I’m following the procedure as used by John Kaupe in Peter van der Molen’s book “Swaziland Philately to 1968”. For this example I’ll be describing what I did in detail and then just do summaries for other twelve 1d stamps.
Working example 1:
Step 1: Perforation 12.5 x 12 – CORRECT
Step 2: Overprint measurement – checking the overall and all intermediate measurements and all are within the limits – CORRECT
Step 3: Nothing that stands out
Step 4: Match the perforation holes to the comb profile for the 1d which, for the 1d only one comb which makes it easier.
I again use my Photoshop skills to do this but the other way is to print large versions of the scans and match it with photocopies of the comb profiles that can be found in van der Molen’s book.
I make a selection of the perforation, paste it onto a scan of the comb profiles and then use an opacity filter of around 50% and resize my selection (keeping the same aspect ratio), to match the length of the comb scan. I firstly look for the obvious characteristics e.g. very wide or narrow tooth or severely raised or dropped perforation holes.
For this stamp the wide tooth after the 5th perforation hole stands out and also the dropped hole on the right. Only comb 8 matches this criteria and overlaying the image on the complete comb perf looks quite good. At this point I’m fairly confident that I’ve identified the right column.
I then proceed to check the side perfs and checking column 8 I just feel that it doesn’t match close enough. I go back to the horizontal line of the comb testing all 10 options and it just fits perfectly on column 1 and not on any other.
I then check the vertical perfs for column 1 and this time it matches perfectly. This stamp is definitely from Column 1.
Step 5: I now check the stamp and overprint again against the constant overprint and plate flaws identified by Kaupe for the stamps in column 1. In this column only the stamp in position 31 doesn’t have a constant flaw so my chances should be quite good to find something.
Checking for all possible flaws in this column I now notice the small dot in the overprint between the “l” and “a” in the position consistent with the top left stamp of the sheet i.e. position 1. In this position a small dot in the “d” is also listed as a constant flaw but it is missing from this stamp.
From all my analysis, in my opinion the stamp and overprint is genuine, the stamp is definitely from column 1 and I think from position 1, the missing spot in “d” leaves me a bit unsure though.
Main lesson for me from this example is that you have to verify the combs extremely accurate and it should match closely.