3p de Heij Nagelopen Door Hans - DNB
3p de Heij Nagelopen Door Hans - DNB
3p de Heij Nagelopen Door Hans - DNB
Email: Hans.de.Heij@dnb.nl
Page 1/84
1. INTRODUCTION
This paper will focus on the selection of public security features for banknotes. Next to the
reproducibility of a new security feature other parameters should be studied too to make a new
security feature successful, like e.g. the time needed to verify the feature, the required human
operation (feel, look and tilt), the location of the feature on the note and the consistency within
the series.
Another purpose of this paper is to give a historical overview of the policy of De Nederlandsche
Bank (DNB) regarding the introduction of new security features.
DNBs tasks regarding counterfeit deterrence are:
- producing banknotes with adequate security features,
- keeping the retail sector and public at large properly informed,
- supporting repressive measures on the part of judicial, judiciary and police authorities against
counterfeiters.
DNB prefers to present these three aspects as a triangle (see Figure 1 [99]). The focus in this
paper will be on secure banknotes; law enforcement, communications and awareness are only
marginally touched.
Central bank
secure banknotes + communication
Cooperation
Figure 1.
For counterfeit deterrence to be effective, maximum cooperation is required. Close collaboration is the best
guarantee that the central bank provides secure banknotes and adequate communication, making retailers
and consumers aware of the security features and that counterfeiters are traced by law enforcement.
Page 2/84
Self defending
89 %
<1%
10 %
Law enforcement
75 %
< 0.5 %
25 %
Figure 2.
Two basic strategies to protect banknotes against counterfeiting. Central banks following the strategy of
self defending banknotes spend relatively more money on the security of the banknotes. Central banks
opting for the strategy of law enforcement will spend relatively more money on law enforcement.
Imaginative figures.
Communication
The Swiss National Bank was the first to issue a public leaflet (1976). Before then this was
considered unwanted, since informing the public is like telling the counterfeiter, so was the
general opinion of central banks. DNB followed the Swiss example in 1983 with a first
experimental leaflet for the NLG 50/Sunflower. Today every central bank makes an effort to
familiarise the public with the (public) security features. Communication costs are relatively low.
Easy to explain, hard to counterfeit
Security features should be easy to authenticate and difficult to counterfeit. This is easier said
than done! Indeed: If you can make it, they can fake it!, the motto of Mr. Martin Mund of the
European Central Bank (ECB). In a Monthly Bulletin phrased as: The main challenge in
developing a new series of banknotes is ensuring that, on the one hand, the new banknotes are
innovative and difficult to counterfeit, and that, on the other, they are easy to check and have
security features that can be easily communicated. [62]. For the US dollar a similar policy is
drawn: future US dollar features should be complex yet easy to explain to the users of currency
[77].
Page 3/84
Features for new banknotes should not only be selected on the basis of perception, but obviously
also on the basis of other studies, like
- communication strategy (e.g. motto like feel, look, tilt, retrieval path),
- capitalizing on public awareness of legacy features (e.g. no more than two new public features
and improving the features that are kept),
- costs (e.g. maximum cost increase 3 %),
- resistance to counterfeiting.
In 2009, DNB conducted a survey of the Dutch publics security feature preferences. It turned out
that the location of security features on euro banknotes has the highest relative importance,
followed by the number of security features. The appearance of the security features is considered
less important. Security features that are verifiable at one glance will give the strongest boost to a
new euro 50 banknote design (see Paragraph 8.5.2)
Page 4/84
made in 1282 in Bologna in Italy, long before the first western banknotes appeared, which
amounts to a lifespan of well over 700 years. Although back then watermarks largely consisted of
lines, watermarks are still used in banknotes (Figure 3). And if we consider that copper plate
engraving was first used for British pound banknotes in 1694, we are looking at a tradition
spanning over 300 years! [24]. The origin of micro-printing is not exactly known, but Dutch
guilder notes issued in 1860 are the first available example of the micro-printing in offset used in
banknotes. Still in use as a security feature in many banknotes including the euro, the application
of micro-printing is a tradition spanning 150 years!
(a)
(b)
(c)
Figure 3.
a) sketch of the first watermark made in Bologna in Italy in 1282. The first NLG notes in 1814 included
such a line watermark.
b) mould-made watermark Little Owl by Jaap Drupsteen for DNB and produced by VHP. Fist watermark
with integrated highlights applied in NLG 100/Little Owl, issued in 1993. Since 1924, NLG notes featured
mould-made watermarks.
c) mould-made watermark Renaissance by Inge Madl for DNB and produced by Arjo Wiggens (VHP) in
2003. The background is an innovation, a so-called digitised area.
Page 5/84
problematic given that banknote dimensions have reduced over time, in some cases over 50 %!
[100].
Foil and silk screen were new features introduced in the 80s to deter colour copying and were
added to the existing ones like watermark, thread and gravure print. Today central banks
maximise the number of security features in their banknotes to around 20, divided over several
user groups [25, 48, 55], including around 6 security features for the public and 2 or 3 features for
the retailer. Contrary to the past, central banks can pick from a wide variety of security features.
The question is: how to make this choice?
Original
banknote
Banknote
reproductions
Above
threshold
Search for
new
feature(s)
New
banknote
design
Figure 4.
Process traditionally preceding the production of a new banknote model in a particular denomination in
reaction to a high incidence of counterfeits of that denomination.
Page 6/84
(a)
(b)
Figure 5.
Communication examples on the use of high banknote denominations in the Netherlands.
a) Sorry, no change (gasoline station, 2007) .
b) These banknotes are not accepted (liquor store, 2009).
c) This ATM issues also EUR 100 notes (ABN AMRO branch, 2009).
Pictures by author.
(c)
Page 7/84
Page 8/84
happened in Canada after the issuance of the new CAD 10 in 1999. A similar reaction by
counterfeiters was recorded in the United Kingdom. Shortly after the introduction of the new
GBP 20 in 2007, counterfeit figures surged to an all-time high in 2008. In the Netherlands, too,
counterfeit figures increased shortly after the issuance of the NLG 25/Robin in 1989.
Page 9/84
Counterfeits detected
per million notes in
circulation [c/mln]
0 10
2.
3.
4.
5.
6.
10 - 50
50 - 100
100 - 200
200 - 500
> 500
Country/currency
Australia, Japan, Latvia, Norway, Romania, Sweden
(Netherlands before euro)
Hungary, United States (excluding 1 USD-notes)
Euro (average in euro area)
Canada, United States, Euro in NL
Canada (in 2004), Great Britain (in 2008)
-
Table 1.
Seven levels of counterfeits and some of the countries where these levels have been established. Number of
counterfeits detected per million notes in circulation [c/mln].
Page 10/84
40,000
35,000
30,000
25,000
20,000
15,000
10,000
5,000
juni 2009
dec. 2008
juni 2008
dec. 2007
juni 2007
dec. 2006
juni 2006
dec. 2005
juni 2005
dec. 2004
juni 2004
dec. 2003
juni 2003
dec. 2002
Figure 6.
The incidence of euro counterfeits in the Netherlands, 12-month rolling average number of detected fake
banknotes in circulation. Figures concern counterfeits passed as registered by the National Analyse
Centre of DNB, 2002 - 2009.
Page 11/84
- numbers,
- in- or decreases in the number of fake banknotes,
- the places where the counterfeits were distributed and/or circulate
- the denominations that are counterfeited (the most),
- face value,
- the places where the counterfeits were produced and by whom or which criminal group.
Area/country
Year
Reference
1. Increase/decrease
2. Distribution: where
3. Circulating: where (mainly)
4. Note counterfeited the most
5. Percentage of most
counterfeited denomination
on total of counterfeits
6. Other counterfeited notes
7. Numbers passed
- absolute
- c/m
- per resident
8. Numbers seized
9. Face value
Euro area
2009
[95]
+ 17 % *
Euro area
Euro area
EUR 20
48.5 %
Netherlands
Before
euro
Euro
2009
2000
[96]
[26]
+ 31.5 % *
+ 30 %
Euro area
NL
NL
NL
EUR 50
NLG 100
82 %
~ 90 %
EUR 50,
100
EUR 20
NLG 25,
50
826,000*
64,056*
~ 140 ***
16,299
~ 10
Bulgaria
criminal
group
Bulgaria;
Utrecht
(NL)
USD 100
between 5
and 30
USD 103
million
(0,01 % of
circulation)
Colombia
Peru
United
Kingdom
2008
[91]
+ 235 %
UK
UK
GBP 20
98 %
686,000
298
GBP 13.7
million
Mafia
digital press
(61 %)
United
States
2008**
[62, 77]
e.g. 17
arrests
*) These figures are simply double the figure for the first half year of 2009.
**) Period October 2007 - August 2008
***) implicitly mentioned (reader has to divide the number of counterfeits by banknotes in circulation)
Table 2.
Typical subject-matter covered by press releases of 4 central banks.
Overview of counterfeit figures in eurozone and some other countries.
Page 12/84
Such reports have a statistical format including some legal remarks. Table 2 provides an overview
of a reporting style as observed by four central banks in this context. All too often these pressreleases fail to inform on technical matters like differences between the fake and original notes.
Increase in counterfeit detection by Dutch retailers since euro introduction
A comparison of the number of counterfeit NLG notes with that of euro counterfeits in the
Netherlands (see Table 2), should take the following into account. The figures for the Netherlands
were only based on the counterfeits detected there and did not include NLG counterfeits detected
in other countries. The figures on the euro counterfeits in the Netherlands also include - of course
- euro counterfeits coming in from other euro and non-euro countries.
Table 2 shows that Dutch retailers were since the introduction of the euro confronted with about 4
times more counterfeits in 2009. However, the total loss incurred by Dutch retailers because of
shoplifting is much higher than the total value of the counterfeits, i.e. around EUR 750 million
and EUR 2 million, respectively.
Probability of receiving a counterfeit in NL
The odds of winning a lottery are usually much higher than receiving a counterfeit in circulation.
In reaction to the first increase in euro counterfeits (2004), DNB calculated the probability of
receiving a counterfeit in the Netherlands. Since the circulation of euro notes in the Netherlands is
unknown, the total volume of passed counterfeits was expressed in absolute numbers.
Nevertheless, the following calculation can be made:
Assuming a total circulation of 13 billon euro banknotes (13.109) in 2008 and a share of 5.71 % of
this total for the Netherlands (capital key of 3.99 % rebased to 100), the Dutch account for 742.3
millions of euro notes in circulation. However, the real circulation in the Netherlands is, as said
before, estimated to be lower, i.e. at approximately 350 million notes. With around 50,000
counterfeits in 2008, the probability of receiving a euro counterfeit in the Netherlands will be assuming the issuance of counterfeits are a constant flow over the whole year - about 140
counterfeits on 1 million real euro notes. The real probability will be lower, since counterfeits are
not equally put in circulation. Furthermore, the calculation is only valid for the person being the
last to use the counterfeited note; it presupposes that only one payment is done with the
counterfeited banknote. Only the last person using a counterfeit will bear the damage of a lost
value.
The probability of receiving a US dollar counterfeit note is believed to be lower, around 5 on 1
million, including one-dollar bills [62].
Perception of the number of counterfeits
At the height of the first media attention for euro counterfeits in the Netherlands, in February
2004, people estimated the number of counterfeits to be much higher than the actual number of
counterfeits circulating. Entrepreneurs estimated that - within a 5-year period - 0.5 to 10 % of the
euro notes in the Netherlands would be counterfeits. Consumers were expecting even much higher
levels: 30 -100 % [48]. Such high figures were also reported in 2004 in research done by the ECB;
49 % of the retailers believed that they have come across a fake euro banknote [34].
In a recent survey by DNB on the safety of payment instruments, 11 % of the Dutch respondents
reported having received a counterfeit banknote or coin at one time or another [93]. This public
perception is remarkable, since it is far above the reality of around 50 c/mln passed or 0.005 % in
2008. Based on these figures the Dutch estimated over 200 % more counterfeits in circulation as
there were in reality (in 2008).
Page 13/84
Confidence in banknotes
In addition to the cognitive average knowledge of public security features, DNB considers the
Dutch citizens confidence in euro banknotes a relevant psychological indicator and started in
2004 to measure the publics trust in banknotes. Figure 7 provides the confidence figures for NL
and Canada.
The publics confidence that euro banknotes in circulation are authentic is not high: around 7 on a
scale from 1 to 10 [48, 55, 85]. The Bank of England has also measured aspects of public
confidence since 2004, but did not publish the results.
In 2004 and 2007 the ECB made a so called cash handlers survey [34, 61]. From these two
surveys it can be concluded that in both years about 50 % of the European cash handlers is of the
opinion that the euro banknotes are sufficiently secure against counterfeiting. For the Netherlands
the opinion is also evenly split. It is most probably on the basis of these surveys that the ECB
reported that Studies show that, in general, the more a banknote is immediately liked, the more it
inspires confidence [66].
10
8
6
4
2
0
2004
2006
6.8
Euro NL
Canada
2005
5.57
5.31
2007
2008
7.1
5.4
5.43
2009
2010
?7.3
7.2
5.55
?
5.6
2011
?
5.6
?
5.64
Figure 7.
Public confidence score for the euro (in NL) and for the Canadian dollar.
Threat indicators
The first counterfeit threat in Europe was already recognised in 1849 by the Acadmie
Internationale des Sciences. Banknotes could be reproduced using palaeography. Palaeography
employs both lithographic and photomechanical techniques. Such techniques became available
with the invention of the photography in 1825 by Nipce and by Daguerre in 1837. Because of
this counterfeit warning, in the 1850s the black print on the Belgium banknotes was replaced by a
more secure blue tint [20].
Instead of probability of a counterfeit or confidence, central banks have recently started
looking for ways to create threat indicators, just as in 1849. Such indicators intend to create a
signal, e.g. on a scale of 1 to 5, that indicates what kind of response is necessary. This signal
could be based on a variety of aspects like: increase in counterfeits since previous quarter,
counterfeit percentage in total banknote circulation, deceptiveness, average counterfeit value or
financial damage.
Recently artificial models have been developed to explain these threats, like threat assessment
(Bank of Canada, 2008), analytical flow model (USA, 2007) and multi criteria decision analysis
(ECB, 2007). They will be discussed in Chapter 6.
Page 14/84
Feb 2002
2,002
70
61
31
5
5
7
7
3
7
11
1
5
3
2
2
11
2.3
Feb 2003
2,015
65
52
13
3
3
5
5
4
8
16
2
9
3
2
1*
3
18
2.0
Feb 2005
1,501
68
49
12
3
4
5
9
4
10
23
5
12
3
3
3*
5
15
2.2
Feb 2007
1,506
65
43
14
4
4
5*
5
4
7
18
4
7
3
2
2
3
19
1.9
Feb 2009
1,058
76
55
15
2
3
9
8
6
14
27
3
16
5
3
0
5
7
2.5
Page 15/84
Instead, security features are getting more and more complex to evaluate. The public already has
major difficulties recalling a feature. The general public is unaware of security features, as is
illustrated once more by DNB's 2009 study of public knowledge of security features (see Table
3).
The group that can not tell any security feature dropped from 19 % to 7 %. Spanish researched
showed that almost 1 in 5 Spanish cannot name one security feature either [73].
Public: some qualitative, but no quantitative knowledge
The ECB also found that the public at large has little knowledge of euro banknotes' security
features. Their qualitative surveys suggest that cash users usually do not pay much attention to the
security features of the euro banknotes and are familiar with only a limited number of these
features. Furthermore, just a tiny fraction of the respondents was able to name or describe their
functionalities. Knowledge is, in most cases, limited to the traditional security features, i.e. the
tactile properties; the watermark and the security thread [61].
So, the public knows some qualitative aspects of banknote security features, but hardly any
quantitative. The relative high confidence in the euro banknotes (Figure 6) is probably one of the
reasons why the public does not check banknotes on counterfeits. An other explanation is given
by Ruud Van Renesse: The probability of a member of the public receiving a counterfeit is
virtually negligible. So why would the public inspect banknotes at all? It is hardly worth their
while. [57].
How does the public know what to check?
Many people will recognise the following situation: Look at me, what do you see?, asks the
wife of the husband. A new dress?, the husband answers hesitantly. No, look at my hair! I have
been to the hairdresser! This is a fine illustration of what kind of images our brain stores: 95 %
of what we see is recollection of old images [67]. For banknotes, this means that people will
probably just look for the watermark and foil, the two best-recalled public security features. This
implicit long-term memory seems to be irrelevant for a full security check, since it does not tell us
how the feature should be checked.
Another example of banknote recollection over a period of 7 years was researched in 2009 by
TNS NIPO at the request of DNB (Table 4).
Page 16/84
The Dutch public was asked which images of former guilder banknotes they could remember.
The Snipe, as the then circulating NLG 100 bill was popularly referred to, was mentioned by far
the most, followed by the Lighthouse and Sunflower.
Only a few people recalled historical persons like Michiel de Ruyter (2 %) and Frans Hals (2 %).
Men proved better able to recall pictures on guilder banknotes than women, mentioning the
lighthouse and Michiel de Ruyter the most. Respondents between 35 and 54 years old could
mention more pictures of guilder banknotes than those in the other age groups: the snipe,
lighthouse and sunflower. The denominations of 25 and 10 guilders were best recalled (33 % and
26 %) [86].
Banknote perception
Besides the overlooked new haircut, other examples were given in the second public feedback
paper [55]. There is a difference between removing an element (a birthmark or a moustache) and
adding one (start wearing glasses). This might also apply to the perception of banknotes. Leaving
out the watermark or a secure ink may go unnoticed, while something added like a stamp or
writing on a banknote will definitely not be overlooked. This explains why counterfeiters can
permit themselves to leave out some security features as we will see in paragraph 8.3 DNB Model
3: Simple method.
There are some other famous recollection examples that might be useful for banknote design, like
e.g. the interior of a room. When people are offered an image of that room, they are very well
able to reproduce an inventory list of all the furniture and other objects in the room. However,
when the same furniture and objects are offered as single images, one by one in a random order,
people are less able to recall all objects shown. This speaks in favour of a little story on the note,
as we will see in paragraph 8.5.4 Communication of security features.
Does the counterfeiter make use of the publics knowledge?
Why do counterfeiters focus on certain features and widely ignore others? Knowing that the
watermark is the best-known public feature, would counterfeiters always incorporate fake
watermarks? Or, in more general terms, do counterfeiters use the publics knowledge as reflected
by Table 3 for optimising their bogus notes?
To answer this question we need statistics on how often security features are reproduced. Such
statistics are only incidentally made. Table 5 provides on the one hand the public knowledge of
security features and on the other hand an estimation of the presence of public security features in
counterfeits.
Public knowledge
(NL, 2009)
76 %
55 %
15 %
9%
8%
3%
2%
Imitated in some
form in euro
counterfeit
Frequent
Almost all
Frequent
Frequent
Some
Frequent
Frequent
Table 5.
Relation between public knowledge of public security features and the presence of such features in
counterfeited euro banknotes (common class, frequency).
Page 17/84
Figure 8.
How to break the vicious circle? Make designs that get attention of the public!
2007
35 %
16 %
4%
3%
0%
45 %
100 %
2008
36 %
18 %
5%
3%
0%
41 %
100 %
Table 6.
Overview of the use of retail devices in the Netherlands (2007 and 2008) [55, 79].
Page 18/84
These results qualitatively match the data for the Netherlands as reported in the 2007 Cashier
Survey of the ECB except for the pen tests for starch content (see Table 7) [61]. Euro banknote
paper does not contain starch and therefore will not leave a dark brownish mark as most
commercially available paper does. The pen is not recommended for authentication of euro
banknotes because it is not always accurate and results can be manipulated using various
chemicals.
Retail device
Ultraviolet lamp
Euro pens indicating whether the
banknote paper is genuine
Infrared viewer
Authentication device/equipment
The till/equipment you use automatically
checks all banknotes
You dont use any tools (spontaneous)
Other tool (spontaneous - specify)
Dont know
NL
44 %
10 %
Euro Area
26 %
9%
9%
22 %
1%
7%
3%
2%
18 %
4%
2%
55 %
1%
1%
Table 7.
Results on the question: Do you use any tools to help you check the authenticity of euro banknotes? If so,
which ones? (multiple answers possible). Cashier Survey 2007, ECB.
A similar figure (43 %) was reported by the Banco de Espaa in 2008 for the presence of
banknote authenticity devices at Spanish shops. They reported also that 47% of their customers
are not annoyed by verification with authenticity devices of the banknotes they proffer to pay for
a purchase; but on the other hand 41% are annoyed by it [73].
Automatic devices are preferred
There are two types of device for retailers:
1) The type of device letting the retailer decide whether the note is genuine or not, e.g. UV lamps,
IR viewers, magnifiers and mirrors (for the colour-changing features when the note is tilted).
2) The type of device that indicates whether a note is genuine or not, often by a green/red light, a
text display or sound (a beep).
The second option is preferred and advised by DNB. Research conducted by DNB in 2006
showed that UV light checks on banknotes are often not properly conducted. This explains why
so many counterfeits with a limited UV imitation are not rejected. That is why DNB discourages
the use UV lamps (see paragraph 8.5.1). Auto detection devices on the other hand provide the
retailer with an easy argument to ask: The detector does not accept this note, would you have
another one for me?
Develop features to assist retailers
It is known that retailers do not wish to spend more than 2 seconds on a security feature [48].
Central banks could be of assistance here to retailers, for example, by:
- ensuring that in the next generation of banknotes all retail features are on one side,
- developing tools, e.g. a barcode scanner, that also check a banknote for genuineness (Figure 9).
- developing a marker and/or taggant readers (see Paragraph 7.1).
Page 19/84
Figure 9.
Cashier feature with three functions: security check, denomination recognition and cash management
developed by DNB in 2001.
Left: a barcode on the banknote is readable with the laser light scanners used at supermarkets (660 nm).
Right: metameric barcode NoCopyCode, developed by Joh. Ensched. A barcode with camouflage
based on metameric inks. Any code could be used, like e.g. EAN-13, EAN- 8, or any other. Metameric inks
are explained in paragraph 8.2.
Page 20/84
Figure 10.
Banknotes are often used without being seriously checked. The flash second is important. Drawing by
author (1984).
In this instant, the value must be verified. The receiver, usually a retailer, will need to assess if
she or he is getting the right value, while the payer must determine whether she took the right
denomination from her wallet. This explains also the research results of both the US Treasury and
the ECB that the touch of the banknote is the most important trigger on counterfeits. In 2002 the
US Treasury reported that 25 % of the cash handlers only check the just received banknote if it
feels suspicious (against 6 % of the general public, the consumers) [29]. The 2007 Cash handling
survey of the ECB reported that the most common security feature checked is tactility for 70 % of
the cashiers [61].
Maximum time spent on checking security features: 5 seconds
The authors paper entitled Public feedback for better banknote design provides data regarding
the time needed to check a security feature [48]. From DNB research, the total time to settle a
cash payment transaction is known to be 19 seconds [36]. Unfortunately, it was not documented
what exactly is included within this time. Therefore, we have to live with the breakdown of 19
seconds as assumed in Table 8 until more studies on time become available. From this
breakdown, it may be deduced that the total time to check a banknote is probably around 5
seconds. This is also consistent with the study initiated by the ECB in 2007 to analyse how people
handle banknotes. The total handling time was 5 seconds on average, 3.5 seconds of which was
spent on exploring the front and the remaining 1.5 seconds on checking the reverse. After 10
seconds the test terminated [52].
Five seconds also seems to be the limit of what is socially acceptable to check a received
banknote without being impolite.
Page 21/84
Estimated
time
1s
2s
2s
1s
1s
5s
1s
2s
1s
1s
0s
2s
19 seconds
Table 8.
Breakdown in time of a cash settlement between a payer (public) and a receiver (retailer). Estimation by
author, based on the reported total time to settle a cash transaction.
Estimated
time
2s
2s
2s
1s
2s
3s
4s
16 seconds
Table 9.
Estimated time to check the public security features in a EUR 50 banknote.
Page 22/84
check is often hindered by the thumb or other fingers covering the feature when the note is being
turned. It is therefore interesting to follow the response to the recently introduced motion thread
on the reverse of the new 50 Danish krone banknote issued in August 2009. Will it be successful?
Other ECB research done in 2007 reported that the majority of the cash handlers check banknotes
on both sides (64 %) with just over a quarter only the front (27 %) [61].
No nested features
There is a clear trend that central banks expect the public to pay more attention, in casu more
time, to security features. The reason is the trend of so called nested features. Nested features
are based on the principle of a feature within a feature, reminding of the well-known Dutch
example of the Droste Cacao box (Figure 11). The first image, a nurse holding a tray with a cup
of cacao and a Droste Cacao pack on it, is interminably repeated in that Droste box.
Droste-effect
Figure 11.
Principle of nested features. On the left: Droste effect. In the middle the original basic feature. On the
right the basic feature with 3 nested features in it. One nested feature contains two more levels of nested
features.
Figure 12.
A switch from dark to light within the renaissance window is seen when the hologram on the euro 50
banknote is tilted.
A hologram is such a nested feature: a plain foil (main feature) is provided with a hologram
(nested feature 1), mini text (nested feature 2), micro text (3) and, from the reverse, a perforation
(4). The hologram itself shows the image of a gate switching from positive (feature 2.1) to
Page 23/84
negative (feature 2.2) on being moved (see Figure 12). Also the numeral 50 may be seen
(feature 2.3).
Another example of a nested feature is a mark in glossy ink (main feature) showing a colour-shift
(nested feature 1) and movement (nested feature 2). Micro-lettering on a security thread is also
seen as a nested feature, just as a feature which becomes discernable when held under UV-light
with a long (around 365 nm) or a short (around 254 nm) wavelength.
Change blindness
The switch from dark to light within the Renaissance window of the hologram on the euro 50
(Figure 12) reminds of a change blindness test done with 20 US dollar banknotes. The two
images of Figure 13 are used in an animation movie showing each 1.5 second a switch from one
image to the other. Most people do not notice that the background behind the portrait changes
from dark to light. Tilting the euro 50 hologram will also be hindered by change blindness. So
called achromatic holograms switch from light to dark and will experience similar perception
problems.
Figure 14 is also a variant on change blindness. Mirroring the portrait of the well known one
dollar banknote is not noticed by most people [80].
Figure 13.
Change blindness illustrated by USdollar 20 banknote. When the above images are shown on a screen most
people will not notice the 2 second switch in the background to Jackson from dark to light.
Figure 14.
Which is the correct USD banknote? [80].
Page 24/84
between these features. This principle is explained in Figure 15 and the DNB-patent Authenticity
Mark [46]. As reading eyes typically focus on sizes of about 30 mm x 15 mm, these dimensions
might serve as guidelines for the dimensions of public security features [67].
Figure 15.
Left: banknote with six individual public security features; stand-alone features without integration or
overlap (island features).
Right: banknote with six integrated public security features; partly overlapping each other. Integration is
further enhanced by overlap of non-secure areas, like e.g. overprinting with offset.
Thread, foil stripe or iridescent band are usually simply overprinted in the banknote design and
are not really integrated. Such features could be more effective by making them leading in the
design process. Also features that are not linked, not physically connected to others could be
improved like the colour changing element (OVI) and the watermark. Such island features
should be avoided since they make life easier for the counterfeiter.
Multi-level security features
One should be careful with features serving several user groups. A security thread for the public
that may also be used as a detector feature in the central banks sorting machines is an example of
a multi-level feature. Such feature might lead to sub optimization for one or more of the
relevant user groups. Ideally each feature should serve just one user group.
User requirements
Time seems to be the most important user requirement for a public security feature, but there are
more, as listed in Public feedback for better banknote design papers [48, 55]. Table 10 provides
an overview of these user requirements, including the new introduced no nested features
requirement. As an example the table is filed in for the euro 50 note.
Scoring of public features in the EUR 50 note is simply done by a Yes or No.
The features best fulfilling the user requirements are:
- nail scratch feature (or ISARD), 7 out of 7
- security thread, 5 out of 7
- see-through register, 5 out of 7
The features that do not match the user requirements are:
- hologram, 1 out of 7
- OVI, 2 out of 7.
Page 25/84
USER
REQUIREMENTS
1. Fast (< 3 s per f.)
2. Easy to use
3. Discrete
4. Reliable (durable)
5. Striking
6. Univocal
7. No nested feature
Total
FEEL
1.1
1.2
Tactile
Nail
relief
scratch
Y
N
Y
N
N
N
Y
3/7 Y
Y
Y
Y
Y
Y
Y
Y
7/7 Y
2.
Watermark
Look
through
Y
Y
N
Y
N
Y
Y
5/7 Y
LOOK
3.
Security
thread
Look
through
Y
Y
N
Y
N
Y
Y
5/7 Y
TILT
4.
Seethrough
Look
through
Y
Y
N
Y
N
Y
Y
5/7 Y
5.
OVI
Total
6.
Hologram
Reverse
N
Y
N
Y
N
N
N
2/7 Y
N
N
N
N
Y
N
N
1/7 Y
5/7 Y
5/7 Y
2/7 Y
5/7 Y
2/7 Y
4/7 Y
5/7Y
Table 10.
Scoring the public security features in the EUR 50 note against the user requirements.
Y = Yes, meeting the requirement. N = No, not meeting the requirement.
3s
3s
3s
---- +
9s
This leads to serious doubt if such features will be used by the public. Most likely, the public may
use it partly, e.g. will view it from the front in transmission.
OVI and SPARK
More and more central banks are growing convinced that Optically Variable Ink (OVI) is not a
strong feature. After DNB and Romania, similar findings are reported by Banco de Espaa. Just
as in the Netherlands, the optically variable ink mark is the least familiar public security feature,
known by 11 % of the Spanish (and 3 % of the Dutch). The OVI on the dollar notes is also not
popular in the United States. The NRC in their NextGen book writes that color-shifting inks are
rarely used by the general public[62]. Dr. Hans Reckers (Bundesbank) is one of the criticasters
Page 26/84
of OVI because of its astonishingly easy reproducibility [78]. Vladimir Finogenov of the
Russian Central Bank agrees: OVI are reproduced with rather high quality [76].
With Spark, the next generation of OVI features, a third, motion-based, nest level was
introduced, in addition to the gloss and a colour (Figure 16). The order of checking the Spark
feature will be:
- motion,
- colour switch,
- gloss.
Figure 16.
The first banknote to feature the Spark is the Chinese 10 Yuan (print run 6 million), issued by the Peoples
Bank of China in 2008 on the occasion of the Olympic Games in Beijing in 2008. Picture published by
Getty Images.
The question is whether it will be successful. Motion is registered more quickly and in a different
part of the brain than colour differences. People will focus on motion rather than colour change
and/or gloss. If it sees motion in a counterfeited Spark-feature, the public might take it for real.
Since the OVI patent has expired, another draw back is the free production by any company of
two elements of the Spark: the gloss and colour switch.
And there are other hurdles for a Spark feature to be successful. The limited range of colours that
are suitable for the purpose is another weak aspect, also from counterfeiters point of view. Once
the good colour flop imitation is found, it may be used for many banknotes, both within the series
and internationally.
banknote colour
turns to gold
Figure 17.
Colour concept for a banknotes series. Easy to communicate: all colours turn to gold.
Page 27/84
From the customers point of view central banks will opt for a colour-scheme as provided in
Figure 17, where seven different colours all change to the same hue of gold. Such a proposal is in
line with the advice given 2007 to develop security features much more on colour concepts, like
e.g. partly coloured foil stripes [55]. The design of the colour-moving features should be extra
appealing, since the public in general is not keen on having to tilt banknotes in order to check
features [48]. Finally, the Spark feature is expensive, making up 15 to 20 % of a banknotes
production costs.
Page 28/84
Usually the printer is the contractor for a new banknote. The printers customer is the central
bank, not the public or the retailer. Semi-finished product suppliers are even further removed
from the final users of the banknotes (see Figure 18). This situation may be explained by the
controversy between technology push (the suppliers) and technology pull (the central banks).
FOIL
PAPER
PRINTER
CENTRAL
BANK
INK
Suppliers
(semi finished)
RETAILER
PUBLIC
Figure 18.
Supplier (too) far from final customers. The central bank should take care of their stakeholders and inform
the private sector by means of a Programme of Requirements.
One of the central banks commenting on this issue is the Bank of Canada. The approach of the
private sector, because it is largely driven by profit motives and short term timelines, is not
completely suitable for the Bank of Canada. () Working through an internal research and
development program allows the Bank of Canada to test concepts that fall outside the core
capabilities of the traditional suppliers. Such exploration allows the Bank to progress toward
features targeted to specifically meet the needs of the Canadian currency user. [82].
An even stronger quote comes from Julian Ashbourn: Suppliers are convincing central banks of
the merits of their particular features, spiced with some attractive sounding theoretical benefits,
couched in the language of the organisation being targeted. With good marketing and publicity a
steamroller effect is created and, with the help of conferences and workshops to reinforce the
message, consumers dutifully start adopting the sales-speak.[83].
The new CEO of DeLaRue is aware of, and even accepts, this criticism when he tells that the
security industry is suffering from complacency. I think there is a real concern that this is
becoming an introverted, narrow and incestuous industry which forgets the outside world be it the
consumer (in the end of the day the user of banknotes). [92].
The banknote industry cannot develop optimal banknotes for lack of input from the retailers and
public. Central banks should provide this input. The central bank is, in modern management
terms, the chain director and should do the consumer research. By means of a Programme of
Requirements the central bank may inform the private sector [65].
What should be developed?
Since there is only one feel feature available (ink relief) central banks are especially in need for
more feel features. Tilt features - and especially colour switching features - are less popular than
feel and look features as is concluded in the first paper on Public feedback for better banknote
Page 29/84
design [48]. Also user requirements for public features are defined in this paper, like a checking
time less than 3 seconds. Further more new retail features are needed as follow up for the
magnifying glass and the UV lamp.
abstraction level
Printer is contractor
It seems that the central banks should develop themselves much more as knowledgeable on the
subject banknotes and should increase their development work, starting first of all to understand
their customers better, especially the retailers and the general public (Figure 19).
5. Science
4. Independent designer
3. Project team
2. Central bank
1. Suppliers, printer
development input
Figure 19.
Schematic view of the estimated input in banknote development by 5 disciplines (estimation is subjective,
made by author). Suppliers dominate banknote development process.
Example
Often developers of security features expect that public and retailers will recognise small design
deviations in a counterfeit. Having a good idea of what a document should look like is essential to
detecting irregularities so is the idea of many suppliers. This may be questioned. Even if a large
element like the main image on the euro notes is altered, as in Figure 20, this is not noticed by
more than 80 % of the people [86, 87 90, 100].
Figure 20.
Change blindness for euro banknotes. Which are the correct notes?
Page 30/84
Page 31/84
human, automatic and forensic. Human is subdivided into unassisted (e.g. general public)
and assisted (e.g. retailer). Automatic is subdivided into cash distribution (e.g. counting,
sorting) and retail (e.g. vending machines) [64].
National Research
Council USA
USD
2007
Analytical flow model
Currency symbol
Year
Methodology
Target
Classes of
counterfeiter
Criteria
Arithmetic
operations
Judgement
Public testing /input
Measurements
Complete note or
single features
Time needed by
counterfeiter
Costs made by
counterfeiter
Final output
Forecasting which
technologies are
favourable for new
security features
1. Primitive
2. Opportunist
3. Petty criminal
4. Professional criminal
5. State sponsored
5 criteria per counterfeit
class (excellent, good,
easy, somewhat, blank)
Not found useful
Bank of Canada
CAD
2008
Stage/Gate process
(Technology Scan and
Threat Analysis)
Selection of designs of
security features for
development by the
central bank
1. Primitive
2. Hobbyist
3. Professional
4. State sponsored
3 in criteria and 7 out
criteria
No
Single features
Dynamic Selection
Process (ranking table)
Central bank experts
Yes, some human
perception experiments
No
Single features
No
Single features
No
No
No
No
No
No
Development programme;
technological directions
where to search for new
features
Stage-gate process,
> 300 at start; still 8
projects running for new
features
Resilience Grade 1 - 6
(best - worst feature)
By Committee (NRC)
No
Table 11.
Overview of three models.
Page 32/84
Page 33/84
Page 34/84
Figure 21.
Conceptual banknote for the partially sighted: clear large numerals, alternating between positive and
negative against different geometric patterns. Secure tactile patterns are included at the short edges
providing a codification for the blind. Maximum attention for the 4 security features in the centre (but not
on the folding line). One security feature has a secure purple colour. Background could be used for other
security features. Design by author (2009) [55].
Page 35/84
Page 36/84
paper) and intaglio lines (extrinsic, print). In addition, it was proposed to keep number reading
(extrinsic, print) [3].
Finally the new sorting machines, introduced in 1981, were to have three security feature
detectors: AQUS (barcode watermark), ISARD (intaglio line pattern) and an OCR-B number
reader (see Figure 22). The ink of the banknote number could be either magnetic or nonmagnetic.
ISARD
AQUS
Figure 22.
Left: first banknote with ISARD, the NLG 10/Frans Hals issued in 1971.
Right: NLG 25/Robin banknote with barcode watermark AQUS, issued in 1990.
Page 37/84
free
sinusoidal barcode
Figure 23.
Left: conceptual proposal for a so called free intaglio detector. Instead of the fixed intaglio line element
shown in Figure 22 any intaglio print area may be used for detection. All relief in this area will be added
up and the sum total has to be above a given threshold value. The sum total of all relief on a counterfeit
will stay below this threshold. The data collection area area should also be freely adjustable. The first
concept was proposed in 1998 by Koeze (DNB) and elaborated further by De Heij in 2003 for a discussion
with De La Rue Currency [31].
Right: The initial idea is to transform the sharp-edged bars into sinusoidal waves. Wavelength variation of
the bars exploits the characteristics of the mould-made paper machine to the maximum! Optical
transmission detection in IR range. Sample produced by Arjo Wiggens (2004). Based on DNB EU patent
on BCWM (2003) [30].
Page 38/84
Security features
Banknote
production process
Internal
1. Paper
2. Foil (hot
stamping)
3. Thin-foil (cold
transfer technique)
4. Silk screen/
rotogravure
5. Offset
6. Intaglio
7. Numbering
8. Perforations
9. Cutting
number + database
(micro)perforation patterns through
finished banknote
shape of edges, notches
Add-on
(delivered as semi finished product)
security thread, micro chips (RFID
tags), UV fibres, other types of fibres
like thin steel, luminescent pigments,
other pigments, markers/taggants like
e.g. botanical DNA
plain foil with or without hologram
Table 12.
Overview of several internal and add-on features that are or could be used in banknotes (the list is far from
exhaustive).
*) Recently first prototypes were printed of flexible displays. Individual pixels are printed by using
inorganic and organic Light Emitting Diodes (LED). The organic LEDs (OLED) are printed using wafer
techniques.
Page 39/84
Taggants
A relatively new development are the so called taggants, specific compounds added to the
banknote paper or ink and therefore classified as add-on features. Taggants are a subdivision of
banknote markers like numbering or magnetic codes. When engineered at a molecular level, these
taggants can provide a unique signature when probed with a suitable reader. Taggants are
typically manufactured using complex rare earth-phosphor compounds that are hard to source.
Their production may also be based on different technologies, e.g. optical, nano or DNA. Since
2004 customized genetic codes can be produced by extracting DNA from an infinite selection of
plants (botanic DNA). The typical structure of cotton makes every banknote unique and could
also be seen as a marker.
Taggants are known security features, but their application in security products has remained
limited. They could be introduced in banknotes as a retail feature, since today they can only be
read by specialised devices operating at slow speed or at standstill. For the same reason, such
features could be used as a counterfeit deterrent system feature or as a forensic feature. Those
readable by high-speed detectors could suit the sorting machines of central banks. Each
denomination could have a code of its own or even every banknote could receive a unique code.
Although classified as add-on features, taggants have also internal characteristics since they have
to be dissolved in paper or ink.
Original
banknote
Counterfeited
banknote
Graphic industry
- primitive counterfeiter
- casual counterfeiter
- petty criminal
- professional criminal
Figure 24.
Schematic presentation of the producers of original and counterfeit banknotes.
Page 40/84
Outside DNB the model was first introduced at the 1991 meeting of the Paper Committee of the
Banknote Printers Conference. One of the conclusions of this paper is that System analysis by
physical and chemical dimensions of reproduction systems leads to a clear development strategy
for new security features. [19]. Over the years the system was developed further [22, 25, 43, 87]
and today the model may be described as shown in Figure 25. The six key dimensions are
specified in more detail in Table 13.
Reproduction system
Original
banknote
1) resolution
2) colour a
3) optical density
Counterfeited
banknote
4) geometry
5) mass
6) material b
Dimension
Units
1. Resolution
2. Colour
3. Opacity
4. Geometry
5. Mass
6. Material
Table 13.
Overview of the six dimensions used in the system approach and the corresponding units (overview of units
is not exhaustive).
Page 41/84
Colour
Density
Resolution
Geometry
Mass
Material
Watermark
o
+
++
+
- 1)
o
Foil/
hologram
o
o
+
+
N/A
o
Security
thread
++
+
o
N/A
+
See-through
register
o
o
o
+
N/A
o
Colour:
iridescence/
OVI
+
o
o
o
N/A
o
Tactility
o
o
+
+ 2)
N/A
o
Table 14.
Strength-weakness analysis of the 6 public features of the euro series 2002 according to the black box
model or system approach. Original compared to counterfeit:
+ + = original much better than counterfeit, + = original better, o = original and counterfeit quite similar,
- = counterfeit slightly better, - - = counterfeit is better than original. N/A = not applicable.
1) because paper weight is similar to standard in copy machines, 2) for nail scratch element.
Gloss
original hologram
512
Reproduction
Sample A
Sample B
Sample C
Gloss
reproduced hologram
283
162
709
Table 15.
The gloss of the original hologram is compared to three different hot-stamping imitations A, B and C using
different foils and fixed on mat adhesive tape. The gloss of the reproduced hologram is measured in gloss
units [41, 43].
Page 42/84
Original
banknote
Reproduction system
Counterfeited
banknote
Although current off-the-shelf dot matrix systems have a high resolution, this does not mean that
an imitated hologram will be that sharp. The elastomer copy process will reduce the resolution to
just 6 lp/mm.
Keep track of changing key specifications
Now what a central bank has to do is to keep track of improved and/or new technologies. This is
an ongoing process. The last three decades have delivered overwhelming innovations, as phrased
by security product designer Joost van Roon: Scanning, imaging and printing have rapidly
evolved. Techniques that were beyond anyones imagination thirty years ago are commonplace
today. [89]. Personal computers became both cheaper and more powerful in the 1990s. Very
affordable image-editing software, desktop scanners and printers became available and delivered
good quality. Today you may buy a 4,800 dpi ink jet printer for just 60 euro!
Since the introduction of home scanners and all in one devices no new reproduction
technologies appears to have emerged. With no obvious new print technology platform in the
offing, innovation lies in the improvement of features. A case in point is the introduction of
digital engraving around 2000, which ushered in a new phase in a long gravure tradition (first
banknote gravure: 1732).
It seems that for the first time, the development of new features, such as a transparent window in
a cotton banknote, is not specifically aimed at outperforming newly arriving reproduction
methods. Table 16 provides a short and incomplete historic overview of counterfeit threats and
the reaction of central banks and indicating the dimension on which the new protection may be
reduced, like resolution, geometry or colour.
Predicting quality of counterfeits
Whenever a new ink jet printer, new imaging software or a new digital press is introduced, the
key specifications of the black box will change. An example is the particle size of the pigments in
colour copy machines. The pigments of the first generation of colour copy machines (1980,
Canon CLC 1) were limiting the resolution, but did deliver some relief to the copies, quite similar
to real banknotes. The third generation of copy machines used much smaller pigments (1994,
Canon CLC 800) leading to higher print resolutions. The relief disappeared, to the relief of the
central banks!
Page 43/84
Year
1282
1661
1694
1694
Changing value of
real notes
Carving
Original
banknotes not
uniform
Photography
Offset printing
Colour copy
machines
1797
1809
1819/183
6
1839
ca. 1850
1855
1867
1876
ca. 1920
ca. 1925
1928
ca. 1960
ca. 1970
ca. 1980
ca. 1980
1988
1989
1989
1989
1990
1991
ca. 1994
Home scanners
ca.1990
1992
1995
ca. 2000
2009
2009
2010 ?
Dimension
Density
Density
Material
Geometry
Geometry
Geometry
Geometry
Geometry
Colour
Density
Material, density
Geometry
Geometry
Colour
Geometry
Colour
Material
Colour
Material, density
Material,density
Colour change
Density, colour
Geometry, density
Material
Geometry, density
Geometry
Resolution
Density
Density
Geometry
Material,
geometry, density
Geometry, density
Material
Material
Table 16.
Overview of several historical threats and the reaction of the central banks.
Page 44/84
With this system approach it is now even possible to predict the quality of counterfeits. When the
colour gamut of any reproduction system is increased, security features based on colour will lose
strength. The new banknote under development should receive better key specifications on
colour than the latest graphical tools can deliver.
Provides insight in dimensions of new features
Another advantage of the system approach is the quick insight it offers into the basic level of
defence of a new feature. The Wings security thread for example can be defined in terms of
geometry and density (Figure 27)
In reflection
In transmission
Figure 27.
The Wings security thread shows up quite differently in reflection and in transmission (Goznak, 2007).
Also, the real security of the Motion feature is not only the lenticular lenses (geometry), but
especially the high resolution of the electro-graphical printing underneath these lenses. A
transparent window in a paper banknote is mainly a matter of material and optical density (Figure
28). But the model also serves to expose the weak elements in an existing banknote, e.g. the seethrough and (intaglio) portraits, losing territory in, respectively, the geometry and resolution
dimension.
resolution 1)
colour 2)
geometry 3)
optical density 4)
mass 5)
material 6)
Micro-optics Motion in
promotional banknote,
2005.
Figure 28.
Characterising Motion and transparent window with the system approach.
Page 45/84
Original
banknote
Reproduction system
Unusable
output
banknote
x
casual/primitive
counterfeiter
Figure 29.
Basic principle of prevention of banknote reproduction: no output.
resolution
colour
geometry
density
mass
material
Figure 30.
Selection of security features for a new banknote could be based on the policy do not put all your eggs in
one basket.
8.2.1 Resolution
For a better understanding of the system approach, resolution is explained here in more detail. By
tradition, banknotes are printed with continuous lines instead of dots. Replication by scanners or
copiers may be recognised because it consists of dots specified in dots per inch (dpi), in screen
values or in pixels or in any other way. If we want to compare the resolutions applied in banknote
production and in the graphic industry, both must be expressed in the same units. For banknotes,
line pairs per millimetre [lp/mm] are preferred.
Page 46/84
The finest elements a banknote printer may print are lines of 30 m wide. If the area between two
printed lines is also 30 m, the line frequency of these lines may be calculated as 16.7 lp/mm. A
resolution of 360 dpi is equivalent to 6.6 lp/mm.
Eigenfrequency
Over the decades several security features were developed based on printed lines, such as lines in
alternating colours, as will be explained in Section 8.2.2 on dry offset printing.
A review of all kind of security features that can be printed by lines and also dots is provided by
Ruud van Renesse in 2002 [28]. One of the classes defined is local screen modulation,
subdivided in screen- and scan traps. Screen traps are dedicated line patterns that interfere with
the screens used to reproduce a banknote with moir fringes as a result. Scan traps are defined as
printed patterns that form aliasing effects when the feature is scanned with similar frequencies as
for example the frequency used in the scan trap (eigenfrequency).
Screen traps using line patterns were first applied in the NLG 10/Frans Hals, issued in 1971. Later
Dr. Peter Koeze found that for being effective, the line width of the printed and the unprinted line
should be exactly the same [4]. The frequencies selected for the screen traps were similar to the
frequencies used in the reproduction systems used by the counterfeiter, e.g. screen 45 or screen 60
and are therefore also scan traps. Both, screen- and scan traps, are security features which trigger
the publics attention. Today such features are not considered to be public features, but trigger
features (or level 0). Such trigger features make the counterfeited note look blurred or brownish,
which triggers people to check for example the watermark and other public security features [48,
55].
When screen- and scan traps are printed too small, people will not notice them. That is why a
large circle was printed on the NLG 100/Snipe (Figure 31). Disturbance by interference (moir
patterns) or by aliasing effects (i.e. eigenfrequency) would disturb the homogenous circle so was
the design idea.
detail mirror
metameric colours
fading to zero
Figure 31.
Left: NLG 100/Snipe with screen and scan traps, issued in 1981.
Right: NLG 250/Lighthouse (reverse) with resolution indicator or detail mirror and screen and scan
traps, issued in 1986.
The NLG 100/Snipe was the first banknote with screen traps based on line pairs with exactly
equal line widths (a) and (b), leading to 50 % coverage (Figure 32). Unfortunately, the best line
pair specification the printer was able to print was between 40 % and 60 % coverage, which made
the screen traps less effective.
Page 47/84
Cut-off frequency
A second defence line is based on the principle that a reproduction system will not be able to
reproduce line frequencies above its eigenfrequency (Figure 32). In fact, the Nyquist theorem
explains that the sample frequency of the system must be at least twice the resolution of the
printed matter (Harry Nyquist, 1929).
a
Table 17 provides an overview of several typical cut-off frequencies of printing units and imaging
software, provided in lp/mm. Not every new technique achieves better performance than previous
ones. A photograph taken about 200 years ago, a Daguerreotype, typically has a pixel size of
about 0.5 m (or 100 lp/mm), the size of a particle of silver amalgam, while a standard digital
photograph today would have a pixel size of 6 m (or 8.3 lp/mm).
Cut-off
frequency
[lp/mm]
1.25
2.4
3
5
6.25
6.6
7.9
8.3
10
11.1
12
13.2
14
15.7
50
100
Imaging software
Cut-off
frequency
[lp/mm]
50 -100
100
100
200
200
700
Table 17.
Characterisation of reproduction systems by resolution in terms of line pairs per millimeter.
Page 48/84
that same note. According to the printer, the 0.2 mm letter height proposed by DNB could not be
achieved, but DNB proved that it could, with the letter font shown in Figure 33 a) and b) [10].
Where several languages have to be used on a banknote, central bank and designers tend to limit
themselves to numerals (e.g. 50) or letter designs (such as EURO). Micro-symbols as
presented in Figure 33 c) and d) may be used to create new images and so may go down better
with the public.
(a)
(b)
(c)
(d)
Figure 33.
a) Micro-letters proposed by DNB for the NLG 250/Lighthouse based on the cut-off frequency of the
intaglio press.
b) The word EXPORT based on the letter type proposed in a).
c) Instead of micro-text, designs may use micro-symbols, combined here into the shape of a bottle.
d) Detail of c).
Page 49/84
dry offset 50 m 30 m 50 m
2 offset plates
dry offset
50 m 30 m
wet offset 30 m
30 m
50 m
30 m
3 offset plates
Figure 34.
Security principle of alternating line colours, based on two colours (left, since around 1925) and based on
three colours (right, since around 1960).
Figure 35.
Two examples of line work in banknotes.
Left: typical detail of line work, including guilloches, in NLG 10/Greybeard, issued in 1934 (design by
C.A. Lion Cachet).
Right: typical detail of line work in EUR 50, issued in 2002 (design by Robert Kalina).
Simultan printing
The successor of the RZ-press at security printer Ensched is the Simultan press. These
innovative presses were first introduced in the 1960s. A Simultan press is a brand name for a
printing press manufactured by a well known company in the security printing industry. This
press collects the print of several separate images - all images on one side - on a rubber blanket.
The same is done for several separated images on the reverse side of the banknote. The
registration of the offset plates within one side is high, today less than 3 m. Next these collected
images are printed in one run - simultaneously - on both the front and the reverse of the banknote
paper.
In traditional offset printing, ink separation is based on the repulsion between ink and water (or
wet offset). The printing plates of these new Simultan presses did not use water, but separated ink
and non-ink areas by using a slight relief, in fact a variant of letterpress. Therefore the Simultan
printing technique was called dry offset. The presses had another property unavailable in the
outside graphic industry: one plate on the front could be printed in perfect register with one plate
on the reverse (+/- 0.10 mm).
Page 50/84
To date the Simultan press has lost its added value. Lines in alternating colours and iris print, the
two typical dry offset features, are no longer a hurdle for the counterfeiter and are less used (e.g.
Figure 35). Also the fit of the front and the reverse, the see-through register, no longer provides a
defence against current reproduction techniques. Its main drawback is its low resolution (around 8
lp/mm); most commercial presses can do better. Also the line broadening because of the slight
relief of de dry offset plates is today a disadvantage for a security product like a banknote. As a
consequence there are no new public features using the techniques of the Simultan press, except
the Multi Variable Colour feature.
Multi Variable Colour
In 2004 a new public security feature was introduced using the three offset plates of the Simultan
press. The feature is called Multi Variable Colour (MVC) and is part of the Russian 100 rouble
banknote (Figure 36). The MVC is a smart construction making use of the high registration
between the three offset plates. The MVC feature shows all kind of rainbow colours when the
banknote is tilted.
Detail MVC
Figure 36.
Muti Variable Colour (MVC) security feature in Russian rouble banknotes, introduced in 2004. When
tilted, rainbow colours appear in the area printed under the denomination numeral.
Page 51/84
A
B
high density
area
low density
0
(c)
(a)
(b)
high density
element
Figure 38.
Examples of features based on low and high density.
a) Original NLG 1,000/Lapwing on white paper (issued in 1994).
b) NLG 1,000/Lapwing printed on slightly red paper (1993).
c) High density element printed in NLG 1,000/Lapwing (dark solid offset area with on top dark intaglio).
Page 52/84
8.2.4 Colour
The system approach also yielded some security features based on colour. Since 1850,
unsaturated colours have been used in many banknote designs, including the Dutch. Such colours
were often based on unique ink recipes. The well known Dutch artist Anton Pieck worked and
lived in Haarlem, also the residence of security printer Joh. Ensched. In the 1950s he regularly
visited the printer because he loved all the nice dark brown ink varieties he could find there.
In 1980, unprecedented in banknote printing, a very bright colour, a highly saturated orange ink,
was developed. The idea was to use a colour outside the euroscale colour reproduction standard.
The bright orange colour, showing up brown in a reproduction, was first introduced on the
NLG 50/Sunflower issued in 1982.
In 1986 some more design studies on colours outside the euroscale were performed. One idea that
came out of this study was to print a reference next to the colour outside the euroscale. This
reference was the colour the outlier colour would show up as in reproduction. In an original
note, the idea ran, the two colour areas should look different, while in a reproduction they would
look the same. Figure 39 explains the principle and provides some examples.
(a)
(b)
Examples for colour B
1. = PMS 021
2. = PMS warm red
3. = PMS rhod. red
4. = PMS 245
5. = PMS purple
(c)
(d)
Figure 39.
Study of colours outside the euroscale reproduction standard, 1986 [9, 11].
a) Basic principle. The colour B will be reproduced as colour B, which ideally should be similar to colour
A). Within the ellipse humans will perceive no colour difference between A and B.
b) The human eye will see two different colour areas (A and B) if the diameter of such an area is about 10
mm, corresponding to an angle of vision of ca. 2 at reading distance.
c) The colours developed, plotted in a CIE-diagram.
d) Samples of the colours developed.
Page 53/84
Metameric colours
An other example of a colour pair based on the colour dimension are so-called metameric
colours also recommended by the NRC in their NextGen report [62]. Metameric colours are two
colours (a colour pair) that are perceived as similar under one light source, e.g. daylight, and as
differently under another, e.g. neon light. Infra red (IR) images are also referred to as IR
metameric ink, since under daylight two inks will look the same, while with an IR-viewer one
ink becomes visible (absorbent in IR spectrum) and the other ink is not visible (transparent in IR
spectrum).
Metameric design in Dutch guilder notes
A green metameric colour pair was designed and introduced in the NLG 250/Lighthouse intended
for use by retailers. Seen through a red filter, a rabbit would show (Figure 40). In the years that
followed some more metameric studies were done by DNB and Joh. Ensched. One example is
shown in Figure 41. An other metameric colour pair, in the shape of a fish, was introduced in the
NLG 25/Robin issued in 1990. This was to be the last banknote model DNB incorporated colour
pairs in, because the design suffered of a lack of colours in the area of the metameric colour pair
and the feature never became popular. In 1996 Singapore issued a commemorative banknote of
SGD 25 using metameric colours (Figure 40). In 2001 DNB proposed a metameric barcode for
the euro banknotes (see Figure 9).
(a)
metameric rabbit in NLG 250
(b)
original SGD 25
(c)
one of the metameric colours
becomes visible using a red filter
Figure 40.
a) Metameric rabbit explained in the leaflet of the NLG 250/Lighthouse (1986).
b) Metameric security feature in Singapore SGD 25. A third brown pattern is used as camouflage.
c) Under a red filter the text 1JAN96 becomes visible (SGD 25).
original
colour copy
Figure 41.
Design study on metameric colours by Joh. Ensched, based on designs made by Hans Kruit. On the left
the original (here of course reproduced) and on the right a contemporary colour-copied reproduction (ca.
1986).
Page 54/84
no watermark
0 points
poor watermark
only dark tints
1 point
good watermark
dark and light tints
2 points
Figure 42.
Six examples of counterfeited public security features in euro banknote forgeries and the assignment of
points according to the simple method.
Page 55/84
Maximum score = 12
Euro banknotes have 6 public features, so the maximum counterfeit quality score is 12. Figure 43
is taken from the report on banknote counterfeit quality in the Netherlands as measured in
September 2009.
12
10
8
6
4
2
50
C
78
50
K2
50 4
P1
2
10 a
0P
12
10
0P
7
20
P2
d
50
P5
50
0P
10 2
0P
1
20 3
0P
3a
Figure 43.
Quality of euro counterfeits in September 2009, ranked according to the simple method. The 200P3 has the
highest quality (10 points) and the 50C78 the lowest (3 points). The average counterfeit quality in this
month is 6.4 points.
Page 56/84
12
10
20
Score
50
6
100
200
4
2
0
Jan
2006
May
Sept
Jan
2007
May
Sept
Jan
2008
May
Sept
Jan
2009
May
Sept
Figure 44.
Weighted average score of the quality of counterfeit euro banknotes in NL since 1 January 2006.
The quality of counterfeited euro banknotes in NL is declining. This is especially true of the euro 50
counterfeits (from 7 down to 5 points).
This exercise yields an interesting conclusion: the quality of counterfeits is not rising, but
declining. From the graph in Figure 44 it can be concluded that:
- there is a difference in quality per denomination,
- the quality of the most frequently counterfeited note, the euro 50, is declining,
- the quality of counterfeits in general is stable if not slightly declining,
- the euro 50 has the poorest quality (!, since it is the most used denomination in NL).
Quality of counterfeited public security features
The simple method also delivers the individual quality of each public feature as provided in Table
18, including the two retail security features.
Clearly the watermark is the weakest feature, in this perspective, since it is imitated most and
best. The iridescent feature seems to be the strongest public security feature.
Public feature
1. Watermark
2. See-through register
3. Foil
4. Security thread
5. Tactile effect intaglio
6. Iridescence/OVI
Retail feature
1. UV
2. IR
Score
(2 points max)
1.7
1.5
1.4
1.3
0.7
0.3
0.8
0.1
Table 18.
Quality of counterfeited public security features in euro banknotes, based on the 10 most frequently
received counterfeit types in the Netherlands (April 2009 figures).
Page 57/84
Figure 45.
Colour copy reproductions of NLG 1,000/Lapwing made by DNB at the RRC in Copenhagen in 1995.
A further development of this exercise would be to test such counterfeits on an external panel.
Retailers, law enforcers and consumer organisations could be invited to sit on such a panel. The
test report could also be part of the presentation to the Board.
Track of the time, expertise and investment
The report to the Board would be even more informative if it also included time, expertise and
investment needed to reproduce the freshly designed banknote. Table 19 presents an imaginary
example of such a method, reporting on the time, expertise and investment needed to reproduce a
newly designed banknote.
Page 58/84
COUNTERFEIT INVESTMENTS
Time
Expertise
Cost in EUR
Push-the button: 1 s
Primitive
0
100 hour
Primitive
100
1 day
Casual, hobbyist
~ 1,000
10 hour
Primitive
~ 100
1 week
Professional
~ 10,000
3,000/hour
Casual, hobbyist
~ 50,000
Table 19.
Overview of the time, expertise and investments needed by the counterfeiter to make a reproduction
Expertise subgroups correspond to those defined in Chapter 2 (primitive, hobbyist, professional and statesponsored). The figures mentioned are estimations. A reference is needed to judge if the counterfeit
resistance of the new note is poor, medium or good.
Counterfeit resistance
1. I
2. H
3. F
4. B
5. C
6. D
7. J
8. K
9. E
10. L
11. G
12. A
Table 20.
Overview of the public preference for 13 hologram designs and the quality of the counterfeits. The results
are part of a DNB study Foil with public appeal prepared for the ECB in 2004/2005.
Green = good, favoured by public, difficult to counterfeit, Yellow = medium, Red = poor, rejected by
public, easy to counterfeit.
Page 59/84
Page 60/84
the same opinion The luminescence of the counterfeit is poor, yet close to that of the genuine
note. [94].
Figure 46.
Banknotes under UV light. Because of the brighter reflection of the UV features in the counterfeited note,
many people accept this note as real.
Page 61/84
Attribute
Attribute level
3. Degree of conspicuousness
4. Degree of complexity
5. Pictorial element used for security
feature
Table 21.
Overview of the six attributes of the security features of a banknote and their levels used in the conjoint
analysis 2009.
Score in %
30
23
18
13
9
6
Table 22.
Relative importance of the characteristics of the security features on euro banknotes.
Page 62/84
Location
Degree of eye-catching
Degree of complexity
Kind of image
Appearance
6*
4
2
Should be kind of hidden*
Should be noticed in one glance
Make effort to check*
Check in one glance
Abstract
Number
Figurative*
Modern, state of the art
Same as the current ones*
Figure 47.
Relative importance of the characteristics of the security features on euro banknotes. People would like to
see features that can be checked in one glance.
To men, the location and the number of security features are slightly more important. To women,
the degree of complexity is slightly more important. Youngsters find the number of security
features most important, as opposed to the elderly, who find the type of image used for the
security feature is most important.
Operational model
The conjoint analysis is a working model. The model is filled with the data obtained in 2009 and
provided on a CD-ROM. The model accepts variable input, simulating different banknote
concepts put together using the several attribute levels mentioned. The model will tell the user the
level of public acceptance relative to another concept, e.g. the existing euro 50.
Page 63/84
four security features, it seems that four features is enough [25, 48]. However, central banks may
want to incorporate more than four PSF and opt for a mix of active and sleeping public security
features. Three features will be actively communicated, while three public features are kept
sleeping. Sleeping features will be awakened - some or all - when one or more of the other
features are heavily counterfeited.
In the example of Table 23 a total of six PSF, three active and three sleeping, is divided over the
front and the reverse. However, different concepts are possible.
Front
+
+
+
Reverse
o
o
o
OLD
SERIES
2 - 4 PSF
NEW
FEATURES
2 - 4 PSF
NEW
SERIES
6 PSF
Figure 48.
Example of the gradual approach applied to the selection of 6 PSF for the new series: 2 - 4 public features
are retained from the previous series and 2 - 4 new features are added.
PSF = Public Security Feature(s).
Page 64/84
Watermark
Hologram/silver foil
Security thread
See-through register
Raised ink, relief
Special ink (iridescent ink/OVI
Public knowledge in
NL (2009)
76 %
55 %
15 %
9%
8%
2-3%
Cherish awareness of
old series as gold!
Table 24.
Keep the best-known features. Cherish familiarity with security features in old series like gold!
Public awareness of existing features is taken from Table 3.
Policy ECB
At the Currency Conference 2007, the ECB presented the desired number of security features for
the next series of euro banknotes (Table 25), based on work done in 2004 -2005 [53, 70, 88].
User group
Level 1
Level 2
Level 3
Number
Public
a) Retailers human
b) Retailers automatic device
c) Third-party sorting machines
Central bank sorting machines
Remarks
6
2
At least 4
4 to 6
2 or 3 in substrate and 2 or
3 applied in printing works
Table 25.
Overview of the proposed number of security features in the Euro Series 2.
The proposed mix of public features may be completed by dividing the 6 features over the Feel
Look Tilt motto. Within each of these three categories, the best known features from the previous
series could be kept. In case of six public features a total of three new features are needed,
preferably also divided over Feel Look Tilt and would lead to a marketing mix as first suggested
in 2007 by De Heij, in this paper reproduced as Table 26 [55].
Public use
FEEL
LOOK - through
- at
TILT
Public
knowledge
8%
76 %
55 %
-
Table 26.
An example of a set of public security features based on a marketing mix.
Page 65/84
Retail features
human
Public features
Retail features
automatic device
Features banknote
sorting third party
Features banknote
sorting central bank
Figure 49.
Strategy for selection order of security features. First the retail and public features should be selected, since
they are dominating the design. Machine readable features have a larger design freedom and there is
much more choice.
Page 66/84
Public
level 1
6
Retailer
human
level 2a
2/3
Retailer device
level 2b, 2
Third party sorting
level 2c, 1
Central bank
level 3
2
CDS
level 4 3
Forensic
level 5 1/2
TOTAL
Model A
+0%
1. Colour outside
euroscale
2. Grey colour
3. Screen and scan traps
4. Feel: CtIP
5. Feel: nail scratch area
1.
2.
Model B
+5%
Colour outside
euroscale
Different paper tints
2.
Model C
+ 10 %
Colour outside
euroscale
Different paper tints
3.
4.
5.
6.
Look: watermark
animal with digitized
area
Look: secure
window die cutting
1.
6. Look: watermark
portrait without
digitized area
7. Look: full embedded
security thread, e.g.
Wings
8. Tilt: foil/hologram,
continuous stripe, 10
mm
9. Tilt: iridescent band
9.
13.
14
15.
12.
13.
14.
16.
17.
18.
19.
20.
21. nano lettering
21
15.
16.
17.
18.
19.
20. botanical DNA
21. UV-fibres extrusion
21
16.
17.
18.
19.
20.
21. botanical DNA
22. nano lettering
22
7.
Look: windowed
thread colour switch
7.
8.
Tilt: foil/hologram,
registered stripe, 12
mm
Tilt: micro optics
8.
Tilt: foil/hologram
registered stripe, 20
mm
9. Tilt: offset, OVI +
blind embossing
10. IR-image
Table 27.
Example of three conceptual banknotes A, B and C with different cost prices and different sets of security
features. In this example a total 21 or 22 security features are foreseen. CtIP = Computer to Intaglio Plate.
Retailer human are features were some aid is need like a polarisation filter or UV lamp and the decision is
human.
Page 67/84
security features
banknotes
central bank
sender
public
receiver
Figure 50.
Basic sender-receiver communication model. The banknote security features communicated by the central
bank (sender) do not get across to the public (receiver).
Let us start with the following statement by the ECB in October 2007: With respect to
communications on the current series of banknotes, such qualitative research has helped to make
communication tools, such as brochures, leaflets and electronic communication media more
easily comprehensible by avoiding technical terms for the security features and by providing
simple instructions on how to authenticate a banknote. [60]. This policy should be the starting
point for a new series of banknotes, rather than being developed once the note is ready for issue.
A communication strategy should already be part of the design process as described in Public
feedback for better banknote design 2 [55].
Series concept
The communication strategy should first of all cover the complete series. A generic strategy is
required, leaving freedom to individual communication concepts for each denomination [25].
The euro banknotes have a so-called split between low (5, 10 and 20 euro) and high
denominations (50, 100, 200 and 500), each subset having partly different security features.
Different public features for low and high denominations make it more difficult to explain and to
communicate. ECB research reported in 2004 that close to 70 % of the cash handlers did not
know that there were two groups of euro banknotes: the low and the high denominations, bearing
different security features [34]. Research by DNB in 2007 reported a similar conclusion: The
distinction between low and high euro denominations is not effective as it leaves both the public
and retailers confused about the security features, besides making the public information tools too
complex. [48].
A split is not appreciated and it is advisable to apply the same security features throughout the
series. For reasons of communication all features should be on the front of the banknote, making
a communication concept even stronger. The recent conjoint analysis indicated that people prefer
public features both on the front and on the reverse. On the other hand, time spent on the reverse if at all - is limited to around 1.5 seconds.
For a new banknote series a central bank should decide between better public recognition (all
features the same) and higher counterfeit resilience, which always is a trade-off as shown in
Figure 51.
Page 68/84
Lower Higher
Counterfeit resistance
Different
techniques,
similar
designs
All features
similar
All features
different
Similar
techniques,
different
designs
Simple Complex
Public information
Figure 51.
Trade-off between similar security features in all denominations and simple or more complex information
tools (e.g. front side of one note explains all). The design of the features can be different, but based on the
same technique.
Since the public is quite unaware of security features, it seems that central banks should focus on
better public recognition. This argues in favour of using the similar technique for each public
feature throughout a series and incorporating all features on the front. Each denomination within
a series could be presented in different designs, such as different watermarks, different paper tints
and different foil elements.
Tell a little story
The optimal banknote design uses features that are easy to remember but still difficult to
counterfeit. In several publications De Heij plead for a little story as was the case for the Dutch
guilder notes [25, 48, 55]. Based on this design philosophy an example was prepared for the
Central Bank of Aruba in 2007. In Table 28 the letters A-R-U-B-A tell a little story about the
island, for each denomination a different story.
Watermark
Florin
10
20
50
100
200
500
A
Portrait of
Indian
See through
(offset)
R
Turtle
Donkey
Pelican
Palm tree
Willem III tower
Carnival feather
Relief print
(intaglio)
U
Dolphin
Rattle Snake
Owl
Dividive (tree)
Bird painting
from cave
Group of people
in carnival-like
clothing
Colour crypt
(silk screen)
B
Shell
Iguana
Butterfly
Cactus, Aloe,
Old coin
Masque
Colour shifting
wide thread
A
5 continuous
silhouettes:
fish, flower, bird,
tree, persons
head, carnival
element
Table 28.
A first proposal for design/communication of 5 public security features. Each public feature is identified
with a small letter symbol, reading from left to right: ARUBA. Because of the low volume, the watermark
is kept the same for each denomination.
Page 69/84
The 10 Florin tells about the animals in the sea and the theme of the 20 is about animals on land.
Animals of the air are found on the 50 animals and the 100 tells us about the typical plants on the
island. Old and new architectural elements are used on the 200 and on the highest banknote
people are filling the carnival theme. To illustrate this concept further a conceptual banknote
design is made for the 100 (Figure 52).
Figure 52.
Conceptual banknote design for a 100 Florin banknote for the Central Bank of Aruba. Five public security
features may be found by following the letters A-R-U-B-A. The study was focussed on banknote identity,
in this case by following design elements of banknote designs of European Union, Netherlands, Caribbean
area, South America and USA. Design by author (2007), published with permission of Central Bank of
Aruba.
Page 70/84
20
BANKNOTE
change
euro
twenty
wallet
one note
Figure 53.
Representation of a euro 20 banknote in the visual working memory. Only a few features of the banknote
(colour, hologram, numeral and flag) are linked to the knowledge we have about banknotes.
Based on the example of a dog in Visual Thinking for Design [67].
Page 71/84
The following is based on the perception of a dog as described by Colin Ware, applied by your
author to a banknote. Let us take the situation where we accept a banknote in return at the
supermarket checkout counter. How do we perceive the change? Assume the change is one 20
euro banknote and some coins. Our rapid eye movements (or saddic eye movements) capture the
banknote. From the retinal image some features are captured and processed bottom-up. The
captured features could be the ones we know from recollection: colour, denomination numeral,
flag and hologram. These two or three features are low-level features, matched by the brain to
relevant patterns (a crude map) and finally to an object (a banknote). This object is not simply an
image of the banknote. The rest of the banknote features are not yet activated, as illustrated in
Figure 53.
4.
What is active?
Eye movements
Eye movement catches a
banknote
Inferotemporal cortex (IT)
6.
7.
8.
Linguistic/semantic judgement
9.
Search PSF
10.
Motto
11.
Check PSF
12.
Cognitive thread
13.
14.
15.
5.
Comments
Occurs whenever our eyes alight (0.04 s).
Banknote image falls on fovea to be analysed by the brain
(0.04 s).
In the IT cortex neurons responds to meaningful patterns, e.g.
a banknote or a car. Rapid characterisation of the banknote,
categorical judgements. Sub area banknote found in 0.1 s.
A banknote is a strong gist that will increase the number of
cognitive cycles concerning the banknote.
Three-dimensional shapes or skeletons are stored in the brain.
Example: 2D images with shades.
Pop-up differentiators in V1 (colour, shape, texture, motion,
depth).
V2s neurons respond to slightly more complex patterns,
based on the processing already done in V1.
In V4 more complex compound shapes are identified from
patterns of features.
The banknote image is processed in both the verbal and
visual working memory (0.1 s 1 s, max 2 s). Limitation: up
to 3 items.
Are we looking for denomination, public security features
(PSF) or something else?
Orientation and navigation: way finding features, retrieval
path, pre-set lay outs.
Link visual information to verbal working memory. E.g.
Feel-look-tilt or Look for Yvonne.
Identification, information: either on colour, geometry,
resolution, etc.
On going chain reaction towards higher levels of pattern
recognition.
Feature OK. Check more.
Recollection of: check more than one feature!
Repeat motto. Etceteras.
Table 29.
Schematic breakdown of the brain process involved in checking a banknote, based on the theory described
by Colin Ware. Best guess by author!
Page 72/84
At the same time the banknote is processed bottom-up, the brain also processes the note topdown. Top-down processes (or high-level attention) reinforce relevant information, causing a bias
in favour of the signals we are looking for. If we are looking for the banknotes value then the
colour and denomination numeral will signal louder. If we are looking for the public security
features, then the watermark and the hologram will have their signals enhanced. This biasing in
favour of what we are seeking or anticipating occurs at every processing stage. Telling the
banknotes value is accomplished by an iterative process of linking and re-linking visual and nonvisual information about the banknote.
Because of the very limited capacity of the visual working memory, most information about the
20 euro banknote will be lost as we trace out the coins. We will most probably not check if the
received note is genuine.
Table 29 presents a first attempt to make a breakdown of the brain process involved in checking a
banknote, based on the book by Colin Ware. More study is clearly needed.
Figure 54.
Example of eye movement planning (follow me) for 6 public security features on a banknote. The feature
in the top right corner is ignored.
Page 73/84
(a)
(b)
(c)
Figure 55.
Explaining security features by marking them in the information leaflet. Why not do this on the banknote
itself?
a) image of HKD 1,000 as used in leaflet of the Hong Kong Monetary Authority (around 1995).
b) image of ZAR 20 as used in leaflet of the South African Reserve Bank (around 2005).
c) Information tool on credit card size by ECB using numbers to indicate the public security features
(2007).
Figure 56.
Dummy note with way finding features used for testing comprehension (self explaining and searching).
DNB, 2003.
Figure 57.
Communication symbols introduced by the Bank of Canada early 2008. The design is based on the wayfinding icons developed by DNB in 2003. An additional colour is introduced for each function. Also each
symbol shows a hand. The motto is TiLL: Touch, Tilt, Look through and Look at.
Page 74/84
The developed way-finding icons became quite popular as a communication tool. After DNB had
used them on its CD-ROM Genuine or counterfeit in 2002, several others adopted them, such as
the European Central Bank (2003) and the central banks of South Korea (2006), Chili (2006),
Mexico (2007) and Canada (2008) [55]. Often the symbols were modified, as in Canada (see
Figure 57). Despite the copyright aspects this is a peculiar habit, since the altered symbols are not
tested on their comprehensibility like is the case for the original proposed symbols [44].
Pubic security features indicated by a letter
Whereas way-finding icons are used in communication tools around the world, the idea of
printing them on banknotes has met with resistance, although it has been proved that doing so
will increase the average number of public security features recognized from around 2 to 4.7 [55].
However, alternative design solutions have been developed as shown earlier in Figure 52. All five
public security features are simple to find, just follow the letters A-R-U-B-A. The same concept
may be used for a euro 50 banknote as shown in Figure 58. Here the word E-U-R-O-P-A is
indicating six public security features. The basic idea behind these concepts is to find first a
meaningful word, supporting the philosophy of the playing man (for the Dutch: humo ludens),
and subsequently detect the public security feature for each letter.
Figure 58.
Conceptual banknote EUROPA with six public features E, U, R, O, P and A. People are invited to find all
features and works as a memory aid. The letters will also be understood by other cultures as (compare the
gate letters at airports. In international championships Greek football players have their names in Latin
writing. In Russia the plate numbers on cars are in Latin, not Cyrillic.
Two variants are possible: the currency code E-U-R (three active features on the front) or E-U-R-O (four
features on the front). Design by author.
Page 75/84
10. CONCLUSIONS
10. General
10.1.1 Public perception of counterfeits in circulation is far above the reality; people
overestimate the reality by 200 % (NL, 2008).
10.1.2 To assist the selection of security features for new banknotes a model is required.
However, there is no such proven model available. Recently some models have been
developed by central banks, but these models are not (yet) mature.
10.1.3 Different selection models lead to different preferred features as shown in Table 30.
Selection of security features should be done by using different selection models.
Public security features (in euro)
Watermark
+
++
++
++
-+
o
o
Hologram
/foil
-+
-+
o
o
--
Security
thread
+
o
o
+
o
+
o
+
Tactility
+
+
o
+
++
++
o
Colour
changing
ink
--o
++
o
-
Glossy
gold
stripe
++
-o
o
++
N/A
o
o
Seethrough
register
+
+
--o
+
Table 30.
Overview of the appreciation of the public security features in the euro 50 banknote, showing 9 different
feature selection models.
10.1.4 Dedicated research of central banks will lead to public and retail security features better
serving the needs of these user groups. Central banks should write the requirements, the
industry should develop.
Worldwide banknotes receive similar public security features, because the introduction of
new security features in banknotes is often triggered by suppliers and early adaptors.
10.1.5
New features are usually a dedicated barrier against the counterfeits coming in
(photography 1850s, colour copiers 1980s, home scanners 1990s). New security features
should be a barrier against the new threats. After 30 years of overwhelming innovations
in the graphic and information technology industry, there seem to be no new (technical)
threats on the horizon.
10.1.6 While the market is offering more and more add-on security features, the choice of
internal security features is limited. Yet, if a choice can be made, internal features are
preferred over semi-finished features, because they can only be created inside a paper
mill or banknote security printing works. By principle internal features are more secure.
Page 76/84
10.1.7 The focus of counterfeit reports is often on statistical data and counterfeits seized,
illustrated with anecdotic stories, usually about the organisation of the criminals. Studies
on counterfeits should be more analytical from the perspective of:
- the range of the dimensions of the reproduction equipment, like resolution, colour,
opacity, geometry, mass and material,
- counterfeiter, like production time, skills and investment costs.
10.1.8 The public is not interested in public security features since the probability to receive a
counterfeited banknote is very low. This explains why the quality of the counterfeited
euro banknotes is relatively low (6.4 on a scale of 0 - 12). Probably this quality will
decline further. Central banks are right that the difference between a real and a
counterfeited note is easy to see, but their message does not come across to the public.
10.1.9 Instead central banks should focus on the retailer, who is the target of the counterfeiter.
The retailer is key in the combat against the acceptation of banknote counterfeits. New
banknote designs should assist first of all the retailer.
10.1.10 Public interest in public security features is key to get public attention. Develop a slogan,
or keep the one that is successful (e.g. feel-look-tilt). Give the new note a name. All
within an attractive design.
10.1.11 From guilloches to holograms; every generation of banknote developers joins in at a new
security technology and seems to forget about the old one.
Page 77/84
10.3
10.3.1 A public security features should full-fill all user requirements: fast (< 2 s), easy to use,
discreet, reliable (e.g. durable), striking, univocal, easy to communicate and no nested
features.
10.3.2 Public security features should have a similar authentication procedure on all
denominations; probably, but not necessarily based on the same technology. Avoid
denominations within a series bearing different public security features like a split in low
and high denominations.
10.3.3 Throughout the series public security features should always be on the same location.
10.3.4 Develop a slogan, or keep the one that is successful (e.g. Feel-look-tilt, Look for Yvonne
or E-U-R-O-P-A).
10.3.5 Within the individual banknotes the design theme of the public security features can be
different (tell within each note a little story using the colour of the note, the name of the
note or something else).
10.3.6 The public will take about 5 seconds to check a banknote (although usually they will not
check at all).
10.3.7 Three features should be checked, so 1 to 2 seconds per feature.
10.3.8 Divide the public features (i.e. 6) in active and sleeping features (i.e. 3 active and 3
sleeping)..
10.3.9 Active features are promoted. Once a preset threshold level is passed, the sleeping pubic
features could be promoted.
10.3.10 All active public features should be on the front for two reasons:
- preference of the public: in general people do not look at the reverse,
- communication: one poster showing the front of one note is representative for the
complete series.
One or two sleeping public security features could be on the reverse. All human
detectable retail features could also be positioned on the reverse.
10.3.11 Public features based on similar human operations like feel, look or tilt could be grouped
together so that these features may be checked in one glance. This is especially true for
tilt and look-through features.
10.3.12 Pictorial elements like way finding features or letters will help the public to trace the
security features.
10.3.13 Start the design process of the individual banknotes with the development of an eye
travel path. Start with the public and retail security features (there is more design
freedom for detector readable features).
Page 78/84
10.4
Retailer
10.4.1 A retail feature should full fill all user requirement, like a total authentication time < 2 s.
10.4.2 Retailers do not use a magnifying glass (neither does the public). From a marketing point
of view there is a need for a follow up. Features based on a polarisation filter seem to full
fill the user requirements of the retailers.
10.4.2 Counterfeits have good imitated UV features, often even brighter than the ones in the real
notes. The retailer should not rely on a human operational check under a UV lamp.
10.4.3 Detectors telling yes-or-no the note can be accepted are preferred over detectors that
need an interpretation by the retailer.
10.4.4 Maximum 3 features for retailers, based on different dimensions. A taggant could used,
since the detection speed can be slow.
10.4.5 All human detectable retail features on one side (reverse).
10.5
10.5.1 Instead of adding CDS-features once the design is ready, such CDS-features should
become part of the design process. Therefore specifications are needed.
10.5.2 CDS-features reduce the design quality of a banknote. Such features require too much
space and provide the note with a blurred and pale look.
10.5.3 An intrinsic and invisible feature using the complete banknote surface would be an
optimal CDS-feature (like e.g. a spectral property or a special fibre added to the
substrate).
10.6
Forensic features
10.6.1 Since forensic features are small, some forensic features could be introduced like e.g. a
taggant, fibres with an extrusion profile and a nano-text (letter height 2 m).
10.7
10.7.1 Tests on counterfeit resistance should be done on the complete new banknote.
10.7.2 Tests on counterfeit resistance should be reproducible by third party and should also
report on needed skills, investments and time.
10.7.3 Self made counterfeits of the new note should be tested with retailers and public
(providing feedback on the banknote design and its features).
10.7.4 Proposing proof prints of the new banknote to the Board should include two self made
counterfeits: a push the button copy and a counterfeit on a professional level.
Page 79/84
10.8
10.8.1 Reporting on counterfeits should be done by counterfeits detected per million notes in
circulation (counterfeits passed in c/mln).
10.8.2 Information about counterfeits in circulation to the public is fragmented and unclear,
leaving the public with a passive attitude and negative feeling. Central banks report in a
statistical and juridical way on their counterfeited banknotes and i.e. not in a more
informative manner like reporting on:
- the chance to receive a counterfeit,
- public confidence,
- the difference between genuine and most counterfeited notes.
10.8.3 Central banks do not collect information on how often a security feature is faked. Such
information is valuable feedback for the design process of a new banknote.
10.9
Counterfeits in NL
10.9.1 Since the introduction of the euro banknotes in 2002 Dutch retailers are confronted with
about 4 times more counterfeits in 2009.
10.9.2 Best known public security features in the Netherlands are watermark, hologram/foil and
the security thread.
ACKNOWLEDGEMENT
This paper is part of the fieldwork of my PhD-study Key elements in banknote design at the
Delft University of Technology, Faculty of Industrial Design Engineering. The paper is a first
investigation of this research area and will be discussed with my tutors and supervisor and
investigated further.
Several people have contributed so far to the content of this paper. Since the paper is extensive,
not all have commented on all subjects. Thanks are due to the following persons and or
organisations for their remarks or support:
- Ms. Jolanda Hijlkema - Duikers (DNB),
- Mr. Marco Wind (DNB),
- Dr. Peter Koeze (DNB, retired),
- European Central Bank.
Further thanks go to Mr. Jan Binnekamp, head of the Currency Policy Department (DNB), for
making this paper possible and to Mr. Fred Collens (DNB) and Mr. Ren Kupershoek (DNB) for
editing my English.
Page 80/84
REFERENCES
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
Page 81/84
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
French, John; On the right wavelength - magnetic particle printing based on a presentation made by
John French, Product & Image Security 1997.
De Heij, H.A.M.; The design methodology of Dutch banknotes, Optical Security and Counterfeit
Deterrence Techniques III, Proceedings of SPIE Vol. 3973 (ISBN 0 - 8194 - 3591- 0), San Jose,
California USA 27 - 28 January 2000.
Bakker-Rijnbeek, L.M.; Jaarverslag falisificaten 2000, De Nederlandsche Bank NV, Amsterdam, 6
September 2001 confidential.
Keyworth, John; Forgery. The Artful Crime. A brief history of the forgery of Bank of England notes
Bank of England Museum, London 2001.
Van Renesse, Rudolf; Hidden and scrambled images - a review Proceedings SPIE, 4677 (ed. R. van
Renesse), 2002, pages 333-348.
De Heij, H.A.M., Lisa a. DiNunzio and Olivier Strube; 'Comparence EUR - USD Banknotes Public's
Appreciation and Knowledge' De Nederlandsche Bank NV, US Treasury BEP and European Central
Bank, Presented at Banknote 2003, Washington DC, USA, 2-5 February 2003.
De Heij, Henricus A.M. and Thomas Buitelaar; Series of security documents provided with a
watermark in the form of a barcode EP 1 273 461 A3, European Patent Office, 5 March 2003.
De Heij, Hans and Ton Roos; Free Intaglio Detector, letter to De La Rue Currency, De
Nederlandsche Bank NV, Amsterdam 27 March 2003.
Next generation of euro banknotes will be different interview with Antti Heinonen, Keesings
Journal of Documents, Issue 3, Amsterdam 2003.
R&D Plans for new Euro Unveiled interview with Brian Dennis, Currency News, Volume 1, No 5,
May 2003.
Quantitative survey Professional cash handlers by TNS/NIPO, prepared for the European Central
Bank, Frankfurt, 7 February 2004.
Church. S and L. Setlakwe; Analysis of counterfeits and public survey results as design input
Proceedings SPIE, 5310 (ed. R. van Renesse), 2004, pages 63-73.
Betalen kost geld, Werkgroep Kostenonderzoek Toonbankbetaalproducten, Maatscahppelijk
Overleg Betalingsverkeer, De Nederlandsche Bank NV, March 2004. A summery The cost of
payments can be found in the Quarterly Bulletin of DNB, March 2004.
De Heij, Hans; Foil with public appeal (for new designs of euro banknotes) De Nederlandsche Bank
NV, prepared for the European Central Bank, Amsterdam 9 June 2004 confidential.
De Heij, Hans and Jeanine Kippers; Efficient cash payments with euro coins and banknotes in the
Netherlands, De Nederlandsche Bank NV, presented at Banknote Printers Conference, Dresden, 6-9
September 2004.
The second series of euro banknotes Annual Report, European Central Bank, Frankfurt 2005
De Heij, Hans A.M.; Life cycle analysis of security features in banknotes; from central bank to
retailer.
De Nederlandsche Bank NV, presented at Banknote 2005, Washington 20-23 February 2005
De Heij, Hans and Javier Gamo; Counterfeit Report Foil with public appeal De Nederlandsche Bank
NV, prepared for the European Central Bank, Amsterdam 23 March 2005
The euro banknotes: developments and future challenges Monthly Bulletin, European Central Bank,
Frankfurt, August 2005.
De Heij, Hans; Evaluation of counterfeits, development of methodology De Nederlandsche Bank
NV, Presentation to RRC Users Meeting, Ottawa, Canada 28-29 September 2005.
Boersema, Theo; Comprehensibility of graphical symbols for clarifying security features Delft
University of Technology, Presented at Annual Meeting of the Europe Chapter of the American
Human Factors and Ergonomics Society Turin, October 2005.
Van Renesse, Rudolf L.; Optical Document Security (Third Edition), Artech House ISBN 1-58053258-6, Boston London, 2005.
De Heij, H.A.M. and T.R. Stange; Authenticity mark Deutsches Patent- und Merkenamt, DE 60
2005 000 658 T2 2007.11.15. European Patent Office, EP 1 607 235 BI. Date of filing: 21 December
2005. Date of publication and mention of the grant of the patent: 7 March 2007.
Jonker, Nicole, Bram Scholten, Marco Wind, Martijn van Emmerik and Marike van der Hoeven;
Counterfeit or genuine: can you tell the difference? DNB Working Paper 121, De Nederlandsche
Bank NV, Amsterdam 2006.
De Heij, H.A.M.; Public feed back for better banknote design DNB Working Paper No. 104, De
Nederlandsche Bank NV, Amsterdam, June 2006.
unused reference number
Page 82/84
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
78.
79.
Van Haeften, Ewout; Counterfeit threat assessment. A methodology approach power point
presentation to the Counterfeit Working Group (ESCB), De Nederlandsche Bank NV, October 2006.
Van Gelder, Alwin; Euro Counterfeits in the Netherlands De Nederlandsche Bank, presented at
RRC Users Meeting, Oslo 2000.
Lingnau, A., Francesco Pavani and Jens Schwarzbach; How do People Manipulate Banknotes?
Study for ECB, Center for Mind/Brain Sciences, University of Trento, Italy April 2007 confidential.
Heinonen, Antti; The euro banknotes: recent experiences and future challenges, Currency
Conference, Bangkok 6-9 May 2007.
First International Conference on the Protection of the Euro against Counterfeiting: The euro is in
safe hands Press release European Central Bank, The Hague, The Netherlands 16 May 2007.
De Heij, Hans; Public feedback for better banknote design 2, Occasional Studies, Volume 5,
Number 2, De Nederlandsche Bank, Amsterdam, September 2007.
Increases all round for Australia Currency News Volume 5, Number 9, Shepperton September 2007.
More variations on opals for security Currency News Volume 5, Number 9, Shepperton September
2007.
Dinjens, Marlies; Winkeliers doen biljet van 100 euro in de ban Volkskrant, Amsterdam 12
September 2007.
Buitelaar, Tom; Counterfeit Threat Assessment presentation to RRC User Meeting, Lausanne, De
Nederlandsche Bank NV, Lausanne, 25-26 September 2007.
Circulation and supply of euro banknotes and preparations for the second series of banknotes ECB
Monthly Bulletin, European Central Bank, Frankfurt October 2007.
Survey Among Cash Handlers in the Euro Area, TNS prepared for European Central Bank,
Frankfurt, November 2007.
A Path to the Next Generation of U.S. Banknotes: Keeping them Real US National Research
Council, Committee on Technologies to Deter Currency Counterfeiting, The National Academic
Press, ISBN 0-309-10575-7, Washington 2007.
Holm, Elizabeth; A flow model for banknote feature evaluation Sanda National Laboratories,
presented Optical Document Security, San Francisco 23-25 January 2008.
Church, S., T. Granzotis, M. Lacelle and A. Firth; Methodology for establishing bank note security
requirements Bank of Canada, presented at Optical Document Security, San Francisco 23-25 January
2008.
De Heij, Hans A.M.; Programme of Requirements: a powerful tool to develop new, secure
banknotes De Nederlandsche Bank N.V., presented at The Conference on Optical Security and
Counterfeit Deterrence, San Francisco 23-25 January 2008.
How the euro became our money European Central Bank, ISBN 92-9181-985-9, Frankfurt 2007.
Ware, Colin; Visual thinking for Design Elsevier, ISBN 978-0-12-370896-0, Amsterdam 2008.
Optical document security - past, present and future Currency News Vol. 6, No. 2, Shepperton,
February 2008.
Dramatic Fall in Counterfeits Brings Canada Close to Target Currency News, Vol. 6 No. 3, ISSN
0895-9080, Shepperton, March 2008.
50 leads the way for 2nd series Currency News, Vol. 4, No. 4, Shepperton April 2008.
Euro banknotes - a tangible symbol of integration Monthly Bulletin, 10th Anniversary of the ECB,
European Central Bank, Frankfurt May 2008.
Security Feature (Usability Quadrant) Perception Study, by Daniel Smilek, Kelly A. Malcolmson &
Jonathan S.A. Carriere, Bank of Canada, presented at RRC User Meeting, Lausanne 2 July 2008 Confidential.
Negueruela, Dari J. and Maria Jos Fernndez; Have people learned to love euro banknotes yet?,
power point presentation, Banco de Espaa, Currency Conference Prague, 12-15 October 2008.
Thick, Jacqui; The Curse of Complexity power point presentation, De La Rue Currency, Currency
Conference Prague, 12-15 October 2008.
Pacreu, Jaime; Interview with Thomas A. Ferguson, former Director of the Bureau of Engraving and
Printing of the United States, Billetaria, No. 4, Madrid October 2008.
Main trends in counterfeit deterrence in the banking system of the Russian Federation, Banknotes of
the World, No. 10, Moscow October 2008.
NexGen 100-Dollar Note Will Become a New Stage in the Long History of US Money, interview
with Larry Felix, Banknotes of the World, Number 8, Moscow October 2008.
Geldsammler 2008/11, November 2008.
HBD Monitor Betalingsverkeer 2008, Hoofdbedrijfschap Detailhandel, Den Haag 2008.
Page 83/84
80.
Alter, Adam L. and Daniel M. Oppenheimer; Easy on the Mind, Easy on the Wallet: The Roles of
Familiarity and Processing Fluency in Valuation Judgments Princeton University 2008.
81. Currency Counterfeiting on the Rise Currency News, Vol. 7 No. 1 January 2009.
82. Firth, Andrea V. and Sara E. Church; Building better banknotes. The role of scientific research at the
Bank of Canada Journal of Documents & Identity, Keesing Reference Systems, Issue 28, Amsterdam
2009.
83. Ashbourn, Julian; Documented thoughts Journal of Documents & Identity, Keesing Reference
Systems, Issue 28, Amsterdam 2009.
84. Schilling, Andreas; Diffractive OVDs and banknote windows Keesing Journal of Documents &
Identity, Issue 28, Amsterdam, 2009.
85. The Bank Note Confidence Survey, web site Bank of Canada. Purchased 20 February 2009.
86. Visser, Julie and Judith Sonke; Euro Banknotes. A study about awareness and recognition of euro
banknotes among the Dutch report by TNS NIPO prepared for De Nederlandsche Bank NV,
Amsterdam, 24 March 2009.
87. De Heij, Hans; Innovative approaches to banknote design power point presentation, Watermark
Conference, Kazan 23-25 June 2009.
88. Towards the second series of euro banknotes interview with Antti Heinonen, Banknotes of the
World No. 6, Moscow, June 2009.
89. Van Roon, Joost; Contemporary passport design Journal of Documents & Identity, Issue 29, ISSN
1571-0564, Amsterdam 2009.
90. Latest developments in payment and settlement systems Quarterly Bulletin, De Nederlandsche Bank
NV, Amsterdam June 2009.
91. Surge in Bank of England Counterfeits Currency News, Volume 7, Number 6, Shepperton, June
2009.
92. Currency provides anything but a tall order for De La Rues new CEO Currency News, Vol 7, No 6,
Shepperton June 2009.
93. Kosse, Anneke; De veiligheid van toonbankbetaalmiddelen, een onderzoek naar de beleving en het
gedrag van de Nederlandse consument. De Nederlandsche Bank, Amsterdam, June 2009.
94. Do not look down on counterfeits Banknotes of the World, Number 7, July 2009 Moscow.
95. Biannual information on euro banknote counterfeiting press release ECB, 13 July 2009.
96. Meer valse eurobiljetten onderschept in Nederland press release DNB, 13 July 2009.
97. Infosecura, 13th year, number 40, July 2009.
98. Balueva, Tatiana; In search of new images interview with Hans de Heij, Watermark, Number 5, St.
Petersburg 2009.
99. Van den Kommer, Esther; Communication on euro banknotes, De Nederlandsche Bank NV, Cash
Seminar, Budapest, 7 October 2009.
100. De Heij, Hans; Banknote design for the visually impaired De Nederlandsche Bank NV, Occasional
Study 2009 Volume 7, Number 2, Amsterdam October 2009.
Page 84/84