Ancient Coin Photos

My experiments with a Nikon 990 Digital Camera - January 2002 update

A page guaranteed to bore all but those interested in coin photography.

When I wrote my first page on coin photography with standard cameras I believed that digital cameras available to consumers were not of sufficient quality for use by serious photographers of ancient coins (or anything else for that matter). This has changed. The available cameras have progressed considerably but are still not perfect. I remain unimpressed by the macro zooms provided on these amateur cameras for the single reason that they require working so close to the coin that there is little room to arrange lights. I still believe that better results are possible with a 35mm single lens reflex with interchangeable lenses but the appeal of the new breed of digital cameras was more than I could resist; I bought a Nikon Coolpic 990. This page is something of a diary of my learning process with this camera. Those of you considering such a purchase may benefit from my mistakes. Certainly I am no expert. Those who think they can educate me in a better way to get better pictures from this camera are asked (begged!) to write. Updating this page has taken longer than I expected. To avoid this page loading slowly for those with dial up connections I have compressed (see below for discussion) images more than on many of my pages.

We begin with a very important point. The current generation of digital cameras are very capable of producing photos much larger and higher quality than needed to illustrate web pages such as this or to use for selling coins on eBay! The photo above shows a heavy antoninianus of Gordian III (shot at an angle to show its thickness). The two separate images were combined into one file and reduced to the size you see here. At the right is a small crop from the full size image (do you recognize Gordian's sideburns?). The lack a sharpness you see in the inset is due to the coin not being sharp enough to use all of the detail recording capability of the camera. Try to imagine how large the whole image would be if it were displayed at the same scale as that sideburn! If your sole purpose of buying a camera is web use, the 3 Meg class cameras are serious overkill. You could do quite well with a camera half their price.

BROADBAND UPDATE: Now that more and more people have cable modem, DSL or some other form of fast internet access, it is possiblle to post web pages that contain really large images. For those interested in seeing a full size coin image from a 990 Nikon, I have added a combination of obverse and reverse totalling just under 1.3 Megabytes. I do not recommend persons with slow internet access downloading this image.

'Overkill' is not a problem if you are printing out paper copies of your photograph. Current consumer grade ink jet printers are capable of placing several hundred dots on every linear inch of paper. The combined file of obverse and reverse 3 megabyte images will allow printing a very good looking 8x10 photo. Lesser digital cameras can do equally well for web page illustration but will produce prints best restricted to smaller sizes. Some of what follows on this page makes good use of the fact that images resampled down to web friendly size will look better than the huge files from which they were reduced. Rather few of my coin subjects can withstand the microscopic scrutiny of this camera. This page is not an advertisement for the Nikon camera. Good results could be obtained by many other models. Allow me to repeat a thought from my photo page: Few bad photos are made bad by bad equipment; usually the problem is a bad photographer. The best lenses are only sharp when properly focused; the sturdiest cameras produce fuzzy pictures when held by trembling hands. Some other cameras may lack features described below while others may have something my camera lacks. I am no expert on all models; I am, as of yet, only minimally proficient even with my own camera. In addition to printing your own, excellent digital prints can be purchased online from a number of suppliers. I have used Ofoto and posted some of my favorite photos on their site where you can order prints up to 20"x30" with decent quality. Ofoto would be happy to sell you prints of my photos (or your own)!

The Nikon 990, in common with most digital cameras, offers a selection of sizes and qualities for saving images. These selections will allow taking between one(!) and 333 images on the included Compact Flash memory card. Larger cards are available that will multiply these numbers. The question that we must address is where we find the best balance between image quality and practicality. The composite photo illustrating this paragraph shows small sections cropped from three sizes of original images. The camera also offers three compression level settings. Due to limitations of these much reduced images, it is pointless to try to illustrate the 'in camera' compression modes here. For use on eBay, the fast loading 'basic' compression and smallest image size (640x480 or VGA) are more than sufficient and, as we will discuss later, should be reduced and compressed further to shorten the time required to load the photo. On the resolution available on a computer monitor, the difference between this basic level of compression and the next higher (called 'normal' by Nikon) is apparent. The highest quality JPG available ('fine') requires a paper print to appreciate the improvement. It is best to remember that we can always reduce and compress images but the camera mode settings provide a ceiling for the final quality of the image. I often shoot at higher quality JPG settings 'just in case' I might want to make a paper print of the image at some later time. There is one higher quality available on this camera where the file is saved as an uncompressed TIFF file (the one that gets only one image on the 16 MEG card) which is a feature that will be used by very few owners of this camera. If you have an application that requires this quality level, you will know the signs suggesting it. If you have to ask why, don't bother - just use JPG's. One use of the larger file sizes would be shooting several coins together as a group. This will produce very nice images that can be separated later or viewed as a whole.

There is one major advantage to using any camera (digital or film) over imaging directly with a flatbed scanner. With a camera, the photographer can adjust the lights and the angle of the coin to make a desired difference in the photo. A direct scan is somewhat like a photo produced by the kind folks who issue driver's licenses; it looks like you but you prefer not to think about that. Sometimes a minor adjustment of light angle can make a major change in the modeling on the portrait or throw needed light into a hidden part of the design. Our example here is a heavily test cut Athenian tetradrachm. By shooting from a slight angle and placing the light so it falls to the bottom of the large test cut, we have documented a coin that suffered greatly but revealed a heart of pure silver. Note, however, the short test cut across the body of the owl has a dark tone. This is a poor photo since it leaves doubt about the reason for this darkness. Is it a copper core, dirt in the cut or just the work of an amateur photographer? The answer is the last. The same coin is shown in an excellent direct scan on my Grading page. I did not claim all of my digital images were good photos; some are merely educational. Some coins will require several attempts to get the best possible photo. Even beautiful coins can take ugly photos and, with skill and luck, some ugly coins can be made to look presentable. Professional portrait photographers make a living making plain people look like movie stars; Paparazzi make a living making movie stars look ordinary or worse. What can be done with people photography can be done with coins (except for the part about making a living at it). Few of my coins are fashion model beauties but some of my photos are. Learn to read photos and decide if what you see in an image is due to the coin itself or to the skill level of the photographer. This is a good skill to develop if you intend to buy coins from illustrated lists or on-line sellers.

Earlier we noted that the camera can produce an image too large for practical web use. The inset here shows the full size image of the ear of Nero taken from the 2048x1536 pixel (3.34 Megabyte) image. First look on screen of the full image is disappointing. The metal is shown granular and contrasty. When reduced to a more reasonable size for viewing on a screen, the surface smoothes into a pleasing rendition of the coin. The point here is that we must be careful to make the image fit its intended use. For selling coins on eBay, the full image would be horrid. The microscopic metal detail would be read by most viewers (those who did not give up before the whole thing loaded!) as a rough surfaced coin. Few ancient coins can stand up to such microscopic scrutiny. No coin should be posted at a size larger than what will fit on one screen. Since many people are still using monitors limited to 640x480 pixels, I recommend never posting a combined (obverse and reverse!) image larger than this. Better yet, 600x300 leaves a little room for borders on the page. Even this is too large for 99% of coin photos. Most successful sellers seem to use 300x150 or 200x100 images. Smaller than this makes me think the photographer is hiding something. Having the larger images available for printing and cropping is wonderful but I can not repeat too often the importance of using reason in selecting sizes. Consider download time; not everyone has a T1 connection. Consider at what point the image gives the same visual effect you get looking at the coin itself.

Compression comparisons

Compressed 25 9.6kb	Compressed 50 7.1kb	Compressed 75 4.9kb	Compressed 90 3.1kb	Compressed 99 2.1kb

While we are on the subject of keeping file sizes small and practical for eBay or other web use, it would be good to consider JPG image compression.

Compressed 80 - 9.0kb
JPG (or Jpeg) is the most often used image format on the web. Its strong feature is a "compression" process by which relatively insignificant portions of the image are thrown away making the overall file size much smaller than BMP, TIF or other uncompressed formats. The degree of this compression is user controlable in most graphics programs. I use and recommend Paint Shop Pro 7 and the rest of this paragraph will refer to compression as it is handled by that program. Unfortunately, all programs do not use the same scale to describe the levels of compression. Im PSP7, the compression number refers to the degree to which image points have been discarded. At '1' very little is discarded; at '99' almost all the detail has been lost. Other programs (including Photoshop) use scales reversing this order with smaller numbers being the worst quality and higher numbers the best. All that is important is that you determine (read instructions or by trial and error) which way your software works and don't be confused by experts that tell you how they work with other software. At first glance it would seem that we would want to use the least compression to keep the best possible quality on our photos. The fact is that low compressions produce such large file sizes that practical matters of load time, file storage and data transfer limits (the one that locks up my Geocities site every day) forces us to consider compression as a necessary evil. The above row of images illustrates five different levels of compression. At the far left is compression 25, the setting I have most often used on this site. Oddly, it appears no better (worse?) than the next, level 50, image which is 25% smaller in size (therefore, faster loading and easier on my data transfer quota). The center image, level 75, is about half the size of the first but still appears nearly as good. The final two images, 90 and 99, do show terrible degradation but could be used for some purposes. . Our larger image (2x linear, 4x the area of the above) illustrates an interesting point. Compressed to level 80, it comsumes 9.0kb or slightly less than the level 25, smaller image in the row above. Which image makes better use of the space consumed? You choose. The purpose of this paragraph is not to tell you how to use you alloted kilobytes but to make you aware that there are choices to be made. To improve the efficiency of this page, I have recompressed several of the images to greater degrees than previously. I hope this will allow more visitors to see it before my data transfer limit is reached without leaving the impression that I have posted ugly images.

High relief Greek silver like this Philetairos portrait of Pergamon by Eumenes II presents a special challenge to the photographer. The high relief presents such a variety of angles that it is next to impossible to avoid reflections that burn out the highlights in the photo. For the most part, the answer is to place these highlights where they will do the least harm. Special lighting techniques and polarizing filters can be used to lessen such problems but these often leave the surfaces looking unnatural as if the coin were made of something other than metal. Further investigations in how to cope with coins like this will have to wait for later updates to this page. For now, it will suffice to note that the lighting that looks best on one side of the coin may be wholly inappropriate for the other. I wish I could tell you a formula that will allow the proper placement of lights to give the best results with every coin. Unfortunately each subject will require looking at the image in the viewfinder and 'reading' the way light plays on the devices.

Lighting for digital photography is like lighting for any photography. The two images above illustrate the same coin (a bronze of Alexander the Great) shot with the same camera. The left image is the result of a light arrangement like that used for most of the coins on this page. The right image resulted from lighting diffused with a white plastic cup used as a 'tent' to provide less directional illumination. This picked up the patina and played down the high relief of the coin. The same lighting on a very worn or low relief coin would show little detail but proves pleasing for this particular coin. The point here is that the digital photographer needs to look at each coin as an individual subject and apply the same 'bag of tricks' used for small object photography before the invention of the computer. There is no single 'best' lighting for every subject! Which photo looks the most like the coin? That depends on what light you use when looking at the coin.

A denarius of Septimius Severus is shown here with three very different lightings. The top image (with inset of the eye) was the best I could do with the coin close enough to the camera to fill the frame. It suffers from burned highlights on the high relief parts of the portrait. Moving the camera further away (and resizing the images to match) allowed the lighting seen on the lower left image. The light was moved very near the camera just to the point beyond which a shadow of the light would have fallen on the coin. This results in a digital version of the pseudo-axial lighting discussed on my photo page. The brightly cleaned fields here reflect back to the camera a bit more than I consider desirable but the modeling on the legends and devices is nice. Note how the minor surface flaw in the right field is emphasized severely. The lower right photo was taken by bouncing light off a circular reflector placed around the lens with no direct light falling on the coin. This produced an image with light from every direction (or from no direction?) that shows the metallic nature of the surface more than the engraved details of the coin design. This last is an interesting experiment but will produce the finest image of very few coins. Which of the three do you consider the best image? My answer is 'none'. This coin is a work in progress. How, when and if I succeed in getting a great image of this coin remains to be seen. At least with a digital camera I will not be spending more on film for the reshoots than I did for the coin.

There is a feature of the camera which will allow filling the frame with a coin and maintaining a greater camera distance. This, however come at a great price in terms of image quality. Zoom lenses on digital cameras are rated in terms of the difference between their focal lengths at wide angle and telephoto. The Nikon 990 has a '3x' optical zoom meaning the telephoto has three time the focal length of the wide angle. The camera also has a rating of '4x' digital zoom. When used, digital zoom will process the central portion of the image to enlarge it to fill the entire frame. The camera program fills in additional pixels needed to give the general effect of the increased telephoto power but there is actually no increase in detail. I do not like digital zooms. The quality obtained is not substantially different than just shooting the coin from a greater distance and cropping to the smaller image. Our example shows the same coin discussed in the preceding paragraph with more pleasing lighting (made possible by the additional room to arrange lights). The sharpness and detail is disappointing. The quality is acceptable at this 150 pixel square image but fails, in my opinion, at larger sizes.

The Nikon 990 lacks a threaded socket for a cable release that would make it easy to take close photos without jarring the camera and blurring the images. Nikon lists (but I have not seen) a remote release accessory that will trip the shutter via the USB connection. It is also possible to release the shutter using the self timer allowing the vibrations to subside before the photo is taken. This, however, requires you to reset the timer and macro options each time you take a picture. The camera needs a cable release socket. That a major company like Nikon would put the design effort into this level camera that is so obvious at every turn and still omit this basic need is truly strange. Our photo of an as of Vespasian shows a slight trace of blur from camera motion. This is easily seen on the full size inset where round dots of light appear doubled or as diagonal stripes. In this case, the amount of blur was slight enough not to be a problem once the image was reduced to 'web friendly' size. After all, I didn't want to post a really terrible photo here so I used this one that is only a little terrible. ;) Note that a completely different version of unsharpness is also seen on this image. The image was focused slightly in front of Vespasian's nose. Depth of field from the use of a small aperture (manually controllable on this camera) extended the area of sharpness nearly from the front edge to the back of the portrait but not to the rear edge of the coin. This would have required a much smaller aperture than is available on this camera or the swings and tilts of a view camera. We need to remember that not everything is possible and asking a camera of this level to take this photo was pushing matters considerably. The fact that this camera allows the user to select the spot on which the camera will autofocus makes this sort of image possible and easy. Many consumer digitals would have selected the wrong plane of focus and would not allow the user to intervene.

To use the camera for coin photography I constructed a simple stand to hold the camera mounted on the upright by means of a mini-tripod ball and socket mount. A series of upright dowel rods set in wood blocks hold coins an appropriate distance from the camera. Photographing coins on dowel rod posts allows lighting flexibility and shadow free backgrounds thrown out of focus. Groups of coins can be shot resting directly on the base. Light is provided by a set of high intensity desk lamps but photo flood bulbs or other types of light could be used. The angle of the light suggested a lens hood should be used to prevent flare. One was fashioned out of a cut down plastic 35mm film can (example at lower left of photo) which just happens to fit the outer diameter of the lens. The camera receives power from an AC adapter and can be connected to a television (or VCR) allowing use of a much enlarged viewfinder.

In the camera's manual mode, everything is controllable by the photographer. A tour of the viewfinder screen information beginning in the lower left and going counterclockwise:

A rectangle with a central spot shows the camera is in spot metering mode. This forces the light meter to read only the central area indicated by the large white rectangle.

'A' indicates aperture priority mode when the photographer sets the f/stop and the camera selects the proper shutter speed. In this case the camera has selected (yellow) 1 second to match my f/7.5.

'+/- -0.7 indicates that I told the camera to underexpose the image by 2/3 stop. This setting I find gives better results of bronze coins. Silver is shot with no correction to the meter setting.

[ 20] indicates that the card in the camera can hold 20 more images at the current size and quality settings. 'Normal' quality setting is displayed but 'Full' image size is assumed since there is no display of one of the smaller size indicators that would be just above 'Normal'. Perhaps this is the place to mention that there are several other features and controls available on the camera that do not show on this screen since they are not in use here. The whole display can be turned off so all you see is the coin if this is preferred.

A small light bulb shows the camera is set for incandescent white balance. This would have to be changed if the rig is used outdoors in daylight or with electronic flash.

A crossed out lightning bolt in the upper right corner indicates the built in flash is turned off. A tulip just to its left indicates the camera is set to focus in macro mode.

The central rectangle that indicates the area the meter is reading is surrounded by a set of 5 brackets that are user selected (red = current selection) to tell the camera where it should focus. For most coins this is the center but some of the angled shots shown on this page required this be changed.

This seems a bit complex but all of these features can be set to allow the camera to select everything for the photographer according to a program. This makes the camera a 'point and shoot' suitable for use taking photos of things other than little coins. Not all digital cameras are this complex but allowing the advanced user to control everything insures the capability of doing the job whatever that job might be. It is not so complicated that the average user will not be able to take decent photos within a few minutes of opening the box. Neither is it so simple that a user will have full command of all features for quite some time! The idea of this page is not to scare off those who might want to use a digital camera to photograph coins but to point out that these cameras will allow you to grow in sophistication and make the best images of which YOU are capable. Near the top of this page we noted that few bad photos are made bad by bad equipment; usually the problem is a bad photographer. The camera is capable. Now the question is: 'Am I?'

Back to Main page