Removing Cancellations...

Faken and others,you are absolutly right...As you see by the picture..I took a Queen stamps and placed a red dot under 1973, the with a permanent magic marker covered the red dot and guess you are right...BUT...what is the arrow pointing at...hmmmm....

Wert,
Sigh. Please understand, this is about pixels. This is no different than anti-aliasing, when a graphics program uses algorithms to 'smooth' out lines. It is software code that simply 'averages' based upon pixels.

Here is the same image you just posted zoomed in to pixel level. The software does not see under anything; it is simply looking at the pixels involved. (click for larger image).


Each pixel here is various shades of black, but only those right over the original spot are close to 0,0,0,0 (true black).

The darkest pixels are what the software uncovered. You simply pressed harder around the original spot to make sure you covered it completely. But in doing so you added a darker area that the software picked up. If you go back and darken a spot lower in the blacked area, it too will 'show up' after you run it through the software.
don

51studebaker..I consider you a friend, and not some one to fight with..Call this good hearted conversation..OK..?

I will post one more thing and then bow to your expertise..Cause I am NO expert...I did what you said...recover the dot with black permanent marker and even covered over the "1973"...NOW, this is the least powerful program that I use, so the outcome would be better if I used one of my powerful forms of software..Below is what I done within 2 minutes or less...You can see the "1973" from under the black and you can also see the actual colour of the :1973"...And rements of my original black spot...I took your blow up of my mark and as it looks, the red pixels are visible...


51studebaker...YOU ARE PROBABLY CORRECT...I am just wondering what any program can do..??

Thank you for the conversation 51studebaker..I have enjoyed it.

Wert,
No problems. Was just trying to explain how the software was working so folks didn't draw the wrong conclusions.
Our eyes (especially as we all get older!) can not discern many of the subtle shades that a computer or software can; for example look at these three squares of black.


They all look the same to our eyes but truth is that all three are a different shade of black any software can quickly determine.
don

Hello, Latin Heart Reader here. In my experience with neural networks I developed a patent for the hardware to mimic the retina. What it taught me is how the eye and brain process data and extract features differently from software written in linear code. More specifically the two positions regarding the red dot are correct. The extraction of the red dot from the dark smudge is done through what is called differentiation and that is what the retina does. My point is when using the tools your eye will give you a 'probability' there is something there and start stretching the difference of those same pixels that get averaged. It is amplifying the difference and thats what subsequent images shown did too, mimicing the eye. Point being this. Any software tool that lets you adjust contrast and color saturation will assist in creating the differences that the 'Shotwell' tool can process. My suggestion being perform the image enhancements by increasing contrast and color saturation that is apparent in Shotwell, before processing it with the tool. Then the results of the two approaches will become more similar. I will try to do an example off line.

Translation: What those digital removal tools are best for is the isolation the TOP colour; in other words isolating the cancel (or overprint) on a stamp, like the example showing the grid cancel on the 3c small squeen.

Another possibility is the permanent marker may be allowing the computer to still discern some difference in color as either a bleed through or a more dense area of coverage. Either one might look the same to our eyes and completely black. Your computer program may still detect the difference in the pixels however. Just a thought.

That makes sense Watermark. The scan is also able to detect the differences in depth too. So an ink mark on an engraved stamp would still show the engrave portion as a slightly different colour.

Now you guys are catching on to what I am trying to say..That is for example the small queen "vampire bit" that I was wanting to see under the cancellation....I don't care what colour it is, just that there is a indication of something in a particular area of the stamp...What I mean is the red mark I placed on the stamp and the black permenant marker I covered over the area, now we know if you mix black and red, you get the colour MAROON....Now here is the trick I was trying to show everyone..If you take a software program like Shotwell or better and diffuse the colour through the "threshold level" part of the program, you can isolate the MAROON colour...Does that make sense...When I try and remove the cancellation, I want to see the so called error shape whatever colour it may be...

NOW IF YOU WANT TO REMOVE THE BLACK AND I THINK THIS IS WHAT THESE TYPES OF PROGRAMS DO, YOU FIRST GET THE SOFTWARE TO REMOVE THE BROWN , THEN THE BLUE..BECAUSE BLACK IS A COMBINATION OF THE TWO...ONCE YOU HAVE ACHIEVED THAT, THEN THE BLACK SHOULD BE GONE, LEAVING ONLY THE STAMP AND ITS ORIGINAL COLOUR.

'Photochopping' an image, whether it be a stamp scan or an image of a model, simply means that the pixels have been 'changed' by the software. There is a way to filter images of people in light colored bathing suits to make them appear almost completely see-thru. But again, this is simply software 'enhancing' the deltas between pixels shades. The results are, at best, the computer interpretations of those deltas. (Same with using programs to assign a '3d' look to a stamp; it simply adds peaks to different color pixels.)

But it isn't a 'x-ray', there is no depth or z axis, it is simply pixel shade differences that are being algorithmically enhanced. No magic, its just pixels.
don

51studebaker...I guess we have beaten this topic to death...haha...Have a good day my friend, thanks for you input.

That is what we want, because we are unable to do it with our eyes.




...unless it is a 3d acanner.