• entry
    1
  • comments
    5
  • views
    83

DISCOVERING VARIETIES USING PHOTOSHOP TO COMPARE COIN IMAGES

1 1
FAS_Coins

391 views

 

DISCOVERING VARIETIES USING PHOTOSHOP TO COMPARE COIN IMAGES

Using Adobe PhotoShop (PS) we can compare images of the same date to view any possible differences and thus establish the variants.

I will be using the Venezuela 5B (1879 to 1936), my favorite coin collection, for these comparisons.

image.png.98600a0ec330aafd9146d0db834d3494.png
1936 was the last date the Venezuela 5B (FUERTE) was issued.

 

 To compare two coins we must first align them and make them the same size. For alignment I use the top of of the shield. Before resizing, crop the image to the borders of the coin. Resize all images to the size of the largest.

image.png.12274c0e26ebc14a39c301d6e1397e5c.png

There are two methods that I use for comparing images. The first method involves placing one image on top of the other and making the top image around 50% translucent. The second method requires more work but I feel is worth it: Make a “mask” of each coin in different colors and the compare both “masks” one on top of the other. Both methods will be clear once we see the finalized product. 

In the case of the Venezuela 5B all but one of the varieties known involve the date so we will concentrate there.

Here is a combination of both methods. A red “mask” was made of a coin graded with the variant “9 HIGH” and a green one for “LOW 9”. Both images were superimposed and the top image was made partially translucent. More examples will be posted as I get the work completed.

image.png.68c37148c2a06c818d9aa38a68190cba.png

 By closely looking at the second 9 it is easy to see that the red 9 in higher and more to the right of the green nine.

image.png.921dad8cf800f83bc4521d91dcd0b4fc.png 

Looking at only the masks is even more obvious. The stars are not part of the date and thus should stay in place always and serve as a guide that we are doing a good job if they match.

image.png.9e87f9af15b796478d9eb68d58f28ee6.png

The following example is the result of superimposing only the two masks of 1926 coins (NGC SN included). As you can see the stars on each side align perfectly. In the case of the year 1926 NGC does not recognize any variants even thogh these two coins are clearly different.

image.png.196259e9eb1ef955d66103612b5e5d38.png

 

NGC ACCEPTED  VARIETIES

NGC recognizes varieties for the following years:

1886

1888

1902

1910

1911

1912

1921

1924

1936 Not attributed any longer by NGC.

 

THE 1921 VARIETIES

I made a mask in red for the 1921 "Narrow Date" and pasted it on top of a coin with no variety attributed. It is clearly a "Wide Date". The major differences are noted in green.

image.png.239d846abdaa66b76051c555bc2d9cfc.png

 

THE 1910 VARIETIES

The are two varieties recognized for 1910, “ROUND 0” AND “OVAL 0” 

image.png.4c2cd9f849c34e4fa92882f7600202d7.png

It’s easy to see the difference in this comparison. Not only is the “ROUND 0” wider but the center of the zeros are very different.

image.png.7c381d24cc935b0a2fdb22a158561781.png

 

THE 1924 VARIETIES

NGC recognizes 4 varieties:

WIDE DATE, NARROW DATE, LOW 9 and HIGH 9.

 

WIDE DATE (RED) / NARROW DATE (GREEN)

The difference is quite obvious

image.png.e936f56dc7fc6b15ab6edc82838024fe.png

image.png.3a669e02af40357147ed98bf84019faa.png

 

LOW 9 & HIGH 9

In the following photo we compare an NGC coin graded “LOW 9” in blue to “NARROW” in

orange and “Wide” in black.

image.png.052f13fb60e394f6ab98ab970db8c7cf.png

image.png.6de72341f40e6a0582c20b4dab0a30e6.png

The “LOW 9” is about as wide as the “NARROW” and all number except the “1” are lower. I did not find a coin graded “HIGH 9” but it is possible that such a coin is no other than the “NARROW” which as we mentioned is higher.

 

Is there another variety for 1924?

Let’s compare two coins.

The first coin is an NGC coin graded AU DETAILS “LOW 9”.

image.png.3289a51083bc42cf0246136d08529d4b.png

The second coin is a PCGS MS 63 with no variety attributed.

image.png.59fadaed27d7314c91a02a114af00644.png

The third image is a comparison of both coins.

image.png.3219a51635eb69de3977ef5a74a57d82.png

As can be clearly seen the PCGS coin is very similar to the “LOW 9” however the NGC coin is clearly wider which is very easy to spot when you look at the “4” The PCGS coin is clearly different from the other varieties “WIDE” and “NARROW” and it’s obviously not the missing “HIGH 9”.

 

THE 1936 (NOT) VARIETIES

In the past NGC has recognized three 1936 varieties: “HIGH 3”, “NORMAL”, “LOW 3”. Let’s take at look at them.

In the following image are two superimposed “masks” of the “HIGH 3” and “NORMAL” varieties. They are identical.

image.png.ae890269dd0af3658b6f35400b75e0a6.png

This is a side by side comparison of the “HIGH 3” and “LOW 3” varieties.

image.png.3e30a3d09574e59c1ceffc0cee104967.png

This is a top on top comparison of the “HIGH 3” and “LOW 3” varieties with the top image being 50% translucent.

image.png.baaa5fc500781d921c4725b46b606d51.png

Can’t see a difference, neither can I!

When I asked NGC, on June 5th. 2020, about the 1936 varieties, the answer I got was “We will not do any attributions for 1936” which I thought was the correct and honest answer.

Here is additional proof of the non existence of the 1936 varieties. The same "mask" has been applied to all coins.

image.png.e95b58ca769826ea6eb46b4af84a9bcf.png

image.png.c065038be632cffadbf59adcf4624c54.png

image.png.f0659bee132ac1b4bb8f95b6f538d09a.png

image.png.c663334f3ab8e724bb2cc4e5229511db.png

image.png.d47a0c3dbf98c3fd6d7aaa0c73f6006c.png

There is an additional image for another LOW 3 but I ran out of uploading space.

THE 1911 VARIETIES (corrected on 06/24/2020)

The image below is composed of  “NARROW” mask and a “WIDE” mask superimposed on top of a “NORMAL” coin so that all three varieties are visible.

The only “NARROW DATE” I found was a PCGS SN 80554128 MS 64. By the way, PCGS has some great images I have used for my research, NGC images are not even close.

image.png.20b2b2f3d7002d9f2f45a507f9cc09fe.png

The following coins graded by NGC as “NARROW” I found to be  “NORMAL”

2669687-013 XF 45
2669687-008 VF 35 
2834210-002 VF 30

 

THE 1912 VARIETIES

The difference between the “NARROW” and “WIDE” DATES is quite obvious. The “WIDE” DATE is represented by the green “mask” and the “NARROW” DATE is below. I looked at the only “NORMAL” DATE that I could find but it was a VG 10 coin and I was not able to notice any difference, I understand that the “NORMAL” DATE is only very slightly wider than the “NARROW” DATE.

image.png.0dfa605b2b545a14fb76bbe2fdfd9f3b.png

 

THE 1886 VARIETIES

 

NGC recognizes four varieties for 1886:

NORMAL DATE
SECOND 8 LOW
TIGHT 8s
NO ACCENT

I first made a green “mask” of an NGC “SECOND 8 LOW” graded coin and applied it to all coins for easy comparison.

Below is a “NORMAL DATE” with the green “mask” on top.

image.png.64ba6bcc4d1058cd8eece719cdb367f5.png

The next image is of the “SECOND 8 LOW” with it’s own “mask”

image.png.7e9448ea424c5e98388401e9c2aecb26.png

The last image is of “TIGHT 8s”

image.png.24b158c8371615f6f7be82c10be64418.png

No image is provided for “NO ACCENT” since that is easily observable in the word BOLÍVAR (BOLIVAR) in the obverse.

 

THE 1902 VARIETIES

 

There are two varieties for 1902, “WIDE” and “NARROW” dates.

 

The image below is a “mask” made with a “WIDE” date and superimposed on a “NARROW” dates. The main difference is marked on bottom right of number “2”.

image.png.c30eea1e54852f16b77f03aef6cb082a.png

 

THE 1888 VARIETIES (CORRECTED ON JULY 3, 2020)

 

The first image is of a coin graded “SECOND 8 HIGH” with a green “MASK” outlining the date.

image.png.17c47bdf049e7910fcd87b500ab4072d.png

The second image is of a coin with no variety assigned but differs from the “SECOND 8 HIGH” as can be seen when the same “mask” is applied to it. NGC recognizes only two varieties so I will assume that this one is “SECOND 8 LOW”.

image.png.dab3fe1e92c2feff065af11bc886229c.png

 

THE 1926 VARIETIES NOT RECOGNIZED BY NGC

 

I will present the case for the existence of four clearly different varieties for 1926 which are not recognized by NGC.

According to numismatica.info.ve/en, a widely respected source, there can be as many as seven varieties. 

I will present the ones I found and can show with photographs.

image.png.974a3f440d15c3cc989edd7fb0362395.png

image.png.22913609402ee4c9b62c38ee09308151.png

image.png.2d73a40826edddccd00b2c42072e187a.png

image.png.4c572505a67c0825ad25b357546f1524.png

Here are the four varieties with their "mask" superimposed

image.png.9dc382ea83f2e04b06a9def5c8cd3c95.png

I separated the four varieties into two groups for a clearer image.

image.png.63bb1fde7094478e30d32e1031f0b41b.png

image.png.4358de015bcdc04cb322e9698aa74ad4.png

 

 

The job has now been completed with the description of each variety. 

Below you will find all necessary "masks" you will need to identify the Venezuela 5B varieties. You need to straighten each coin using the top of the seal as a reference, crop the image to a square that barely touches the four side of the coin and make the coin the size of the mask provided. Next you will open the file containing your coin and the file with the mask that matches. The mask will be above the coin image and since it is transparent you will see your coin below. In general only one mask is necessary to identify all varieties since all varieties have been identified and you will see that if the "mask" doesn't match then it's not that variety it's the other one. It's impossible that images will match exactly since there is necessarily some small errors when doing all the needed straightening and cropping, however you should be able to move the mask around to make it fit the necessary space. Making the side stars match is a good starting choice.

image.png.21109db10fcdeb831970cb33039816f1.png

image.png.c928f0722283f97cb8c46aaa00a4452d.png

image.png.4be28a1c1146f02169c5dfc188ef1260.png

image.png.4473919b26bc7b22a6e56d0e9d495a46.png

image.png.b6646f885243b51d21a5afc4257ab60e.png

image.png.2e93ae139fa965c37d0248c4061cbc2d.png

image.png.768aa4b71cff062b88011f34fea419bc.png

image.png.8448cf15e107247357b4e16ef68e9947.png

image.png.1e36503306b7ae729b6ba3fea710d6f3.png

image.png.066b258bad0864d3a87343ebdbb64550.png

 

 

 

I have not been able to remove the image below.

image.png

1 1


5 Comments


Recommended Comments

This is interesting.

A couple of questions: How were the dates punched into the hub or die? In other words, was the complete date punched into each die, or into the working hub, or into the master die, etc? Or was it the last two digits, as in earlier US coins? Or, something else?

And, since I have never used Photoshop, this next question may be ignorant, but is there any possibility that cropping, resizing or other manipulation of the image might cause distortion? Obviously, the "high 9" is a known variety, so I am not implying  that the misalignment was caused by the manipulation. I am just wondering exactly how accurate the photos are once the images are all sized the same. If it is completely accurate, this seems like a really good method of comparing like coins.

Share this comment


Link to comment
55 minutes ago, Just Bob said:

And, since I have never used Photoshop, this next question may be ignorant, but is there any possibility that cropping, resizing or other manipulation of the image might cause distortion? Obviously, the "high 9" is a known variety, so I am not implying  that the misalignment was caused by the manipulation. I am just wondering exactly how accurate the photos are once the images are all sized the same. If it is completely accurate, this seems like a really good method of comparing like coins.

The main challenge to my mind would be getting 2 very consistently made images of the coins and getting them properly rotated and aligned so that any differences you're seeing area actually differences in the coins and not in the image creation / editing process. This is more a camera / image creation issue than a photoshop issue.

It's going to be very difficult to so this with a camera unless you have the camera on a tripod, have the camera on a wireless trigger so you don't touch it at all after you set it up and you lock the focus ring into manual only and you'd want to have things set up with the placement of the coin / slab so that you place both coins in the exact same place, preferably at or near a perfect right angle to the lens of the camera. If you don't do all that just right, with the small size of the features you're looking at, you could easily be seeing distortions and not real differences. You'd probably want to keep the coin just towards the center of the camera / image frame to minimize barrel distortions and other optical distortions that happen in the lens.

Alternatively, you could just scan the coins on a flatbed scanner and hope for the best.

Share this comment


Link to comment

Hi! I don´t have an answer for the die question. Regarding Photoshop I can assure you that the program creates no distortion when resizing. With regard to the distortion when taking the picture I agree one has to be extremely careful. Up to now I´ve worked with photos obtained from third sources, grading and auction companies. I´ve been stuck in Madrid for 5 months, when I get back home I´ll take photos of my own coins. Regarding the validity of the method, I believe, if you do this with extreme care it is surely better than comparing two coins with a loupe.

Share this comment


Link to comment
18 hours ago, Revenant said:

The main challenge to my mind would be getting 2 very consistently made images of the coins and getting them properly rotated and aligned so that any differences you're seeing area actually differences in the coins and not in the image creation / editing process. This is more a camera / image creation issue than a photoshop issue.

You can get a pretty good idea if the images are correct because once you compare them everything matches except the dates (variants).

Share this comment


Link to comment

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now