518 Colors

Here are 3 "ugly" stamps. It would be nice to be able to fill the 518b hole in my album. Obviously it is near impossible to determine color accurately across the Interwebs, but I'm curious to hear your opinions.

All are perf 11 so the possible color varieties are:
1. dull to dark violet brown
2. violet black
3. deep brown



Here are enlarged areas of each stamp that might help the comparison:

Attached is a scan showing a 518, a certified 518b and 518. A darker brown shade of 518b also exists.

Clark

Thank you Clark. Compared to your examples, it seems that mine all have some violet/too much red in their color to be a 518b. Your scan does a great job of comparing the color difference of the 518b.

My "A" could be the darker brown variant, but seems nearly impossible to determine without a side by side comparison.

Any and all opinions are welcome.

Clark, is there and established date range for when the 518b color was produced and used? Also, in searching Seigel and the PF data bases, I'm seeing somewhere in Ohio (several) and a pre-cancel in St. Louis for uses--are there other known towns/cities where 518b was predominately used?

Here's a link to the dark brown "subshade" of Scott 518b as shown in a 2012 Siegel Auction:

http://www.siegelauctions.com/dynam...2020,%202012

Note their comment:

Al E. Gator, one of the Siegel examples is noted to have a cancel from Rochester, NY.

I'm having trouble posting a link to it. Use a Siegel power search for 518b and the keyword "Rochester."

It looks completely different than mine. I'm not even sure if this Siegel example is a cancel or precancel.

I've been through both Seigel's sight as well as the PF cert. data base. So.. St. Louis, Rochester and some place in Ohio as of now for known uses and Johl says "early in 1917"--I guess that rules out an extended period of use?. The color as shown in Seigel and the PF's data base(at least on my monitor) are too helpful. I'm probably like most of us in that its a bit difficult to determine that color difference. I was just wondering if it were possible, as it has been for other classic issues, to narrow down a time frame and known locations. I have 518s singles and multiples used on bank tags all in the early 1920s, as well as off-cover examples, most of which have ample violet in them to know they are 518s but a few that are a little more difficult to be sure about. If there is a known time frame and usage locations, it would help me determine if those I have are worth continuing to ponder. Thanks a lot for your help!

I just realized that the "Johl" that Siegel refers to a set of books now out of copyright and available on Siegel's site:

http://www.siegelauctions.com/enc/pdf/

Links to the 4 files:
http://www.siegelauctions.com/enc/p...hl_Vol_1.pdf
http://www.siegelauctions.com/enc/p...hl_Vol_2.pdf
http://www.siegelauctions.com/enc/p...hl_Vol_3.pdf
http://www.siegelauctions.com/enc/p...hl_Vol_4.pdf

They are huge pdf's and the 518 is located in volume 1. Apparently the 518b designation wasn't listed in Scott at the time.

"There is, however" a very scarce early shade which should not be confused with later printing. This is a very dark brown, almost black, with just a suggestion of the violet. It is listed as black brown. There was also another dark brown shade which is also quite scarce, this is lighter in general appearance than the other shade and has much more of a violet cast, it is not noticeable under artificial lisht. This latter shade is known as dark violet brown. The later and ordinary color is violet brown, the tone checking with the previous issues."

"Plate used: Imprint, "A" and number. 5782."

I just took my 3 for a walk outside. In natural light, the differences between them are very easy to see and look different from my scan. My C seems to be similar to the Johl description for the dark brown variety. However, A and B now look like the 518b in Clark's scan. Surely I didn't just stumble upon 3 518bs. Darn this non-color corrected computing equipment. Three Bill Weiss $5 opinions will be worth it just to know what does not qualify as a 518b.

I've done the walk outside---multiple times, bought and used an OttLite, Used magnification ---everything but a Ouija Board. As with you, I doubt that I've got one--or more--so ya, I guess its time to sent one to Bill. I print off the pages from the Johl book--not all that helpful.

For funsies, I did a similar "enlargement" of the same areas for Clark's examples. Anyone have recommendations for lower end scanners with decent color calibration? That is without using a reference card and adjusting the color space of your photo editing software. It seems that a scanner's light source greatly affects it's color accuracy. Clark's scanner seems much better at this than mine.

The scan was made with an older CanoScan 9900F in 2006. I purchased the scanner when the model first appeared, probably in 2002. I had to replace the scanner with a CanoScan 9000F last year when I retired the last Windows Server 2003 system. Cannon refused to provide drivers for Windows Vista, Windows 7 and later versions of Windows Server. I believe the new scanner is a little more accurate with color. Both models are "photo" scanners capable of handling paper documents, negatives and color transparencies (slides). Unfortunately, the newer Canon software is a dumbed down version of the older software. I found that 64 bit Photoshop does not support TRAIN making scanning very inconvenient. The workaround for now is to continue to use the ancient 32 bit Adobe Photoshop 7.0.1 to manage and process scans. The new scanner is much faster, but it may not work as well scanning coins. (Scan coins upside down for better appearance.)

For stamps, it is helpful to check the levels and expand the tonal range setting the white and black points at the edges of the brightness histogram actually used. The following tutorial may be useful to Photoshop users:

http://www.cambridgeincolour.com/tu...s/levels.htm

While the scanner may not be entirely accurate, scanning a color reference and the stamp of interest in the same scan with the same black or 18 percent neutral gray background will be helpful. Viewing colors is best done with a 18% neutral gray background.

Unfortunately, 518b is a stamp that cannot be easily be verified without using a color reference. Two shades are said to exist. My reference is the lighter shade.

Clark

Let me put in a good word for the CanoScan LIDE210 letter size scanner. For $100 it gives terrific bang for the buck. The software is made for dummies, but it grabs an image with a nice set of options. I don't edit with it. I use Photoshop CS5 to tailor the image to my needs for size, color correction, whatever. But the combo is very consistent and reliable, and when the scanner eventually dies, it won't hurt to replace it. The thing is, the choice of scanner is only half the matter. The other half is your editing software, and not just the free stuff that comes bundled with the hardware.

My scanner is an all-in-one type from 2009. Despite age difference, your 2002 photo type scanner may have better color accuracy. I'll pick up a photo type at some point.

Your link is quite useful. It's quick and easy to set the black and white points to get a better picture. It had a barely perceptible effect on the coloration of my scan. However it did provide better contrast. Setting the contrast with a dedicated slider can be inaccurate. The histogram makes it easy to set without loosing information.

The TWAIN issue in 64 bit versions seems quite annoying. I sat confused looking at a disambiguation list for "TRAIN" until I finally realized what you meant to say.

Would you describe your left stamp as Violet Black and your right one as Dark Violet Brown? It won't serve as a reference, but I am curious.

For comparing with a 18% neutral gray background, do you use a physical card behind the stamps or do you crop the stamps to their designs and create a 18% neutral gray layer beneath them? For video production color editing, I found it interesting that there is benefit to painting the whole room in that tone.

Thank You,
Ryan

Paint.net seems adequate when better software isn't available. It took me a while to realize that the bundled software was making adjustments without me opting for them.

There is a good third party scanning app, named VueScan. Not only does the interface make scanning batches easy, it often allows you to identify and use older scanner even when the manufacturer does not supply a driver.

One last note, it is critical that folks scan with the scanner cover closed for the most accurate imaging and color. Ambient light plays a big factor in how the software and hardware render the resulting image. This is easy to test, simply make two scans (one with cover open and one with cover closed). The scanner and software will try to accommodate and account for the ambient light coming into the scanner bed if the cover is open. Scanner manufacturer go to the extra cost of putting covers on the device for a reason (to deliver consistent results).