Venera-13, Venera-14 Lander Images, Images generated from raw digital telemetry Options

[tasp]

As I recall, (sorry I can't cite a reference) the landers are expected to always be colder during their descent than their surrounding atmosphere, so there is a concern that 'crud' in the atmosphere will tend to condense on the lander. In the case of a sensor trying to study the 'crud' this might be a good thing. In the case of a camera lens, not so good.

[Guest_DonPMitchell_*]

I can confirm to you that it is indeed RAW imaging data : please find herebelow, an example of the data I'm using to produce my final images (four of the V13 cam2 images for blue, green and red channels)

I worked for a while with that set of raw images from Brown (Jim Head sent it to me). But there are mysterious problems with that data. Here is an example of one of the Brown U "raw" images, together with the same section from the original 9-bit telemetry (which I got from RNII KP in Moscow):

[attachment=5953:attachment]

When I was working with the Brown data, I noticed that there are sections of the vertical scanlines missing. Not whole scan lines, but parts of scanlines, with little horizontal fault lines then running across the image. You can see by comparing the lengths that about a dozen full lines of data are gone. I cannot imagine how that happened.

Another problem with the Brown U. data is that it is only 8 bits deep, and the 8th's bit is randomly scrambled, possibly also by photoshop, so there is really only 7 bits of data. One of my goals was to compute both photometric response and aperture correction for the camera. Having all 9 bits of data per pixel was very important for this (the raw images are logarithmic encodings of photometric brightness).

[vikingmars]

[quote name='DonPMitchell' date='May 30 2006, 08:11 PM' post='56311']
I worked for a while with that set of raw images from Brown (Jim Head sent it to me). But there are mysterious problems with that data.

Maybe it will help you : I was staying at Brown U. in 1984 when the tapes from Russia came to Jim's office with this very data on it : I think that Brown's is the "rawest" original imaging data available...

==> Congratulations, Don, for your nice imaging work !!!

[Guest_DonPMitchell_*]

Maybe it will help you : I was staying at Brown U. in 1984 when the tapes from Russia came to Jim's office with this very data on it : I think that Brown's is the "rawest" original imaging data available...

The very "rawest" data from the Venera probes is a series of repeated transmissions of 9-bit digital data. So there are actaully 3 or 4 versions of each camera transmission, with different noise and defects.

With all 9 bits, it is possible to reconstruct a 10th parity bit (which was sent from the lander to the spacecraft bus), and fix many portions of the images where the bit stream got out of sync. A lot of what looks like streaks of white noise is actually perfectly good data, just with the parity bit sitting in the high-order position. Then with all 3 or 4 transmissions, I was able to recreate an almost noiseless version of each camera transmission.

Deriving the photometric response of the camera (and the logarithmic amplifer) was another problem, and this has not been done correctly before. So it was possible finally to get linear brightness, and then convert that to the sRGB standard (which compensates for the response of television and computer monitors).

I still believe the color calibration has not yet been done correctly. I am still working with the Russian engineer who built the Venera cameras, to derive a true color balance. So don't completely believe the colors in any Venera images you see, including mine.

[Guest_DonPMitchell_*]

I should add that in the late 1980s, Carle Pieters at Brown University took a crack at the problem of color calibration. They created a panorama from the red, green and blue Russian images, balancing the color channels. Their choices were informed by experiments on the thermochromic shifts of oxydized basalts and iron oxides at high temperatures, and from Venera-9 data which returned a reflection spectrum from the nearby surface that went into the infrared.

Pieters, et al

Here is the image they produced:

[attachment=5960:attachment]

This is probably the best color result so far, but it has two probelms. First, it was based on an incorrect linearization of the data. And second, a somewhat more accurate result can be obtained by solving the spectral inverse problem for the camera (from the wavelength-dependent camera response).

Thermochromism and piezochromism are problems, because even the metal-oxide pigments that the Russians painted on the color test board were shifted in color by heat and pressure. Yuri Gekin measured the spectrum of the paint in their Venus simulator, under the proper heat and pressure, so that is helpful now. But the best result will come from data on the internal lamp used to create the calibration signals (seen at the top or bottom of the "raw" images).

[Bob Shaw]

Don:

Are those *more* colour calibration squares on the circular base of the lander, below the teeth? I'd never noticed them before!

Bob Shaw

[ugordan]

Why does there appear to be color fringing in the color composite, as if the 3 channels weren't correctly registered?

[paxdan]

Don:

Are those *more* colour calibration squares on the circular base of the lander, below the teeth? I'd never noticed them before!

Bob Shaw

I can make out CCCP

[tasp]

Why does there appear to be color fringing in the color composite, as if the 3 channels weren't correctly registered?

Are the colored dots on the edge of the probe (by the 'teeth') resultant from pebbles being blown onto the lander between the different exposures needed for color?

Why does there appear to be color fringing in the color composite, as if the 3 channels weren't correctly registered?

On the far left of the image, there does seem to be some color registration fringing. Perhaps the probe shifted ever so slightly between exposures.

Another idea; the outer skin of the probe would be expanding as it heated up, thus shifting the location of the imager just a bit between exposures.

[Guest_DonPMitchell_*]

Don:

Are those *more* colour calibration squares on the circular base of the lander, below the teeth? I'd never noticed them before!

Bob Shaw

Yes, six standard polysiloxane enamels were painted on the extended calibration board and on various parts of the lander (red, green, blue, white, grey and black). The letters "SSSR" (looks like CCCP) are printed there too, many people don't recognize that because they are very blocky and covered with dirt. Internal reports from RNIIKP give the spectrum of those enamels at normal pressure and under the temperature and pressure of the surface of Venus.

But you are also seeing false color effects from the wind blowing dirt around, during the time that the scene is being scanned though different color filters.

[Borek]

I've just noticed this article on space.com (via reddit), which is about Don Mitchell's work on Venera imagery:

http://www.space.com/scienceastronomy/0609...nus_images.html

Borek

[ljk4-1]

I've just noticed this article on space.com (via reddit), which is about Don Mitchell's work on Venera imagery:

http://www.space.com/scienceastronomy/0609...nus_images.html

Borek

Discussion on the images, plus more, started here:

http://www.unmannedspaceflight.com/index.p...ost&p=67529

[climber]

Don,

On the October issue of "Espace Magazine" (in french), there's an article called "Russian Back to Planetary Exploration" (approx translation ).
Talking about Venus they show your pictures (you're credited) and they say that your work is explained on your website : www.mentallandscape.com

[tedstryk]

I have put my Venera images back up. I am working on some new versions, but I figured that I needed to at least have the old ones on my site.

[aconnell]

Hi all. I guess like everyone else I was totally dumbstruck when I saw Don Mitchell's perspective renderings of the Venera 13 and 14 sites for the first time. From what I understand there is some colour information buried in the original dataset but there are various problems and challenges in extracting it to give a true idea of surface conditions as the human eye might see them. In the meantime I was wondering whether any of you image wizards out there might consider producing 'best guess' colour versions of these fantastic views based on the information already out there - and maybe the Venera 9 and 10 sets as well ? I know this is a big ask - particularly from a 'newbie' - and you are all busy people. But I for one would be really grateful and it is nearly my birthday. In two and a half months. Well more like three actually. Al.