PDA

View Full Version : Proof! Bio Station Alpha is Just an Image Artifact



Fraser
2011-Jun-11, 02:50 PM
It’s time for another episode of “Conspiracy Theory of the Week.” This one is involves a supposed secret space station on Mars. The You Tube video showing “Bio Station Alpha” (below) went viral and was even reported on some mainstream media outlets. The station is supposedly a 700 ft x 150 ft structure on Mars [...]

More... (http://www.universetoday.com/86497/proof-bio-station-alpha-is-just-an-image-artifact/)

Buttercup
2011-Jun-11, 02:52 PM
It's a COVER UP!! :mad: NASA is lying like usual! :( Don't allow yourself to be blinded to THE TRUTH that it really IS a space station!

redshifter
2011-Jun-11, 07:46 PM
Geez, another conspiracy theory ruined by a bunch of pesky facts...:)

chrlzs
2011-Jun-12, 12:26 AM
I have no problem with the explanation, and not for a moment did I suspect it was anything but a glitch... BUT, why on earth has it been post-processed so poorly? Do NASA *want* to encourage CT's?

If that little streak is so badly distorted and enlarged and has so much added false 'detail', then everything else in the image is also suffering from the same damage*. I do a lot of p-p and analysis of imagery, and if I saw *that* result and compared it to the original from which I was working, I would know I was doing something awfully wrong.

I took a look at the 'original' link (http://hrscview.fu-berlin.de/cgi-bin/ion-p?ION__E1=UPDATE%3Aion%3A%2F%2Fhrscview.ion&image=5620_0000&image1=3+images&lat=71.822&lon=330.450&zoom=10&mode=nd&scale=12.5&pview=North&exag=1.5&viewport=600x400&image0=5620_0000&code=095491209&UPDATE.x=298&UPDATE.y=198) provided, and that is *not* the original - it clearly shows compression artefacts and jpeg block boundaries. They aren't as bad as in the image under scrutiny, but they are already a problem.

However, this cropped version (http://www.universetoday.com/wp-content/uploads/2011/06/MarsExpresspct.jpg) does appear to be largely uncompressed, and near enough to an original... Examination of that one suggests the cosmic ray strike was resolved down to pretty much exactly one pixel wide. Yet the post processed image shows false detail and compression artefacts spanning approximately 30 pixels!

I have to ask - what is the point of such a distorted/over enlarged/over processed/over compressed image?




* - Ok, that's not precisely correct, as the extent of the artefacts will depend on both the contrast and width of the details, but the point is made...

HenrikOlsen
2011-Jun-12, 12:18 PM
Ask Google, it's their algorithms which took the image and mapped it to their viewer.

And naturally the point is to shot the images mapped to the surface of Mars in a cool way. Nobody sane does planetary science by looking at Google Mars and analyzing the images as see there.

chrlzs
2011-Jun-13, 02:43 AM
Ask Google, it's their algorithms which took the image and mapped it to their viewer.

And naturally the point is to shot the images mapped to the surface of Mars in a cool way. Nobody sane does planetary science by looking at Google Mars and analyzing the images as see there.

Oops. I thought the mangled image had been attributed to NASA (clearly I read too many conspiracy threads!). And yes, of course Google'd images are going to have to be resized and processed in order to get them to match up in image mosaics, but it seems pretty obvious that their processing leaves a lot to be desired.

And NASA are not innocent in this arena - they too offer up badly processed images quite frequently. The SOHO beacon images are a 'good' example - similarly over enlarged and compressed - and they have often given CT-ists good fodder...

But you are absolutely right - no sane researcher would ever be caught not going back to the *real* originals. The prime audience for this stuff is, sadly, not very discerning, so the 'researchers' can throw up pretty much anything..

ravens_cry
2011-Jun-13, 05:05 AM
Considering that an oft repeated conspiracy claim that many have done youtube videos of is camcorder video of a television playback of degraded video, I doubt many CT understand anything about primary sources.

Jeff Root
2011-Jun-13, 07:06 AM
I have no idea how the Mars Express images are colored.
Can anyone explain it here? I would imagine that the distortion
of the cosmic-ray-hit-pixels into a Mars base is an unavoidable
result of the colorization process and the extreme contrast
between the bad pixels and the surrounding area.

-- Jeff, in Minneapolis

chrlzs
2011-Jun-13, 01:18 PM
I have no idea how the Mars Express images are colored.
Me neither. But the colorisation isn't really the problem here - the problem is with what might be called 'edge artefacts' - in other words, in the processes of enlarging and possibly sharpening/contrast enhancement and then finally recompressing the image, much added 'detail' has appeared. This false detail is in the form of 'ringing'/sharpening haloes, compression block effects, even what may well be *double*-processing effects from enlarging an already compressed and therefore artefacted image..

These problems are well-known to anyone who tries to enlarge a digital image (esp. a jpeg) without the requisite knowledge of just how far you *can* enlarge any given image, and the very careful use of appropriate software. Effectively, once you get to '100%' (ie one original image pixel for each display device pixel), further enlargement *will* introduce false details and unless you use extraordinary care, these sort of problems will rapidly escalate.


Can anyone explain it here?
I'm trying to.. :)


I would imagine that the distortion
of the cosmic-ray-hit-pixels into a Mars base is an unavoidable
result of the colorization process
As above, it's not so much the colorisation (although you can introduce some issues like posterisation in that way), but...


.. and the extreme contrast between the bad pixels and the surrounding area.
That's the key. The original image (http://www.universetoday.com/wp-content/uploads/2011/06/MarsExpresspct.jpg) doesn't have much in the way of 'bad pixels' - the cosmic ray strike (if that is what it is) is rather nicely and sharply recorded in a single row of pixels, with little if any surrounding distortion.

But the key point is that the detail is at single pixel width in the original, in other words it is right at the extreme limit of the sensor's resolution. Which means that ANY attempt to enlarge that to two, or five (or in this case THIRTY!!!) pixels simply has to guess at the intervening pixels. And different enlarging algorithms will use different approaches to that guessing process and will give different results. And none of them know the truth - it's all false detail! When you then add the process of sharpening, which creates a slight halo around contrasty edges PLUS the compression artefacts that arise when a jpeg is saved, you end up with all sorts of false detail. Add in excessive enlargement, and hey presto, a Martian base. And if they also break the cardinal rule of doing this twice (or more), then you have an unmitigated imaging disaster on your hands.

Give me five minutes and I'll blow it up twenty times further, and I'll be able to start pointing out aliens in Escamilla fashion - I'll just need to try several different enlarging algorithms, sharpening settings and compression levels until I get something...

There's a good example here (http://www.general-cathexis.com/interpolation/index.html) (may take a while to load images), showing the different approaches that various enlarging algorithms take. Essentially there is no free lunch - once you get to 100%, further enlargement just gives you guessed pixels - as you can see, none of those is 'correct' or even particularly useful...