Free Republic
Browse · Search
Bloggers & Personal
Topics · Post Article

To: Polarik

I would like to commend you on your effort and very good research.

Several months ago, myself and a friend of mine noticed immediately the CoB posted on Fight the Smears was a glaring fake.

He and I both first noticed the grey/white pixilation between the characters after zooming in.

I immediately knew why this was. I have forged several images (not CoBs though).

The granulated dithering and pixilation is due to two image layers. There was an original background image with the native text.

Because you can’t convert back a JPG or GIF to an editable image format, the original scan must stay JPG.

So, to forge, you have to use the clone tool to cover the existing text with the native background colors.

This is where the problem occurs. Because the graphics editor doesn’t have the original layer in an editable format (i.e. fireworks png etc.), it “guesstimates” the color correction with a color correction algorithm.

Because it can’t quite make a perfect color match (not the original layer), it just dithers or attempts to get as close as possible, which is pretty good to the naked eye.

When you zoom in though, it jumps right out at you, because you are viewing the pixels up very close.

So, the image was done in least two layers. The original scanned image and the text overlay.


59 posted on 11/25/2008 5:03:10 AM PST by jetxnet
[ Post Reply | Private Reply | To 57 | View Replies ]


To: jetxnet; LucyT; pissant
Thank you, Jetxnet, and blass you! You are the only other person who knows how this forgery was created, and who agrees with my pixel analysis 100 percent. Can I add your name as a corroborating expert?

You were spot-o in your discussion, but what I learned from creating a clone of the Kos image, is that the forger never had a real, paper COLB in his or her possession. Although the forger received several scan images, one of them had been converted (or saved as) a bitmapped, color-indexed image, and the overlays were done in a bitmapped format before it was compressed and saved, not once, but twice: the high amount of image compression coupled with the large image size and high color count (about 70% or lower could only have been achieved in this sequence. I made about 415 test images until I got it right - but the intermediary step involving a bitmapped image was the hicker because every other combo did not obliterate the green pixels to the degree as they weere shown.

I'll email to you the steps I took, to see if you can also recreate the forged Kos image.

60 posted on 11/25/2008 5:38:20 AM PST by Polarik
[ Post Reply | Private Reply | To 59 | View Replies ]

Free Republic
Browse · Search
Bloggers & Personal
Topics · Post Article


FreeRepublic, LLC, PO BOX 9771, FRESNO, CA 93794
FreeRepublic.com is powered by software copyright 2000-2008 John Robinson