Coil Stamp Authentication Is Too Hard

The rotary coils are all good, there is nothing to fake them from. However I suspect that all the flat plate pairs will be no good.

Sounds like "If you are going to collect W/F coils, especially the flat plate coils. only buy ones that are certified".

Pretty much. I have heard multiple experts claim that 90% of all the flat plate coils that are out there are fakes. Even the inexpensive ones.

rev, that sounds about right. I've gotten lucky with some flat plates getting clean certs in the past and the two collections these come from have for the most part been better than most at accuracy. But I won't be surprised if what you say turns out to be true.

I recently tested the flat plate coil waters, sending three different pairs to PF for certification. Two of the pairs I had recently purchased with the caveat that they could be returned if certification failed. The third pair was one that I had purchased back at the end of 1982. Of the two recently acquired pairs, I thought one likely a fake and the other likely genuine. I held no opinion on my 1982 pair but did hope that, since no refund was possible, it would not be a fake. Results … both of the recent pairs were judged to be fakes and have since been returned & refunded. The pair acquired 37 years ago was the only one judged to be genuine.

One thing has become immediately clear to me. I've been focusing too much on the perfs. Trying to determine perfs that are too round and smooth has generally been my first look. This has also been my frustration. So many perfs seem - sort of ok - not clear enough for me to make a determination. I really should be looking at more basic things like the edge to edge size and rotary vs flat before I start worrying about perfs. A great example is the horizontal pair on the 5th row of the first page I scanned. The perfs look too good to be true, so I would write it off as a fake. But it is clear that this is a rotary printing (by comparing the design size etc) and that the height is within the acceptable range of 25mm for a horizontal coil. So actually it is probably legit. So before I get into precise examinations I should be able to eliminate a lot of fakes with simpler measurements and simplify my task. One of my goals in selling stamps was to get better educated. These kinds of exercises are definitely helping in that regard. Still much to do.

I'll look at the coils closely and if I think they look good, i'll list em on ebay. Problem for me is that there are sooooooo many low priced coils in collections I pick up that it's just not worth the time and effort to authenticate them and list them for a few $$s. What happens most is I just toss em into a "I don't feel like messing with these right now" box and they stay there.

I've even gone so far as staying away from collections that are loaded with them just to avoid the headache, unless they are just a very sparce and tiny part of the collection.

Inkman, I completely agree. I don't bother to look at low priced ones at all. This time I just have too many to ignore. I've been putting it off. Part of the reason I posted here is because now I feel like I have committed myself to doing the work - lol. If some higher value ones pass muster they can get certed and that should command a better price at the end of the day.

Out of 40 flat plate stamps, 12 survived my intense scrutiny. Only one perf 12 survived. A few are low value. A few didn't meet my criteria exactly, but I couldn't eliminate them. They have high enough catalog value that they will be worth sending into the PF to let them decide.

Here are the tests I performed:

Double check that they are flat plate

Measure and check perfs not too sharp

Check edges are parallel and not too sharp

Check edge to edge size is within .3mm (not .03 - I believe this is a typo from Bill Weiss - have found a couple sources that a greater tolerance is accepted by authenticators)

Check watermark (all are single line)

I made heavy use of the Stamp Smarter ID and "commonly faked" pages.

Here are the Scott numbers from left to right and top to bottom. Notes in parentheses are where things were imperfect, but not bad enough for me to eliminate them. Where used stamps are higher value I cannot confirm the cancellation. Those say "cancel?" in the notes.

If you have any suggestions on something I missed, please feel free. All except for a couple low value items (under $20) will go to the PF.

385, 413, 442 (cancel?), 446 (questionable watermark),

412 (edges too rough?), 390 (cancel?), 392 (cancel?), 395, 447 (parallel enough? cancel?)

442 (measures 22mm, cancel?), 445 (measures 24.5mm), 447

You are getting lots of advice on how to spot the fakes. . .I think the bigger deal is the process you need to go through. You are off to a good start by scanning them. I'd then break up the scans into individual hi-res images with indexed names - not presumed scott numbers. .names like 4c001.png, 4c002.png, etc. Then put each stamp on a dealer card with the index number and put a spreadsheet together with all the pertinent information: watermark, apparent perforation, paper (if you might have bluish), rotary vs flat plate, type, etc. Then I'd put columns for everything you are checking for authenticity: evidence of filed perfs, parallel perfs, evidence of trimmed perfs, parallel straight edges, etc. I'd put a column for apparent Scott number. For reference, I'd list the Scott numbers of the fakes they could be made from. Then I'd just start woking on the stamps one by one and filling out the spreadsheet. You can work on it as you like. Once you think you have (or might have) a genuine one, note that in the spreadsheet. You should send those out for certs. This is really an organization problem. . .it is very easy to get mixed up and forget what you did.

I doubt the pair is genuine. The holes are too round, and the perf holes line up exactly, which is all but impossible.

Philzilla, that's pretty much what I did. Quite a few hours went into this.

Rev, take a closer look at the holes on that pair:

The 4-cent perf 8 1/2 Scott 395 is genuine and does not need a certificate. However, it should be carefully checked for faults. One cent perf 12 vertical coils 348 or 385 are the most common perf 12 coils and have the best chance of being genuine. The other perf 12 coils are scarcer than their value suggests, perhaps because so many potential collectors unknowingly wind up with fakes in their collections. A lay down test with a couple of low value perf 12 sheet stamps should reveal whether the perforation holes are spaced correctly. Until Richard Kiusalas published his articles suggesting that English measurements were more useful than metric, some fakes were metric, exactly 12 perforations in two centimeters instead of the more accurate spacing of 0.066" center to center. Also, a lay down test should reveal subtle differences in perforation hole size or general appearance of the perforations.

Another footnote. An earlier post mentioned that one coil edge was rough and the other was not. Normally this occurs if the perforations were trimmed off one side of a straight edge. But this can also happen with early auto wound coils (single line watermark perf 8 1/2 and possibly 388 or 387) because the sheet was cut in half to create ten stamp wide rolls. Initially ten knives were used, leaving the roughly cut edge. Later an 11th knife was added to allow special coil sheet layouts to be used. This is really in the weeds but is sometimes overlooked even by specialists. Some understanding of the evolving coil production process is very helpful when expertizing flat plate coils.

One final note. Some of the perf 10 examples appear to be trimmed. The edges should be parallel to within less than 0.001" or 0.025 mm. Images in the group scan above are small and may correctly or incorrectly render uneven looking imperforate edges.

Also check watermarks. Fakes made from booklet pane singles will sport a vertical watermark (sideways to the design).

Thanks. The way I checked for parallel edges is by using a graphics program that can move sections around without rotating them So I get images like this below. It was interesting to me just how fuzzy the edges look at extreme magnification. The only one that I have any doubt about is the 447. It has a funny bit where both edges move out of alignment at the same time. I won't be shocked if it doesn't certify, but it is in that fuzzy zone and valuable enough I'll let the PF decide.

I still don't like them.