[
Quote:
Originally Posted by Bad. Guardian
Jon, it doesn't make sense to attempt to use circumstances on Mars, e.g. haze and sun angle, to explain changes which occurred _after_ the image was received on Earth.
|
I agree, but the basic image is hazy, has short diffuse shadows, and low contrast and was taken off axis, all of which are due to the conditions under which it was obtained. Processing can greatly improve on this, but cannot eliminate it entirely.
Quote:
Originally Posted by Bad. Guardian
You seem to be making the mistake of trying to view the "raw" image without first normalizing its contrast. The human visual system isn't well adapted for viewing images composed of pixels whose values are crowded together in the middle of the histogram. If you want to understand the processing that was applied to the image, you'll need to normalize the contrast of the raw image, before comparing it visually to the processed image. (In fact you won't need to normalize the contrast yourself, because I've already done it for you, in all the images that I've uploaded.)
|
If you normalise contrast you are no longer looking at a raw image, but a processed one. The raw image is this one:
http://www.msss.com/mars_images/moc/...elease/raw.gif
Quote:
Originally Posted by Bad. Guardian
The word "selectively" bears no such connotation. Non-human processes can be selective in their effects. However, I can understand how you could misinterpret my use of the word "aggressively." You may consider that word retracted, and replaced with "severely."
|
But we are not dealing with non human processes are we? We are dealing with image processing that is done by people. When you write “selectively eliminated" in this context then there is the clear meaning that this was done deliberately. “severely eliminated” does not get you out of it, because once again we are dealing with deliberate actions by the image processes. If the contrast is suppressed, it is because the processes have deliberately suppressed it.
But they have not. When you look at
http://mars.jpl.nasa.gov/mgs/target/CYD1/ the MIPL image is captioned “Contrast enhanced”, a fact that is obvious when comparing it to the raw image. Are you accusing them of lying?
Quote:
Originally Posted by Bad. Guardian
If the cropped version (before-after-slope.gif) shows too small an area for you, you can look at the uncropped version (before-after.gif). They are the same animation, except that one has been scaled down 50%, while the other has been left at 100% but cropped and highlighted.
The first frame (in both animations) is from cydonia1.tif, the image that was returned by the spacecraft (with left/right inversion corrected, and contrast normalized). The second frame is from cydonia1c-s.gif, the image that was used in the press release. I've applied a simple streak removal algorithm to both frames.
I'm attaching for you a side by side comparison, with a highlight showing the exact location where "before-after-slope.gif" was extracted. The left image is from the spacecraft, the right image is from the press release. Minimal processing has been applied, as in the two before-after animations, described above.
|
Thank you, that is very helpful and makes it much easier to compare. It is not a case that one is the raw image and the other processed, both are processed, the left hand image is by Parker, the right hand by MIPL. Those these images were released together by JPL. In the same press release.
Quote:
Originally Posted by Bad. Guardian
No, my analysis is not wrong. Whether they knew it or not, they used the "catbox convolution" that I described in my first two articles in this thread.
|
What do you mean when you say “Whether they knew it or not, they used the "catbox convolution" that I described in my first two articles in this thread.” Are you accusing them of not understanding what they are doing? This link, which we have both used, lays out the processing they did.
(1) The image is sized down by interpolation by a factor of two to reduce some of the noise.
(2) A long, narrow high-pass filter is applied in a vertical orientation to help reduce some of the instrument signature. This signature is seen as the streaking that is noticeable in the original data.
(3) A long, narrow low-pass filter is applied in a horizontal orientation to create an intensity average for the image.
(4) The results of these filtering operations are then stretched to approximate a Gaussian distribution.
(5) The results of the high-pass and low-pass processing steps are averaged together to form the final product.
(6) The image is flipped about the vertical axis to correct for the camera orientation.
Since this bears not relation to what you say they did:
a very simple 1x31 convolution, are you accusing MIPL of lying?
Quote:
Originally Posted by Bad. Guardian
Almost nothing can be seen, because you're trying to view the "raw" image without first normalizing its contrast. Don't try to do that; your human visual system simply isn't equipped for that task.
|
I am not trying to. I am pointing to it to illustrate the extremely low contrast in the original images and how all the processed examples improve on it.
Quote:
Originally Posted by Bad. Guardian
This is because all three (MIPL and Parker and MSSS) normalized the contrast. You need to do the same.
|
So you admit the MIPL normalised contrast? This means that the contrast is enhanced, but suppressed. This is quite contrary to what you claim they have done.
So you also admit that the Parker image is processed? This is not consistent with your use of it as representing the raw image in your thumb nails.
Quote:
Originally Posted by Bad. Guardian
The published description of MIPL's processing was vague; it wasn't detailed enough to allow outside people to replicate the effect exactly. My description of one part of their processing (the part that "flattened" the landform) is 100% explicit; it includes enough detail (even source code) to allow anyone to replicate the effect exactly.
|
Why would they do this? It can’t be a simple typo, leaving out an additional step, as we know that JPL corrected typos, as the amended release of the 7/4/98 shows.
So in your opinion not only are they lying, they are with-holding information.
Quote:
Originally Posted by Bad. Guardian
All images are contrast-normalized before publication. The "catbox" image, however, had another very important processing step, which selectively _suppressed_ all large-scale brightness variations, applied to it before its contrast was normalized.
|
And I am supposed to believe this do this, based on your say so?
Why would they do this anyway?
Quote:
Originally Posted by Bad. Guardian
I know the facts, and have known them since before I initiated this thread.
|
Unfortunately your “facts” are not very factual. You use the Parker image as the raw image, when it is highly processed.
Quote:
Originally Posted by Bad. Guardian
Thank you for your advice. I stand by my use of the word "clueless" in reference to that particular image.
|
So MSS, Parker and MIPL are all clueless as well as dishonest liars?
And your expertise for passing judgement on them is what, exactly?
Quote:
Originally Posted by Bad. Guardian
That is well and good. But one of those images, which was very widely published in mass media around the world, was extremely deceptive. Its creation and dissemination were Bad Astronomy. The failure to retract it was Bad Astronomy.
|
So you don’t like the MIPL processed image, that is your prerogative. All three images processed images were released on the same day, two (MIPL and Parker) in the same press release. It is not JPL’s fault that the media used the image you don’t like. Take it up with them.
Jon