APPENDIX


A TECHNICAL DISCUSSION ON THE DIFFERENCES BETWEEN TRUE AND CLONED PORTRAITS


By using special filters ("clones") in the painting software, anyone with little or no artistic talent can make an acceptable image that in many respects looks like a real painting, but it's not. It's neither artistic nor "painterly" and is more a photograph than a painting.

Some paintings are partially cloned and in part truly painted. In portraits, you will often find the faces of the subject cloned while the background may be truly painted.

There is nothing wrong with cloned or partially cloned paintings as long as they are represented as such. Photographers do this to enhance their photographs, but they are selling photographs, not paintings. A misrepresentation comes about when the "artist" says his painting is a real oil painting when it's not. This is evidenced when he claims on his web site or elsewhere that "no filters were used," or "each stroke was painstakingly applied," or "the work took 10 to 15 hours to do" when, in fact, the cloning filters were used and none of the above was true.


A true painting uses the computer software to provide the canvas, the paint and the brushes, not the artist. The artist is a real portrait artist who scans and places the photograph in his computer side by side with a blank canvas. He then uses the brushes and paints supplied by the software to paint on the blank canvas while eyeballing the photograph the same as he would with a live model. This process is time consuming but results in a true "painterly" painting.


A non-true or cloned painting will take 4 to 5 minutes to do and will look exactly like the photograph. A partially cloned painting may take 25 to 30 minutes with a simple background (like the example below which took 30 minutes to make) or may take longer if real images (like flowers or trees, etc.) are truly painted in the background. On the other hand, a 100% true portrait painting of one person may indeed take 10 to 15 hours to do, and even longer if more people are involved.

HOW CAN YOU TELL IF A PORTRAIT PAINTING HAS BEEN CLONED?

If it looks like art, it probably is. If it's 100% cloned, it's easy to tell. It looks exactly like the photograph, including all the background clutter.

If it's partially cloned, it's more difficult to tell. You can observe the painting closely and try to tell if certain "body parts" such as eyes, mouths, teeth, ears, noses and hands look surprisingly similar to the photo. These are the most difficult and time-consuming things to paint and there may be a temptation for the artist to take a short cut and clone them. More than likely any color portrait made from a black and white photograph will be truly painted.

There is a simple test which, in most cases, will detect partial cloning. The image of Duncan below was partially cloned. Only the background sky was truly painted. Duncan in the foreground was cloned (that's why he took only 30 minutes to "paint"). Drag the image out onto your desktop so that it copies as a file (not all browsers will do this so if your's doesn't work, try a different one). Then open the file in one of many programs you may have that has a magnifier or ZOOM function. See below for examples of zoom-capable programs.


While focusing on Duncan, Zoom In until he becomes pixilated (i.e. until you can see all the little squares or pixels around the edge). Stop zooming when this first occurs. It should take four or five zoom clicks (400% to 500% magnification) for him to become pixilated. Now look at the sky, which was truly painted, and you will notice it is not pixilated and remains smooth. This confirms your suspicion the painting was partially cloned. In fact, you would have to zoom eight or nine times before the sky became pixilated.

This phenomenon can perhaps be explained by the fact a true painting is "layered" with several coats of "paint" as the artist passes his brush back and forth over the painting and blends it in in order to obtain the effect he wants. A cloned subject has only one coat. Because of the thinness of this cloaned coat, it gets pixilated before the heavier layers of the true painting. It's analogous to several coats of varnish on a piece of furniture building up a scratch proof surface while only one coat flakes off quickly.

When you test a 100% true painting by zooming, it not only becomes pixilated at a higher magnification than a cloned painting, but it will get completely pixilated all at the same time.

Note: This test doesn't work where a photograph of the painting instead of the image file itself was put directly onto a web site. Some true artists take photogaphs of their work to show on the web. You can tell this is the case when there obviously is a regular picture frame on the picture or it's shown on an easel or there are objects showing behind the picture. These photographed paintings would get 100% pixilated at the same time at a rather low magnification (depending on the camera resolution), but this doesn't tell us if it's true or cloned.


DRAG THIS IMAGE TO YOUR DESKTOP FOR A SAMPLE TEST
(Follow directions above)

Partially cloned
painting of Duncan

Below is the result you will get. This portion of the painting of Duncan was enlarged with 5 Zoom clicks (500%). Notice Duncan's image, which was cloned, is pixilated (blurred) while the portion of sky shown, having been truly painted, is still smooth.

Duncan enlarged 500%

SOME COMPUTER APPLICATIONS WITH ZOOMING CAPABILITIES

ACDSee, Adobe Photoshop, Apple ColorSync Utility, Apple Preview, Apple QuickTime Player, ArcSoft PhotoStudio, Canon Image Browser, Canon PhotoStich Viewer, Corel Painter, Lemke GraphicConverter, Microsoft Windows Picture and Fax Viewer

Products and companies referred to herein are trademarks or registered trademarks of their respective companies or mark holders. The owner and developer of this web site has no affiliation with any of these companies except, in some cases, as authorized user of the referenced products and makes no representations or warranties whatsoever as to the authenticity or accuracy of said products in performance of the tests outlined herein.