts who are present, and those with limited experience like myself, can provide the answers necessary to allow the speedy production by the MC of a guide of the type envisaged.
I have computer-drawn two five-color O maps to date, in each case using the Adobe Illustrator program with an Apple Macintosh computer. The advice of Flemming Nsrgaard, who is presenting this paper for me. has been of invaluable assistance throughout and much of his advice is incorporated in what follows. I thank him on all accounts and fm sure that he will agree that any errors, however, are all mine.



1. System Requirements

1.1 Macintosh
- minimum hardware: Macintosh LC 4/40, hi-resolution monochrome monitor
- preferred hardware: 68030 CPU (SE/30 - portable, but in-built 9"
screen very limiting; IIsi - with coprocessr; IIci 5/40 - ideal), with double-page monochrome monitor and hi-resolution color monitor (or double-page color monitor!)
At least 4MB seems necessary for comfort given known system file and drawing program aims. Color monitor helps with checking; double-page monitor almost essential for layout work.

- software: Tine Type or Adobe Type Manage for high-standard text display
Adobe Illustrator 3 or Aldus Freehand 3 as drawing program

A study of the specifications for these two programs aaggeat that Freehand should be preferred, but Illustrator is already established in the O world.
In summary:

 
Illustrator
Freehand
main advantage `tried and true' - help available including legend of O symbols and patterns draw in layers; multiple levels of undo
main disadvantage single-layer drawing does performance match specifications?
who can you turn to for help?

1.2 IBM
- minimum hardware: 386-SX?
- software: CorelDRAW!

1 need help here. Trevor Hughes (AUS) should be asked for comment as he has recently finished a couple of maps using CorelDRAW!

1.3 Peripherals
- black-and-white scanner
- black-and-white laser printer
- color laser printer for final OR (the ultimate) a color photocopier with Postscript nterpreter (eg Canon PS 300)
- imagesetter far film output

Ready access to a black-and-white lass pnnter is essential for proof ng at regular intervals. You can rely on a bureau service for the other facilities although it is preferable to be able to do your own scanning.


2. Fieldworking and Preparing the Survey Draft

2.1 Is it necessary to prepare a survey draft?

 
advantages
disadvantages
preparing a survey draft for scanning - field work discrepancies resolved
- result more likely to be auto-traced succesfully
- errors can be introduced
- time-consuming
no survey draft (ie scan fieldwork) - the final map will be prepared direct from the actual fieldwork
- saves time??
- fieldwork must be done on separate sheets* (see 2.2)
- fieldwork presentation must be of a high standard (see 2.3)
- rub-outs, scrape-outs, dirt etc. may show on the scanned image.

*This is not a bad idea anyhow. Having tried each method. I sm not convinced that time is saved overall by scanning the fieldwork luelt What is the consensus?

2.2 Dividing the survey draft into parts
The black-and-white scanned image of a composite survey draft would be difficult and confusing to manual-trace accurately, and almost useless for auto-tracing. Consequently, it is necessary to divide the draft into three (or perhaps two) registered sheets, with each sheet being scanned separately to produce a sequence of templates for use at the drawing stage. Brown features should be on one sheet, black and blue on a second sheet. and vegetation boundaries on the third (although, if there is not too much overlapping, these may be included on the second sheet).

Some exceptions to this scheme of division may be desirable. For example. I think there is a good case for including cliffs on the contoar (brown) sheet. What do others think?

Use a different color for each sheet (or even each type of feature) so that a color photocopy of the registered composite draft can be made for reference during the drawing stages. Avoid using green, though, as this may not scan well.

It may seem that using a color scanner would avoid ft need for distributing the survey draft between two or three sheets. However, although the result may be better for manual tracing, it will be no better for auto-tracing until either color processing software is developed to the stage where the scan can be separated into its `pen/pencil-color' components or auto-tracing algorithms can distinguish such colors from each other. This is something to look forward to!

2.3 Specifications for the survey draft
Firstly, as for conventional cartography, it is important that actual symbol sizes arc used and minimum separation distances am observed. Additional requirements are designed to help optimise the clarity of the scanned image and maximise its potential for auto-tracing.
Some suggestions follow.

A map with extensive form lines seems to present a good cases for preparing a survey draft prior to scanning ones auto-tracing becomes the norm rather than the exception. The fieldworker can then draw them as brokf line in the normal marina, whereas the survey draft can show them as continuous lines (in a different color to the contours for easy tefetame during the drawing stage). Yes or no?


3. Scanning

3.1. Scanner Settings
In AppleScan terminology, the draft map (survey draft or fieldwork) should be scanned as line art at 75 dpi with the More accurate size option enabled. This means that the screen display of the scanned image will be the same scale as what was scanned. As mentioned in #4.2, the quality of the image is far from ideal for tracing, however, and I wonder whether it would be better, if the facilities exist, to scan a 50% reduction of the draft map at twice scale. Worth trying, or have I overlooked a technicality?
I'm amuning that the computer monitor is the Apple standard of 72 dpi, or thereabouts. It would a great boon to manual tracing if there was a sudden development of 150 dpi screens!

3.2 The mechanics of scanning
It is likely that you won't be able to scan the entire draft map in one go - perhaps because it's too big (maximum size for the Apple Scanner is 21,6 x 35.6 cm, although A3 scanners are relatively common) or because you are drawing the map piece-by-piece as the fieldwork is completed. There are two possible approaches to `piecing it together'. You can either build-up a full-size template piece-by-piece using SuperPaint 2 (or equivalent software) to put the scanned pieces together, or you can draw each piece separately and use the cut-and-paste facility to combine the drawn pieces with your drawing program. (Which method is preferred?) Large maps may need to be drawn in pieces anyhow as the time taken to manipulate the complete drawing file may become frustratingly long.
Whichever approach is taken, the survey draft sheets need to be carefully gridded to ensure that each scanned piece can be registered exactly with the corresponding piece on the other sheet(s), aligned properly on the scanner glass, and pieced together correctly with its neighbours. The latter can only be achieved successfully by ensuring that each scanned image overlaps with its neighbours. Or are there other ways?

Care should be taken to get the alignment and registration exact to save subsequent frustrations. You should also experiment with the threshold (or equivalent) setting to obtain an image which has maximum 'crispness' in its lines yet contains all the fine details (eg rocky ground dots).

3.3 Preparing the wan for tracing
The scan can be 'cleaned-up' by deleting stray and unwanted pixels using the editing facility of the scantling program. An 'off square' image can be totaled into alignment using SuperPaint 2. Indeed, it is sometimes convenient, or even necessary, to scan some pieces of map 'south m north', or at right angles, and then rotate the images.


4. Drawing

4.1 Basic approach

In artwork mode, the final map will be an almost unrecognisable complex of crisscrossing hairlines. It is essential, therefore, that you adopt a systematic approach to drawing from the start to ensure that colors or layers are not confused in the final product. A recommended approach (at least for your first map) is to work with the following layers (from bottom to top): 1. yellow map detail; 2. green map detail; 3. brown map detail; 4. blue map detail; 5. black map detail; 6. magnetic north lines (with the Overprint option specified if they are not black); 7. all 'non-map' information (eg border, tide. legend, credits). (Is there agreement with this recommendation?)
There are arguments for some exceptions - for example, grouping lake and road in-fills with their boundaries in layer 1. However, such exceptions should be carefully considered and kept to an absolute minimum (if in doubt, don't make an exception!). The simpler you keep it; the easier it will be to proof and the less likely that features will be unintentionally hidden in the final product.

Freehand is designed to allow you to work in layers (or levels as they am called) which it keeps separate. With Illustrator, it's up to you to use the Group, Lock and Hide commands to keep your layers separate. As drawing proceeds, this requires concentration and a practised routine of steps. What is the situation with CorelDRAW?

The order in which you draw your layers differs from their final 'stacking order'. It is recommended that you draw the brown layer first, the black and blue layers next, then green and yellow, and the 'non-map' information layer last. (Or should blade and blue come first?) You should reduce the drawing to final scale before adding the 'nonmap' detail so that WYSIWYG!

The legend and layers can be drawn at twice stale in another document, reduced to final scale, and then placed as EPS files or pasted into the map via the clipboard.

Your systematic approach should include making back-up copies of the map file at regular intervals, and keeping each version until the map is printed. This provides a safeguard against continually discovering errors introduced at a much earlier stage.

Such a situation can occur when you work for some time with part of the map that is already drawn moved out of the work area, and thus out of viewing and printing range. On two occasion I have somohow lost some detail when it has been out of the work arcs for some time. In both eerie 1 lit all isolated detail in one or more colors - it local point and area features, and isolated line features (eg cliffs), I suspect that in at least one case the was caused by the inwniption of a macro for grouping, hiding or locking, so now I stick to keyboard shortcuts (command-G. etc) where they are available. But then again, Illustrator is a large and complex program and the cause may be a software bug. Has anyone bad similar problems?

4.2 Drawing contours
Much help is needed here - this is a biased personal account!
There are three basic approaches: you can use the autotrace tool, the freehand tool of the pen tool. The quality of the template is crucial to all three approaches but especially, of course, to autotracing. Otherwise, autotracing depends directly on the sophistication of the autotrace algorithm, freehand drawing on your ability to smoothly and deftly trace with the cursor, and drawing curves with the pen tool depends on your ability to place anchor points and direction lines appropriately and economically.

The first two approaches would seem to offer speed advantages over the pen-tool approach but, for me at least, this is not the case because of the excessive amount of subsequent editing with the pen tool that I find necessary to achieve a perfect trace. To snake matters worse, the relatively poor quality of the 75 dpi template means that I generally find myself drawing at 2:1 scale and editing at 4:1 (it 8 times final map scale/). This converts to MUCH longer than when drawing contours with pen and ink, and more than offsets the considerable speed advantage otherwise evident (in drawing lines, placing point symbols, filling areas, etc). No doubt those of you who draw neaps more often this way are much quicker than I am - you'd have to be if you are doing it for a livingl I suspect, however, that my experience would be fairly typical of any spare-time map drawer and this leads me to conclude that the unreserved endorsement of this whole approach to O map making must await the development of better autotrace algorithms. Then there may well even be a time gain over pen and ink drawing to accompany the acknowledged gain in quality of the printed map (with its consistent line thicknesses, dash lengths, dot sins, etc). Fair comment?

Has anyone tried scanning contours at 300 dpi and using a specialist tracing package such as Streamline 2? Has anyone used the freehand tool with a' 'Mouse Pen'?

4.3 Over-printing
You should be aware that, by default, drawing programs do not overprint so that any part of an object that is overlapped by - drawn behind - another object will not overprint unless the Overprint attribute is selected for the object. This is very useful in certain drawing situations (see #4.4) and means that, with good registration, colors will be 'true' throughout on the final map (eg watemourses across open land won't appear green as the yellow, separation will have effectively been 'masked' by the blue (and brown!) separations), Note, however that black always overprints.
Three situations come to mind when you must overprint: blue MN lines - else they will disappear behind road in-fills; yellow (or green - whichever color is in front in the new settlement 'pattern'); and registration crosses - else only the top one (black) will actually exist. Are there other cases?

This discussion assumes the layer ordering recommended in #4.1

4.4 Drawing tips
o Most importantly, obtain a library file of O symbols and patterns, drawn to IOF specification, from the (IOF MC? Compucort? Harveys?).
o Watch-out for 'bits of dashes' at the end of tracks, watercourses and faro lines. Control them by extending the line, or by using the scissors tool to shorten the line or to lengthen the last dash (this occurs because the line, and the dash pattern, start again at the cut). Do this when you draw each line to save a lot of extra proofing later on?
o Handle situations like a large open area with isolated forest and thicket areas by drawing appropriate white and green patches in, front of a 'non-holed' yellow fill.
o Take advantage of the default 'blocking-out' of overlapped areas to achieve perfect abutting in situations like adjacent light and medium green areas.
o Use a Copy and Paste-in-front (or in-back) combination to save re-drawing segments of roads, fences, etc which are also vegetation boundaries.
o Split the screen into two side-by-side windows to preview your artwork simultaneously in Illustrator. (I think you can draw in preview mode in Freehand). This is handy when you arc experinxndng with the distance between objects and just where dashed lines are cut, etc, but of course slows the program down.
o You can 'speed-up' the working of the drawing program by Hiding those layers not needed at the time and, with a color monitor, working in black-and-white.
o The Direct-select tool can be used in Illustrator to save Ungrouping layers (and thus inviting disaster). Beware, though, that if the object you select is a grouped object (eg a circle) then you are not selecting all of it!
o Tidy-up your pattern fills in illustrator by moving the pattern within an object using the Transform option in the Paint style dialog box. Remember, though, to elect to Transform pattern tiles in subsequent transformations (including moving and resealing).
o Come on guys, share those secrets! How do you get a dashed line to 'reverse' so that the dash pattern begins at the other end, without redrawing it from that end? (Cut-it, then rejoin??)

4.5 A word or two about layout
Remember the map, and not the surrounding information, is what is of prime importance to the competitor. Resist the utnptadon to use your new found drawing skills to experiment to excess with ornate borders and titles, colorful backgrounds, fancy layers and outer eye-catching distractions.

Do include some registration marks for the competitor's reference. Although they are not needed by the printer, since the color separating software will add more. than enough, these latter ones will be trimmed-off.


5. Proofing your work

5.1 Black-and-white prints

Colors, other than black, will be dithered. Obtain a solid black print of each layer by firstly hiding all layers, then selecting the layer as a group and giving it a common stroke color of 100% black. This won't affect any fills in the layer, and may add an outline to some symbols, but you can easily allow for this in your checking - or you can avoid these problems to a large extract by grouping line symbols within each layer. Be careful no to save the changed

5.2 Color Prints
You should get a composite color print for final proofing. The best result for checking is obtained by assigning the process colors cyan, magenta and yellow to the blue, brown and yellow layers respectively in the copy of your file that you send for printing (leave green as PMS 361). Don't forget to include any placed image files on your disk, and remind the bureau to select the `Preview and print patterns' option. It is strongly recommended that you get another color print done if you find it necessary to make any significant changes.

5.3 Limitations
Beware that Overpinting cannot be previewed on printers or monitors - it must be checked on the separations. Also, some fine patterns (eg difficult running/good visibility) and small dots (eg rocky ground and broken ground dots) cannot be shown truly at 300 dpi. `Banding' will occur in the former, and the latter will be larger (and hence closer together and more distinctive) than on the separations.

It seems to me that some of the specified symbol sizes are too small. For example, when not used in a screen effect, I think the minimum size brown or black dot should be the maximum size plus (at least?) 20%! Also, the untagged cliff should be +20% in thickness and minimum lengtth, and the minimum size scalene triangles on the front of West Fantasia are indistinguisable from the black dots at 1:15000. Comment?


6. Imagesetting

6.1 Choosing a bureau

Make sure the bureau you choose has the equipment to do a good job by asking some pertinent questions. For example:
o Is their computer set-up sufficiently powerful? Eg a 68030 Mac with at least 4MB of RAM?
o How new is the imagesetter and what RIP (raster image processor) are they using? There have been dramatic advances in RIP speed over the last year or so. Anything older may not be able to handle the job in a reasonable amount of time.
o Do they often provide film output and do they use o denshometer to check emulsion quality regularly? Some bureaus only offer bromide output.
o What is the maximum page size their machine can handle? With many irnagesetters you are restricted to a maximum width of 30 cm: others can handle a maximum width of 45 cm. In some cases there is still room for the registration information, etc that is added by the separation software; in other cases, this must be allowed for within the maximum width. Obviously, this is information you need to take into account before undertaking any drawing!
o How do they charge for graphics work? Will you have to pay a fixed charge per separation (yes please/), or is there an additional charge per time unit? A time loading can make the process very costly if there are imagesetting problems.

6.2 Separation settings
A resolution of around 1200 lpi is good enough (and will be about half the price of 2400 lpi). Choose the default screen ruler setting for the resolution you select - this will be in the range 120-150 dpi. (Is this right?) Positives are much easier to proof than negatives and should be acceptable to your printer (but check first). For film positives and negatives, specify Emulsion Down (right-reading) or the equivalent, Emulsion Up (wrong-reading).


7. Printing

7.1 Registration

The printer will be pleased to have computer-generated positives or negatives as they should register exactly. However, the general absence of overprinting makes it even more crucial than usual that the printer achieve exact registration and this should be explained beforehand. Don't forget to leave instructions as to how the map is to be trimmed, and check any crop-marks you added when drawing to make sure they were not alwed when the separations were done.