A sign that all is not what it seems in the MER data - a strange mark has appeared.

One of the many programs I sometimes run on the Spirit and Oppo images is one which shows the areas of pure black in a picture as red spots of colour.

See here


Normally when I do this, there are splots here and there where the jpg has had the contrast increased and I know there is nothing which can be seen in the shadows if I lighten the whole photograph in my image editor.

However, while doing this the other day, a very regular shape appeared and I wondered what could have caused that.

It cannot be seen by the normal viewer as it was in an area of pure black rock 'shadow' and probably was not meant to be seen by me either. Therefore I post it here for any ideas which may have caused such a mark to appear in the pure black area covering the 'natural' shadow against this stone.

Close up image


Since it is only a bit of fun, I wont bother to give the full url as I do not think you will want to look at it. However, should anyone want to examine this picture, it is Sol 2153 1P319319001EFFABCXP2372L5M1

Reminds me of 'masked layers' in graphics programs. Hmmmm....

This thread seems to be intended for the realm of photography.
Could be related to the myth of the Leica glass in that case.

There is an on going debate amongst photographers about the properties of Leica camera lenses.
The effect sometimes called Leica Glow appears to increase depth of field.
Narrowing the debate to an apparent crux it has been claimed that Leica glass produces an image with higher color depth thus more information.
Opponents claim that simply adjusting F stop will produce the same effect.

Miles Davis the musician is a Leica patron and he claims that he had the camera shop adjust his Leica camera settings for him and he didn't touch the settings ever again!

Makes some sense to me that if a camera is optimized for a particular modality then trying to force changes yourself would probably just degrade the results.

I'm sure most of us photographers experienced F stop issues when we moved away from chemical emulsions. Is this tomorrow or the end of time.

Now could something like this be going on with the space imagery? I hope they get it resolved before the Gaia mission. Essentially GAIA will be sending two parallaxed streams of images to be correlated back on Earth. Over a billion stars located by their juxtaposition.

reply to post by qmantoo


Does that mean that you have never seen an "X" shape in a JPEG? They are relatively common.

reply to post by qmantoo


OP --

What happens when you use a higher resolution version of the image instead of the lower-res version (which has visible jpeg compression artifacts)?

Here is a higher-res version of the image you used. Could you please do the same after-effects on this image as you did to the version in the OP?:



Thanks.

That X cross reminds me of the old navigation devices we used to play with as kids before we could afford cameras.

@SoylentGreenisPeople - it is actually the same. Just now, I downloaded the one from your link and used that in my program and the results are identical to the close-up I posted previously. I think this would have been the version I used in the original running of the program.

The program itself does nothing special. All it does is to show red pixels where the RGB value is (0,0,0) and leaves the rest as they are.

The point is that it is not just any old cross, but is symetrical, so it is unlikely to be as a result of a compression or contrast adjustment algorithm..

Having done this many times before,I would expect to see some smaller areas of red pixels where the contrast adjusting program has reduced the greyscale value down to black which is zero (since they were near zero anyway). This would cause large areas of deep shadow to possibly become red pixels in my program since they had all become zero.

To have a rogue area in the middle of a large block of 'shadow' which shows up as a cross is unusual.

In a gif from the pds, it is unlikely to show as these have been balanced and the greyscale values are evenly distributed, however, if there are any areas of pure black, they will show as red in this processing from my program.

Originally posted by qmantoo
The point is that it is not just any old cross, but is symetrical, so it is unlikely to be as a result of a compression or contrast adjustment algorithm..

The fact that it is symmetrical is what makes me think of a compression artefact (that and the fact that it's not the first time I see one of those). Whenever you see something that fits perfectly in a 8x8 pixels square in a JPEG image then it's very likely to be a compression artefact, as JPEG uses 8x8 pixels squares to make the compression.

To have a rogue area in the middle of a large block of 'shadow' which shows up as a cross is unusual.

It's more unusual to you because you look only at those very dark areas, if you change the contrast (to make it more noticeable) on JPEG images you will see many artefacts and you will, eventually, find another cross.

In a gif from the pds, it is unlikely to show as these have been balanced and the greyscale values are evenly distributed, however, if there are any areas of pure black, they will show as red in this processing from my program.

GIFs use a different compression algorithm, so you will never see something like this (unless it was really on the photo) in a GIF (or PNG) image.

The fact that it is symmetrical is what makes me think of a compression artefact (that and the fact that it's not the first time I see one of those). Whenever you see something that fits perfectly in a 8x8 pixels square in a JPEG image then it's very likely to be a compression artefact, as JPEG uses 8x8 pixels squares to make the compression.

If, as you suggest, it is the result of JPEG compression, why are there no more compression artifacts or non-zero greyscale values within the black 'shadow' area marked in red?
Although you have put up a feasible explanation, it does not explain why it is the only non-zero set of greyscale values within the fairly large expanse of 'shadow'.

I think it is unlikely you are correct in this instance and so I do not think it is caused by the JPEG algorithm.

Originally posted by qmantoo
If, as you suggest, it is the result of JPEG compression, why are there no more compression artifacts or non-zero greyscale values within the black 'shadow' area marked in red?

Because the conditions weren't the same for the algorithm to produce a "X" shape.

Although you have put up a feasible explanation, it does not explain why it is the only non-zero set of greyscale values within the fairly large expanse of 'shadow'.

The reason for the non-zero values to be there has no relation to what I was saying, the values were there because that part of the shadow was not that dark.

I think it is unlikely you are correct in this instance and so I do not think it is caused by the JPEG algorithm.

I said that I had seen this before, so I went looking for the last time I saw it, and here is the image where I saw it. It was saved as a PNG to avoid recompression artefacts.


At the top of the "beam" there's one of those "X", as you can see in the image below (more or less at the middle of the "beam")

I don't know if that's related to the appearance of the "X", but we can see that the 8x8 squares to the right of the "X" have a smooth, artefact free appearance, like those black areas in the photo you posted.

PS: in other versions of the photo that I found on the Internet the "X" was not that noticeable or it was a little distorted.

Edit: here's a different version of the photo you used. This one was converted from an IMG file.

Thank you for the gif image from the pds.

As expected, there are no (or only about 6 pixels) of pure black in the png file but the 'shadow' area we are discussing is completely devoid of any detail - even in this image scientists are supposed to be looking at to do their investigations.

As you say, this cross appears to be a JPEG artifact, however, I have to say that in light of the difficulty of obtaining any detail from it(see below), I suspect that it may have been marked with a cross for censorship. Of course, that is pure specultion.

To bring controvercy into this, the shadow area is not showing shadow from the sun shining on a rock at all, but is some kind of black colour which has been placed there. Zooming into that area, it can be clearly seen that there are pieces of rock and surface which are hidden by the black area. Normally, shadow regions are not completely without detail, particularly close to the edge where some sunlight still reaches due to reflected light, but this area is totally obscured for some reason unfortunately.

Originally posted by qmantoo
As you say, this cross appears to be a JPEG artifact, however, I have to say that in light of the difficulty of obtaining any detail from it(see below), I suspect that it may have been marked with a cross for censorship. Of course, that is pure specultion.

Remember what I told before, if it fits perfectly in a 8x8 pixel square is most likely a JPEG artefact, regardless of shape.
Also, no one in their right mind works directly with JPEG files, and in NASA's (and other organizations working with this kind of images) case they use (as far as I know) a suit of programs called ISIS, and the programs use mostly a format specifically made for it, .CUB files, that allow them to work with different layers (like having a layer for each filter, for example, and creating images with the filters they want) and non-image data.

No need to add a "X marks the spot" if they wanted to censor any thing.

Having said that, I think I have seen somewhere that those MER photos that were published were published before any serious processing, they just converted them from the original data into a viewable format, took a look at the photo and published it.

To bring controvercy into this, the shadow area is not showing shadow from the sun shining on a rock at all, but is some kind of black colour which has been placed there.

How can you know? If you look at the original, non-processed image (not the one I published) you will see that the quality of the image was not that good, the whole image was too dark and with a small number of shades of grey.

I will post it when I come home at night.

Yes, the image was of poor quality,. but I can handle that. What I cannot handle is the black areas which are definitely not on the original which came down from Mars. How do the scientists see anything useful in those images and why dont they complain about it? Or more to the point, why dont they ask why it has been blacked out?

How do you explain those areas then - or is that not a subject you want to discuss?

Originally posted by qmantoo
Yes, the image was of poor quality,. but I can handle that.

The quality is very low on the original IMG file (I shouldn't have used the radiometrically corrected one before, sorry ), as you can see below.
(click for full size, I guarantee that there's an image there)

What I cannot handle is the black areas which are definitely not on the original which came down from Mars.

How do you know that?

How do the scientists see anything useful in those images and why dont they complain about it?

I think that the images are 12 bits per pixel images, so they can work with more than the 8 bits per pixel of the PNG or JPEG files.

Or more to the point, why dont they ask why it has been blacked out?

Now you are assuming that there was an area that was blacked out and that it was done before the scientists have looked at the images.

How do you explain those areas then - or is that not a subject you want to discuss?

Shadow.
Politics and economics are the subjects that I like the least, but I also discuss them, there's no taboo subjects for me, as should be any for anyone interested in the truth.

All I get is a black image, was I supposed to see something else?

How do you know that?

I know that the black areas are not on the original because (as I explained) I can see half-features on the rock which are just outside the border of the black censored area.

Shadows have a small area of grey before they go into deep black (even on Mars/Moon/etc, there is reflected light just at the border of light/no-light places). This means that normally, a lightening of the pixel values will show some extra detail but of a poor quality. Where the image has been artificially blacked out or censored, there is no light/dark transition area.

I think that the images are 12 bits per pixel images, so they can work with more than the 8 bits per pixel of the PNG or JPEG files.

So who gets to look at these 12-bits per pixel images? Where do they get them from - certainly, they do not get them from the pds like we can. If the PhDs at our universities are writing scientific papers based on this absolute poor quality then I do not think our science will progress very far.

Looking at the png you posted earlier, the border between land/sky does not show any pixels and has STILL got compression artifacts over the whole picture (although they are a different type). As I am sure you know, if you enlarge a photographic negative, you can see the grains of the silver iodide that have reacted to the light. If you zoom into a regular digital image you eventually see the individual pixels making up the image. Where are these then? Gif images go 'bad' and blocky when zoomed too much but these dont. Why not? Because we get a very very poor quality second or third-hand image whereas the one that gets downloaded is the best money could buy at the time the rover was manafactured.

Now you are assuming that there was an area that was blacked out and that it was done before the scientists have looked at the images.

I really cannot say if "the scientists" (who are these scientists who get exclusive access to 12-bit images?) have seen these images before they were blacked out or censored. However, I do believe that we dont get access to the best quality images which are available.

Shadow.

Ha! where have you seen shadows on Earth like this? (see the first paragraph of this post)

reply to post by qmantoo


FWIW, I saw a quake which was started from some friends throwing rocks (provoking) at the red landscape. After the ground shifted, flying X's came scrambling, from the ground, and bloody fast. It was, obviously a dream.

You may have something here. I hope so. I'm fairly certain that the dumbest rock on any planet is smarter, more wise, just, and universally in tune, than just about anyone here.

Originally posted by qmantoo
All I get is a black image, was I supposed to see something else?

If you only see black then you need to adjust your monitor, the image is there but only has 15 shades of grey.

I know that the black areas are not on the original because (as I explained) I can see half-features on the rock which are just outside the border of the black censored area.

An image with too few shades of grey would do that.

So who gets to look at these 12-bits per pixel images? Where do they get them from - certainly, they do not get them from the pds like we can.

If I'm not mistaken about the 12 bits per pixel images, that's how they are captured by the cameras on the rovers. I have to look again at that, I may be giving you wrong information.

Looking at the png you posted earlier, the border between land/sky does not show any pixels and has STILL got compression artifacts over the whole picture (although they are a different type).

Sorry, I don't understand what you mean by that

I really cannot say if "the scientists" (who are these scientists who get exclusive access to 12-bit images?) have seen these images before they were blacked out or censored. However, I do believe that we dont get access to the best quality images which are available.

The scientists that have access to the 12 bits per pixel images (if I am not wrong about it) are the ones working in the Rover's team. As far as I know, the IMG files are also 8 bits per pixel.

Ha! where have you seen shadows on Earth like this? (see the first paragraph of this post)

We are not looking at the shadow, we are looking at a bad photo of the shadow.

My doubts had some reason, here's the explanation about those 12 bits per pixel images, they are only used in the Rovers.

Each rover has a total of 9 cameras, which produce 1024-pixel by 1024-pixel images at 12 bits per pixel,[39] but most navigation camera images and image thumbnails are truncated to 8 bits per pixel to conserve memory and transmission time. All images are then compressed using ICER before being stored and sent to Earth. Navigation, thumbnail, and many other image types are compressed to approximately 0.8 to 1.1 bits/pixel. Lower bit rates (less than 0.5 bit/pixel) are used for certain wavelengths of multi-color panoramic images.

ICER is based on wavelets, and was designed specifically for deep-space applications. It produces progressive compression, both lossless and lossy, and incorporates an error-containment scheme to limit the effects of data loss on the deep-space channel. It outperforms the lossy JPEG image compressor and the lossless Rice compressor used by the Mars Pathfinder mission.

Source

ArMaP - OK, thanks for the information. So, what you are saying is that the pds is a farse and that scientists (or others) who are not part of the MER rover team do not have access to the 12-bit images that the team do? The only difference the pds offers is a stretched and prettied-up version of the raw contrasty images posted on the marsrover/nasa site, we dont get any better quality as fas as ultimate detail goes?

Is there anywhere where the better quality versions are held or are accessible to university scientists who are NOT members of the MER teams?

I suspect what you will say is that reputable institutions can request access from NASA for their research projects.

reply to post by qmantoo


No, if you read the post just above the last one you made you will see that the 12 bits per pixel images never leave the rovers. The cameras take 12 bits per pixel images and then change them to 8 bits per pixel and compress them. Those 12 bpp images are the equivalent of the raw image formats in common cameras, they just allow for a better initial image, before being converted to 8 bpp.

And no, I don't think the marsrover/NASA site has better images than the PDS, the fact that they are JPEGs is a sign that they cannot be better.