Mars Hale Crater Looking West

[Shrike]

http://en.wikipedia.org/wiki/JPEG

Scroll down to where they down sample a picture.
Behold rectangular shapes where there shouldn't be.
It gets even worse if there's transformation from one format to another.

As the typical use of JPEG is a lossy compression method, which somewhat reduces the image fidelity, it should not be used in scenarios where the exact reproduction of the data is required (such as some scientific and medical imaging applications and certain technical image processing work).

ESA down samples a picture for web down load.

[tholden]

Problem is, you're talking about seeing individual pixels and you'd have to zoom those images I posted thirty or fifty times to see pixels; I'm only talking about zooming two or three times.

This is what you're talking about:




and this is what I'm talking about:



If you can't see the difference you really need to have eyes checked.

[Shrike]

No im talking about this.







No not magnification but sampling quality (down sampling)

And please don't make it personal !
Nothing wrong with my eyes

[tholden]

The effect is still spread out over the image, and you still can't produce anything like the image I posted that way.

[tholden]

In fact you might want to try it. If down sampling of the original image is producing the structures in the images I posted then those kinds of structures should be there throughout the imaage. Try it. Convert to grey scale, zoom a couple of times and adjust the contrast and brightness until you see the structures I'm referring to in the lower half of the image and then check out the upper half. All you'll see is rocks.

[Shrike]

Ok i did that and i see similar structures all over that image except for the really bright parts.
the picture is to large to post here.
But of course it is not exactly the same as the area you mention but quite similar

[tholden]

One section of the upper part of the image, same settings which show structure in the lower part other than brightness set high enough for the upper area. Just rocks and again artifacts of down sampling would be everywhere on the image, i.e. I would not be able to come up with any part of it which did not show such artifacts.

[Shrike]

Ok i copied 2 sections of the picture i made there's rectangles all over the place.
Actually one can see already rectangles in the original picture.
But that's all because of the jpeg algorithms everything you do to a jpeg causes information loss or information alteration due to jpeg coloring and magnification algorithms.
I know for sure that this is not the original picture.
Because esa does not send a multimillion dollar satellite into mars orbit to only get a JPEG of only 1.2 mega byte size.

Here's what i did from original picture: (http://esamultimedia.esa.int/images/mar ... ater_H.jpg )
Original is already severely down sampled (first step of information loss/alteration)
transform to gray scale (second step of information loss/alteration)
magnify from 100% (original) to 200% witch means 1X magnifaction. (third step of information loss/alteration)
If i would do another step of magnification but then image would be even more compromised by jpeg magnification algorithms. (jpeg is for pretty pictures anything you do to it is not true any more).
adjusted brightness to minus 83 % adjusted contrast to + 70% (fourth step of information loss/alteration )





I do admire your sense of investigation and all your effort.
But what you are doing to a jpeg image is just not justifiable. Maybe you can ask them for the original hi-resolution image and then do the same. (if they are ever going to give you that.)
It would be an image of 10 to 100 mega bytes in file size.

[tholden]

I mean, yeah, I can see what appears to be jpeg artifacts in that but that's all. There just aren't any tall rectangular buildings in that and there are in this:



Huge difference.

[Shrike]

I'm sorry but i don't see any buildings.
I only see jpeg contrast and brightness and even more magnification enhancement.

[tholden]

I'm sorry but i don't see any buildings.
I only see jpeg contrast and brightness and even more magnification enhancement.

Then you should assume I'm posting this stuff for the benefit of those who can.

[Aardwolf]

After looking at the available images my suspicion is that it's due to colour banding. You can actually match the structures from different perspectives in available colour images but when you look at the black and white images the structures dissappear. Probably because of the greater contrast available.

[tholden]

There actually are other people who see it...

http://www.marsanomalyresearch.com/evid ... idence.htm

[webolife]

Pixelation and greyscaling artifacts.
I couldn't help reading the article author's closing note:

NOTE: Remember that you will not be seeing any civilization evidence detail beyond the faint organized patterns in this official image unless you are prepared to do graphics clarification work with it. If you choose that path, you'll need to first loose the pretty, highly reflective but obscuring haze and ESA colors by converting to standard black and white grayscale format as well as invert the image. This will partially defeat the obfuscating top layer and better reveal the evidence. Then proceed with darkening and contrast work to clarify and then add color if you wish.

Joseph P. Skipper, Investigator

The image inversion in particular artificially creates "outcrops" out of depressions, and the the enlarging process diminishes the truer rhomboidal shape of the pixellated regions.

[Shrike]

Don't you just love how Joseph P. Skipper calls it "evidence".

The effect is still spread out over the image, and you still can't produce anything like the image I posted that way.

That's the correct term 'produce' it is not in the actual image it is produced by software algorithms.