The Exploitation of Technology as an Aid to Philately
By Dr John Horsey, County Philatelic Auctions
Image Comparison Process in Action as a video
The video process is described in detail later in this document
As part of the research for my book The £5 Orange, it was necessary to make extensive use of modern technology. This led to several new discoveries, which would probably not have happened using more traditional methods. The book covers new insights into surface printing by De La Rue, how and why the £5 Orange was used, its numerous plate varieties as well as details of the many forgeries and related Cinderella items. The research was based on over 3,500 stamps (mostly images, I hasten to add). The book also covers sections on the exploitation of technology as used in the research. The technology is not limited to the £5 Orange and is applicable to almost any stamp. As such, this article should be of interest to all true philatelists. Some of the techniques described in the book are outlined below.
Revealing Detail with Technology
Traditional approaches to philately have worked well for years but some can be enhanced with modern technology. In certain circumstances, simple devices like the magnifying glass can be surpassed using computers and scanners.
For simple high definition viewing, all one needs is a suitable computer and a scanner. Figure 1 illustrates part of a £5 Telegraphs die proof from a scan at 1200dpi (dots per inch) displayed on a standard 20 inch (50cm) screen. The image shown happens to be in Adobe Photoshop but, at this simple level, the Windows Photo Viewer software that comes free with current versions of Microsoft Windows is just as good.
The detail visible is incredible; the Queen’s head is almost a foot (30cm) across. It is not possible to see such detail across a large area with a normal magnifying glass. A typical Sherlock Holmes style magnifying glass shows a large area but usually at only 2x or 3x enlargement. Those magnifying glasses offering 10x or better magnification show the detail but of a limited area and often have distortion at the edges. A scanner suffers none of these problems and yields high magnification over a wide area without peripheral distortion. The computer screen can be viewed at a comfortable distance and is more restful on the eye than peering through a glass close-up to an object bathed in strong light.
A typical computer screen has a resolution of 72dpi, so a 1200dpi image can be displayed at just over 16x magnification without loss of definition. With a 600dpi scanner 8x magnification is still available which may be sufficient in many cases.
Some caution is needed with 1200dpi scanners. Some have that capability designed specifically for 35mm film and only the central portion of the scanner bed is at 1200dpi; not the entire width. Also a 1200dpi colour scan of an A4 page is over 100Mb and some scanner/computer combinations have limitations on overall image size. With these, a scan of a small area may be at 1200dpi, but a full A4 sheet can result in much lower resolution even though the scan might still claim to be 1200dpi. To check, simply enlarge an image to the extent shown in Figure 1 – if the image is blurred at 16x magnification, the scan is not a true 1200dpi.
While this detail could also be seen with a good magnifying glass, the computer image is easier to study and you can zoom in and out of the image with ease. It is also possible to arrange two half screen images in different windows side by side for comparison. Better still, images can be stored for later reference. Colour images can be converted to black and white for greater contrast and, with multi-colour images, colours can be filtered out for greater clarity. While computers offer major advantages, scans take time and effort. Many will prefer the ease of use of the magnifying glass.
Image Comparison Techniques
Collectors compare stamps or images frequently if only to identify their stamps against catalogue pictures. Often you look for small differences such as Die I or Die II in the Line Engraved penny stars of Great Britain or Types I and II of the Wildings 2½d. For these the naked eye or a simple magnifying glass is adequate but in both cases the catalogue tells you exactly what you are looking for.
It is less simple when progressing beyond the catalogues and you want to compare stamps to detect if they actually are the same or different. You are probing into the unknown. The simplest and cheapest option is again the magnifying glass by looking at parts of both stamps in turn to detect any differences. With such simplicity, it is easy to miss important differences. For example, it is seldom realised that 3d Wildings exist with heads at slightly different tilts.
Around 1974, there was a wonderful device called the Comparatec. This was an illuminated magnifying viewer that gave a 10x image of an area 40mm x 30mm on a viewing screen. In it you placed the two stamps you wanted to compare and positioned them so that the images of each were exactly superimposed. Once aligned, the device could alternate between the images. If the two stamps were identical, the image remained steady. In cases of small differences between the two, most of the design remained static but any differences flickered. It was most useful comparing overprints such as the GB Officials. In simple terms, you could line up the overprints and compare your stamp against a known genuine example (from the same setting) and determine whether it matched or not.
The Comparatec is seldom seen these days. It had a limited field of view which made it unsuitable for large items such as blocks and die proofs. Help is at hand with modern technology in the form of a scanner and image processing software such as Adobe Photoshop. Photoshop is extremely powerful for editing images but quite expensive and needs a fast processor with ample memory. It has more function than most people will ever need. Basically any image manipulation software that has a layers option and the ability to resize and rotate images may well be just as effective. The discussion that follows is specific to Photoshop: the terminology in other image editing software may be different.
The £5 Die Proofs exist with several different dates. This raises the question of what changes were made to the design as the engraver progressed to the final image. Some one might detect with a magnifying glass; others needed a better approach...
First scan the items you want to compare at high resolution. For preference 1200dpi or better although 300 or 600dpi is satisfactory in many cases. In Photoshop, store each scanned item as a separate file in TIFF format rather than the more usual JPEG files. JPEG files are compressed and some information in the stored image can be lost each time it is manipulated and saved. TIFF files are much larger but retain the full information content of the image. Any JPEG files should be converted to TIFF files in Photoshop before editing.
To illustrate the process, in Figure 2, we have images of two different die proofs dated Nov 16, 1876. Open the images in Photoshop but not in full screen mode. This lets you see both images side by side – if necessary drag an image to one side so that both are fully visible. Let us call these images ‘A’ and ‘B’. Using image ‘B’, select the whole or part of it and drag and drop it onto image ‘A’. This creates a new layer in image 'A' containing image 'B'. It is like having the images on two separate sheets of paper. If the sheets partly overlap, some of image ‘A’ will be hidden by ‘B’.
You now need to align image ‘B’ exactly over the one that is hidden. Using the layers tab, highlight layer ‘B’ (green arrow D in Figure 2) and change its opacity to 50% (blue arrow C). This makes the top ‘sheet of paper’ semi-transparent. You can now see through it and view both images as in Figure 2.
Within the combined image and with layer ‘B’ selected (green arrow), the free transform option in Photoshop enables you to move, rotate and proportionally re-size ‘B’ so that it exactly overlaps ‘A’. Figure 2 shows image ‘B’ aligned horizontally with ‘A’ and made the same size but offset from ‘A’. If ‘B’ is now moved up and to the left, the images can be superimposed exactly – the lettering of FIVE POUNDS will become sharp rather than doubled as they become aligned. (In Photoshop this alignment can be automated by selecting both layers simultaneously and using the ‘auto-align layers’ edit option. This caters for some variation in size and orientation but only works if the software can ‘latch on’ to sufficient common characteristics of both images).
Once correctly aligned, restore the opacity of ‘B’ to 100%. Being the ‘top sheet of paper’, only ‘B’ is now visible: ‘A’ is hidden beneath it.
Now for the clever bit: the layers panel has a marker (indicated by the violet arrow E) against each layer that determines whether that layer is visible or not. Rapidly and repeatedly click the marker against layer ‘B’. This will alternately display images ‘B’ and ‘A’. Any difference between the images will show as movement as in a cartoon film; identical parts of the images will remain static. The technique provides a very much enhanced Comparatec function and Photoshop allows you to zoom in on areas of interest.
Image Comparison Process in Action - the video repeated to be near its detailed explanation
Extending the Process to Multiple Layers
The Author took all 56 images from a British Library digital photograph of the black proof sheet of the 1877 £5 Telegraphs to create an auto-aligned 56-layer stack. When using multiple layers it is wise to allocate each layer a name – in this case the corner letters. Imagine this stack as a pile of 56 sheets of paper each with a different lettered image and with stamp AA at the bottom. If the other 55 sheets are made invisible, you have a clear view of stamp AA. Now repeatedly click the visibility marker against layer AB. The rapidly alternating images of AA and AB will show any differences between them.
This was repeated for each image to compare it in turn with stamp AA. The process highlighted some hitherto unknown plate varieties that exist throughout the life of the £5 Orange. It also drew attention to some surprising differences in the corner letters which in turn has challenged the established thinking on how surface printing plates were made - details are included in the book.
The same techniques were also used to reveal differences between various die proofs and also to compare postmarks. By detecting small displacements in the letter positions, some cancels that appeared the same were shown to be from different devices. In other cases, it confirmed that identical cancellers were used and, for some, proved cancels to be forged. It has also been employed to show if similar ‘Specimen’ overprints are from the same or different handstamps.
This modern simulation of the Comparatec is an extremely powerful philatelic tool and opens up whole new avenues in philatelic research which, of course, is not limited to the £5. Additional technological applications are included in the Author’s book. It can take a while to become familiar with the complex software and initially may involve extensive use of its help facility, but the user can soon learn shortcuts and become proficient in comparing images. Who knows what you might discover!
Dr John Horsey runs County Philatelic Auctions , the highly successful world-wide postal auction based in Basingstoke, UK. The £5 Orange is a superb 312-page hardbound book in full colour which was awarded a Large Gold Medal at Stampex in 2013. To obtain a copy, contact John via the County Auctions website.