Besides the camera on the Venera-9 lander, there were also two cameras in the orbiter. Violet and ultraviolet linear cameras that scanned the planet as the orbit swept the probe by. Not many have seen these images, made 18 months before the Pioneer Venus orbiter. (Indeed, many Pioneer Venus publications claim it to be the first artificial satellite of Venus!).
Click to view attachment
Here is a raw image, scanned on Dec 11, 1975 from 12:26 to 13:03 Moscow Time.
Click to view attachment
From the times of the scan, and the parameters of the elliptical orbit of the probe, we can correct the view to create something closer to perspective (actually a spherical x orthogonal view). In addition, I did some deconvolution and contrast expansion. Actually looks like Venus now!
Full Version: Venera-9 Orbiter
The Venera 9 and 10 orbiters had a very substantial scientific payload with some quite capable instruments. Their science return was considerable, though the imaging return was limited.
Mars 2 and 3 made a very poor comparision with Mariner 9, though the Soviets tried to puff-up their capabilities by counting the number of "experiments" on board, quite a few more than the Mariner's 2 camera Imaging System, 1 infrared spectrometer, 1 infrared radiometer, 1 ultraviolet spectrometer, and 1 radio science system. Unfortunately, the Mars 3 (2 did very little reported science) experiments were mostly rather crude photometers and radiometers that returned a few channels of data from single stripes of measurements limb-to-limb as the orbit carried the fields of view across Mars. Mariner had 2 highly capable spectrometers (the IR radiometer was like a Mars 3 instrument) and a scan platform.
Venera 9 and 10 had real spectrometers and the other instruments were more sophisticated. I suspect they had a very substantially higher data rate and total data return than the Mars '71 orbiters as well.
Pioneer Venus improved a lot on some (but not all) the Venera data, but it was still a small, limited capability mission compared with what a purely hypothetical Mariner 13 (the Voyagers were originally to have been Mariners 11 and 12) orbiter might have done.
Mars 2 and 3 made a very poor comparision with Mariner 9, though the Soviets tried to puff-up their capabilities by counting the number of "experiments" on board, quite a few more than the Mariner's 2 camera Imaging System, 1 infrared spectrometer, 1 infrared radiometer, 1 ultraviolet spectrometer, and 1 radio science system. Unfortunately, the Mars 3 (2 did very little reported science) experiments were mostly rather crude photometers and radiometers that returned a few channels of data from single stripes of measurements limb-to-limb as the orbit carried the fields of view across Mars. Mariner had 2 highly capable spectrometers (the IR radiometer was like a Mars 3 instrument) and a scan platform.
Venera 9 and 10 had real spectrometers and the other instruments were more sophisticated. I suspect they had a very substantially higher data rate and total data return than the Mars '71 orbiters as well.
Pioneer Venus improved a lot on some (but not all) the Venera data, but it was still a small, limited capability mission compared with what a purely hypothetical Mariner 13 (the Voyagers were originally to have been Mariners 11 and 12) orbiter might have done.
Wow!!
Great stuff.
Great stuff.
Mars-3 had six specialized spectrometers and photometers, some to just measure water vapor and to measure the depth of the atmosphere by comparing CO2 to non-CO2 IR bands. Besides all other factors, trouble with the radio on Mars-3 probably limited the amount of data returned.
Mars-5 did much better work, and it included a polorimeter designed with Audouin Dollfus. Venera-9 carried a lot of these improved Mars-5 devices and some Venus-specific IR thermal-balance type experiments.
Mars-4 and -5 also carried (in addition to the two phototelevision cameras) a linear camera like the ones on Venera-9.
Click to view attachment
This is one of two panormas sent from Mars-4 as it whizzed past. An electronics failure prevented it from performing its orbital insertion burn.
Mars-5 did much better work, and it included a polorimeter designed with Audouin Dollfus. Venera-9 carried a lot of these improved Mars-5 devices and some Venus-specific IR thermal-balance type experiments.
Mars-4 and -5 also carried (in addition to the two phototelevision cameras) a linear camera like the ones on Venera-9.
Click to view attachment
This is one of two panormas sent from Mars-4 as it whizzed past. An electronics failure prevented it from performing its orbital insertion burn.
QUOTE (DonPMitchell @ Jun 4 2006, 12:57 AM) 
Besides the camera on the Venera-9 lander, there were also two cameras in the orbiter. Violet and ultraviolet linear cameras that scanned the planet as the orbit swept the probe by. Not many have seen these images, made 18 months before the Pioneer Venus orbiter. (Indeed, many Pioneer Venus publications claim it to be the first artificial satellite of Venus!).
Wow. My surprise is that they designed cameras just for violet and UV. I thought the detail of Venus in that part of the spectrum was known only after Mariner 10, meaning that the Soviets sort of hustled to put instruments with that purpose on the craft. Mariner 10 flew by Venus in Feb. 1974, and Venera 9 was launched in June, 1975. That's a respectable response time.
Astronomers had known for a long time that you could see cloud features on Venus in UV.
What they DIDN'T know (before Mariner 10) was whether there were any small or low contrast markings at visible wavelengths, or if there was any "relief" to the clouds, sticking up above the haze which seemed to dominate atmospheric models based on spectra and photoemtry.
That's why they added the "piggyback" wide-angle cameras to the Mariner 10 imaging system, using a fiber-optic "coupler" in one filterwheel position to periscope the image from the wide angle camera to the narrow-angle cameras' vidicons. The flyby trajectory approached the planet from the nightside and took them past the terminator pretty nearly at closest approach. They did wide-angle visible-light mosaics up and down the termniator as they swung through closest approach looking for anything like cumulus towers and cloud shadows and didn't see anything at all like that.
That's why they added the "piggyback" wide-angle cameras to the Mariner 10 imaging system, using a fiber-optic "coupler" in one filterwheel position to periscope the image from the wide angle camera to the narrow-angle cameras' vidicons. The flyby trajectory approached the planet from the nightside and took them past the terminator pretty nearly at closest approach. They did wide-angle visible-light mosaics up and down the termniator as they swung through closest approach looking for anything like cumulus towers and cloud shadows and didn't see anything at all like that.
QUOTE (edstrick @ Jun 4 2006, 11:28 PM)
What they DIDN'T know (before Mariner 10) was whether there were any small or low contrast markings at visible wavelengths, or if there was any "relief" to the clouds, sticking up above the haze which seemed to dominate atmospheric models based on spectra and photoemtry.
And it still somewhat of a mystery. The Russians studied their (rather crumby) Venera-9 images to see if the UV features were raised (white stuff above dark, or dark stuff above light). As I understand it, the white cold collar is somewhat raised, but the dark UV features appear to just be some kind of chemical mixing. But I'm not sure if anyone really knows yet. VIRTIS should give us lots of dramatic pictures of the lower cloud layer, whcih seems to be much more inhomogeneous.
I am still keeping my fingers crossed the the Venus Express Fourier Spectrometer will get unstuck. Otherwise, we will not learn anything new about the composition of the clouds or what the UV absorber is.
The dark-marking material seems well-mixed with the upper haze layer, as the markings show good contrast up to and at the limb. If there was haze without the dark-marking material above the markings, the markings would fade out toward the limb as the line of sight through haze above the markings gets more oblique and more opaque.
Back in 98, I was able to get Galileo Venus images off the net and did some experimental enhancements on them. "Venus_1" is an essentially raw image from late in the encounter, "Venus_2" is a cropped version of the image after it's been bandpass-filter enhanced to progressively boost contrast on local detail. At both low and high latitudes, contrast of UV or Violet (I'm sure of the filter bandpass) markings remains essentially constant from mid-disc to the limb at any given latitude. (You can see 8-bit quantization near the terminator and that the gray levels are not evenly spaced. You can see this more in later images.)
Ven_40A is a bandpass filtered short wavelength image from much earlier in the encounter while Ven_41A is a Near-InfraRed image, similarly bandpass filtered. Contrast in the IR data is much lower than in the Violet/UV, and significantly decreases toward the limb, as overlying haze gets more opaque. Notably, there is UTTERLY no correlation between features in the two images, and the east-west IR markings do NOT have nearly as much poleward "<" shaped tilt as the short wavelength markings. Apparently, there is much less circulation of atmosphere from equator to poles at the depth of the IR markings than higher up.
Finally, Ven-c1 and Ven-c2 are 2-band color images composited from these two frames, the first one is the raw data (most artifacts removed) and the second is the bandpass filtered data. The latter emphasizes how little the markings correlate.
Back in 98, I was able to get Galileo Venus images off the net and did some experimental enhancements on them. "Venus_1" is an essentially raw image from late in the encounter, "Venus_2" is a cropped version of the image after it's been bandpass-filter enhanced to progressively boost contrast on local detail. At both low and high latitudes, contrast of UV or Violet (I'm sure of the filter bandpass) markings remains essentially constant from mid-disc to the limb at any given latitude. (You can see 8-bit quantization near the terminator and that the gray levels are not evenly spaced. You can see this more in later images.)
Ven_40A is a bandpass filtered short wavelength image from much earlier in the encounter while Ven_41A is a Near-InfraRed image, similarly bandpass filtered. Contrast in the IR data is much lower than in the Violet/UV, and significantly decreases toward the limb, as overlying haze gets more opaque. Notably, there is UTTERLY no correlation between features in the two images, and the east-west IR markings do NOT have nearly as much poleward "<" shaped tilt as the short wavelength markings. Apparently, there is much less circulation of atmosphere from equator to poles at the depth of the IR markings than higher up.
Finally, Ven-c1 and Ven-c2 are 2-band color images composited from these two frames, the first one is the raw data (most artifacts removed) and the second is the bandpass filtered data. The latter emphasizes how little the markings correlate.
The whole issue of the nature of the UV markings always makes me wonder why they pulled the camera from Mariner 5. A Mariner-4 camera, with the light leak fixed and operating at 7 or 8 bits per pixel (granted it could have only stored a few images, but this wouldn't have taken very many pictures to solve) would have resolved this. Incidentally, they almost had Mariner-4 use its camera at 5 bits per pixel so that they could take more pictures. That would have been a disaster!
It's really good to see these great images. Thanks for showing them.
Phil
Phil
QUOTE (edstrick @ Jun 5 2006, 02:27 AM)
Apparently, there is much less circulation of atmosphere from equator to poles at the depth of the IR markings than higher up.
Fascinating image analysis. Have you done something similar with the Mariner-10 images? There are UV, Orange and IR images available there. The zonal winds (the superrotation) vary a lot with altitude. It's hard to say if we're seeing meridional or zonal flow differential, given the angled shapes of the dark features.
Back in 81 or 82 as a grad student, I had an acquaintance in the JPL image processing labs attempt to process a Mariner 10 orange/blue/UV (I think) image sequence to try to composite as color. The automatic batch processing put some nasty dropouts in the images and while I shot slides of the raw data which I still have, I never was able to find time to try to clean up the damage for tests on the digital data, which I lost access to as my PhD program collapsed.
The primary data set was a small set of rather small full-disk images taken a day or two after the encounter. There was so little expectation (it seems) that multispectral imaging would be worth much that very little was done at Venus.
The images are now available, but not decalibrated. If I had access to them, preferably in a raw 16 or 32 bit form (preventing data aliasing during decalibration, ideally), I could play my "tricks" on them. I recall that the orange filter image did have a distinct light and dark tilted band in the high southern latitudes that were approximately matched by UV markings.
I will be very interested in seeing the Messenger multispectal data, particularly the evolution of very low contrast details as wavelength changes from near IR through visible to blue (where the low contrast versions of the UV markings show up) In particular, I'm interested in seeing if the Blue, long-wavelength-UV and (if they get it) short-wavelength-UV features match one-for-one or if there's some subtle color differences in that part of the spectrum that may sort out more than one absorber or provide info on absorber altitudes.
The primary data set was a small set of rather small full-disk images taken a day or two after the encounter. There was so little expectation (it seems) that multispectral imaging would be worth much that very little was done at Venus.
The images are now available, but not decalibrated. If I had access to them, preferably in a raw 16 or 32 bit form (preventing data aliasing during decalibration, ideally), I could play my "tricks" on them. I recall that the orange filter image did have a distinct light and dark tilted band in the high southern latitudes that were approximately matched by UV markings.
I will be very interested in seeing the Messenger multispectal data, particularly the evolution of very low contrast details as wavelength changes from near IR through visible to blue (where the low contrast versions of the UV markings show up) In particular, I'm interested in seeing if the Blue, long-wavelength-UV and (if they get it) short-wavelength-UV features match one-for-one or if there's some subtle color differences in that part of the spectrum that may sort out more than one absorber or provide info on absorber altitudes.
Attached are 3 images for an amusing demonstration of my bandpass filtering method applied to slightly different data.
I work for what used to be a small company that built and sold industrial CT scanners -- Scientific Measurement Systems, of Austin Texas (now defunct). What used to be their "scanning services" division is now Pratt & Whitney Austin, where we do CT scan inspection of (mostly) jet engine turbine blades, plus other exotic and oddball items, including rocket engine parts (shuttle included) and other stuff.
I do image processing and analysis with a company home-brew image processing system that can handle 3-dimensional image "cubes" of up to a thousand or more images in 32-bit real-number-format, with total file sizes limited to 2 gigabytes by the Sun sparkstation file system file size limits. It gives me enormous power in some areas, but I'm often crippled by a total lack of commands in the software to do what sould be simple tasks, but which were never written and never needed for the industrial imaging uses.
Anyway, here's a quick demo on a familiar object:
Apple_r1 and _r2 are digital radiograms of an apple (a trace over-ripe, with a bruise or two). R1 is the raw data, directly comparable to a conventional x-ray. R2 is a bandpass filtered version of R1. Besides nearly invisible (in R1) fibrous internal structure, you can see a bruise shadowed on the right side of the apple and shallow bruises "on the limb" along the edges. Apple_1 is a CT (computed tomography) slice approximately through the "equator' showing some of the radial structure, the skin and the bruises.
I work for what used to be a small company that built and sold industrial CT scanners -- Scientific Measurement Systems, of Austin Texas (now defunct). What used to be their "scanning services" division is now Pratt & Whitney Austin, where we do CT scan inspection of (mostly) jet engine turbine blades, plus other exotic and oddball items, including rocket engine parts (shuttle included) and other stuff.
I do image processing and analysis with a company home-brew image processing system that can handle 3-dimensional image "cubes" of up to a thousand or more images in 32-bit real-number-format, with total file sizes limited to 2 gigabytes by the Sun sparkstation file system file size limits. It gives me enormous power in some areas, but I'm often crippled by a total lack of commands in the software to do what sould be simple tasks, but which were never written and never needed for the industrial imaging uses.
Anyway, here's a quick demo on a familiar object:
Apple_r1 and _r2 are digital radiograms of an apple (a trace over-ripe, with a bruise or two). R1 is the raw data, directly comparable to a conventional x-ray. R2 is a bandpass filtered version of R1. Besides nearly invisible (in R1) fibrous internal structure, you can see a bruise shadowed on the right side of the apple and shallow bruises "on the limb" along the edges. Apple_1 is a CT (computed tomography) slice approximately through the "equator' showing some of the radial structure, the skin and the bruises.
Anyone got a walnut handy? We could see what's inside Iapetus!
Bob Shaw
Bob Shaw
Very cool, never seen a CT scan of an apple before.
For image processing, I use a C++ class I wrote that stores images as an array of single-precision floats. It takes up space, but it elimiates all the annoying problems of dynamic range. And I have 8 Gb on my PC, and am running x64 Windows. Now I wish I had put 16 Gb in it!
I use photoshop a lot too, but it is not very good at filters and doesn't do gamma correction properly. I haev GIMP installed too, just to check it out, but its basically an ersatz version of Photoshop, OK for freeware though. Personally, I would recommend Photoshop Elements to people who want a cheaper solution.
I also use ACDsee a lot, for browsing and quick editing. I have a love-hate relationship with that software, and I desperately wish I could find a better replacement. I've looked at ThumbsPlus, FastStone and Irfanview, and they suck. I just can't use them, their interfaces don't support my workflow at all. But ACDSee is the buggiest most horrendously written software on the planet. I have to use the old version 5 because the newer versions crash all the time. Version 5 makes mistakes about locking down folders and accessing files, it messed up jpeg files, there is just no end to it. But I'm stuck with it until someone writes a decent replacement.
Image Analyzer is cool, I only use it for its frequency-domain editor, which is primitive, but the only one available. Its deconvolution filters don't seem to work right, but fortunately Photoshop CS2 has "smartsharpen".
For image processing, I use a C++ class I wrote that stores images as an array of single-precision floats. It takes up space, but it elimiates all the annoying problems of dynamic range. And I have 8 Gb on my PC, and am running x64 Windows. Now I wish I had put 16 Gb in it!
I use photoshop a lot too, but it is not very good at filters and doesn't do gamma correction properly. I haev GIMP installed too, just to check it out, but its basically an ersatz version of Photoshop, OK for freeware though. Personally, I would recommend Photoshop Elements to people who want a cheaper solution.
I also use ACDsee a lot, for browsing and quick editing. I have a love-hate relationship with that software, and I desperately wish I could find a better replacement. I've looked at ThumbsPlus, FastStone and Irfanview, and they suck. I just can't use them, their interfaces don't support my workflow at all. But ACDSee is the buggiest most horrendously written software on the planet. I have to use the old version 5 because the newer versions crash all the time. Version 5 makes mistakes about locking down folders and accessing files, it messed up jpeg files, there is just no end to it. But I'm stuck with it until someone writes a decent replacement.
Image Analyzer is cool, I only use it for its frequency-domain editor, which is primitive, but the only one available. Its deconvolution filters don't seem to work right, but fortunately Photoshop CS2 has "smartsharpen".
QUOTE (tedstryk @ Jun 5 2006, 12:53 PM)
The whole issue of the nature of the UV markings always makes me wonder why they pulled the camera from Mariner 5. A Mariner-4 camera, with the light leak fixed and operating at 7 or 8 bits per pixel (granted it could have only stored a few images, but this wouldn't have taken very many pictures to solve) would have resolved this. Incidentally, they almost had Mariner-4 use its camera at 5 bits per pixel so that they could take more pictures. That would have been a disaster!
Had Mariner 5 carried that camera, its photos would apparently all have been close-up like Mariner 4's, vastly reducing its ability to make any sense out of the cloud patterns. But then-- judging from the science team's official document on the science payload (which I have a copy of) -- the main goal for the TV experiment would have been to look for chinks in the clouds through which Venus' surface could be seen.
I am discussing what was surprisingly not done, not what some old planning document said was considered. Yes, I realize that they had considered another strip of closeups, but why they never considered an approach sequence is beyond me.
This is a "lo-fi" version of our main content. To view the full version with more information, formatting and images, please click here.