Measuring Coins

[Mits]

Apart from cirfumfarance and thickness ... specifically on Indian coins there is a need to measure letter size, emblems etc.

What is the best way to do the same?

[FosseWay]

Most of my collection consists of machine-struck coins and tokens so the following answer may not be entirely relevant to hammered coinage.

Personally I find a relative description of the varieties in question together with good photographs of both/all much more useful than measurements in isolation.

Take this thread posted today by Globetrotter (though you might equally take any of countless similar threads of his). From a measurement perspective the deciding factor is the distance between the edge and the portrait, but IMV it's much more instructive to know to look for "relatively more" or "relatively less" space (or alternatively relatively larger or smaller portraits) alongside pictures of both, than it is to know that one is 1,5 mm and the other 0,8 mm from the edge.

Using calipers is an inexact procedure if you can't get the object between the calipers, and even more so if you're concerned about scratching the surface of the coin and therefore try to hold the calipers away from the surface. If the difference between two varieties is say 0,5 mm or more, and the varieties are reliably one size or another and not on a scale between the two, simply holding a ruler up against the coin will probably do.

Another alternative is to scan the coin at a high enough resolution (would recommend at least 300 dpi, preferably 600) to allow the degree of accuracy you require. You can use a measured background, but this may not be that useful if the feature you're measuring is distant from the edge of the coin. However, if you open the scan in an image manipulation program that (a) has marginal rulers and (b) has the option of quoting measurements in standard linear measure (mm or inches rather than pixels), you can accurately measure any feature.

1. Open the scan in a suitable program (e.g. Adobe Photoshop).
2. Rotate the scan if necessary so that the feature is horizontal or vertical - you can't draw guides at any other angle AFAIK.
3. Magnify the view so that the rulers are at the required level of accuracy (e.g. 0,1 mm).
4. Place a guide at the nearest ruler "tick" (nearest 0,1 mm if that's your chosen accuracy) to one end of the feature you're measuring.
5. Move the scan so that the end of the feature rests precisely on the guide you've placed.
6. Place another guide at the other end of the feature.
7. Read off the number of mm between the guides - you should be able at least to get an accurate reading to your chosen order of magnitude, possibly greater depending on the resolution of your scan.

[Mits]

FosseWay

Thank you very much for this detailed explanation.

Regards
Mitesh Shah

[Pabitra]

Another alternative is to scan the coin at a high enough resolution (would recommend at least 300 dpi, preferably 600) to allow the degree of accuracy you require.
.
.
.
(b) has the option of quoting measurements in standard linear measure (mm or inches rather than pixels), you can accurately measure any feature.

If I know correctly, pixels and dpi ( dots per inch ) are same measure except they are used in context of display and scanner respectively.
If you see an image on a screen and the cursor position is shown, its x and y coordinates are displayed in pixels. You point cursor at two different points, note down their coordinates in pixels, calculate the distance using Pythagoras formula and then use dpi detail to calculate the actual distance in inches. Convert them to cm or mm as per your requirement.
Kindly correct me if I am wrong.

[FosseWay]

That sounds entirely logical. But it saves a lot of faffing around, not to mention scope for error, if you can get the software to tell you the distance in millimetres in the first place.

[Figleaf]

Some electronic microscopes come with software that allows exact measurements. Don't have that, so cannot comment any further, but maybe someone else has it.

Peter

[Pabitra]

That sounds entirely logical. But it saves a lot of faffing around, not to mention scope for error

You are absolutely right that if computer can do it, why should you do it.
The absence of such software should not deter in doing the desired job.
For measuring either vertical or horizontal distance, one has to just keep the concerned co-ordinate same. Then the exercise goes down just one subtraction and one division operation.

The detailed operation is required only when distance is neither of the above and then too, it is just a 4X2 cell spreadsheet and it is just a click away on a standard computer.

Anyway, I concede that you are right since I assumed certain math ability from the user.