I've been wondering what the focal lengths are for the two New Horizon cameras (Ralph and Lorri). Curious to know how much Alan et al. were able to squeeze into their weight budget.
The sad thing about the optical calibration issues that appear to be cropping up on so many current/future missions is that 35 years ago, they had the thing done and dusted:
http://members.tripod.com/petermasek/mariner9.html
"...star images were taken during the interplanetary cruise and also while orbiting Mars to determine and then monitor other geometric properties of the cameras. The narrow angle camera had a detection threshold of 9th visual magnitude, enabling at least a few stars to be imaged in any long exposure picture. Focal lengths and relative camera alignments were computed from the star images and were found to be stable to an accuracy of a pixel throughout the mission."
Of course, that was before all the fun with Hubble!
You'd think, though, that after Hubble, and MCO, and Genesis there's be a sorta checklist cum diary on each of the mission staff's desks - you know, things like:
* Fix Focus by Friday
* Go Metric on Monday
* Set pyros after Sunday (AKA This Way Up on Wednesday!)
The Mariner 4 images of Mars were badly degraded by a design-defect light leak. They were salvaged by digital image processing. The 6 bit data was inadequate for lower exposure images and after picture 11 of 21 1/10, the imaged degraded rapidly into digitization contour patterns
The Mariner 6 and 7 Mars images were badly degraded by analog tape recorder noise (the oxide flaked in flight and accumulated on the tape-heads), and electronic interference noise patterns, together with considerable geometric distortion of the images, severe shading across the images and bad residual images.
The Mariner 9 images were all-digital once the vidicon image was readout, but the cameras still had distortion, shading, and severe residual image problems. In addition, the color filter wheel stuck some 70 days into the mission, ending up at a polarized-orange position which was acceptible, but didn't help. The narrow angle camera also had severe defocus in images that had high exposure levels (not saturated)
Mariner 10 (Venus-Mercury) fixed the residual image problem by putting lightbulbs INSIDE the camera to create a uniform saturated-frame residual image (after the saturated frame was erased), permitting for the first time reasonable decalibration of the images, but the cameras had real problems with dust specks and lint shadows, and the spacecraft stability was inadequate for the narrow field of view, causing severe pointing wander. Look at Mercury flyby-2 mosaic segments for example. The real time 100,000+ bits/sec communication (experimental) was barely adequate at Mercury so the majority of images (other than the relatively few that could be taped for later playback) had severe salt-and-pepper noise.
Viking orbiters used twin cameras with silicon vidicon detectors instead of selenium-sulfer vidicons (I believe). The data storage system couldn't handle the 2 second (approx) raw readout rate from one camera while the other was taking an image in a rapid-fire left-right-left-right sequence for 50 meter/pixel landing site mapping from periapsis. The 7 bit (inadequate... orbiter images often have digitization contouring) data were split into several pixel channels and dumped in parallel tracks onto the tape recorder, which were then read out sequentially. Pictures (and there were many) where all tracks didnt' make it back to earth are missing columns of pixels in varying amounts which have to be filled in by interpolation. The zero-level of the camera ended up negative, instead of small positive numbers, so a blank exposure of black space was all ZERO's, and it was impossible to directly measure the dark-image shading of the cameras. Inflight calibration of shifts in the camera shading was difficult and generally inadequate. As each picture was being readout the other camera was shuttering it's exposure, then resetting the shutter. The mechanical shutter slap vibrated the cameras, resulting in a series of sine-wave interference patterns for a number of lines in the other camera's image, one near the bottom of the frame, one near the top.. due to "microphonic' vibrations of the photocathode in the camera that was beign read out.
Voyager's cameras were again selenium-sulfer vidicons, using the same lens as Mariner 10 for the narrow angle camera (with improvements). Dust specks were minor, calibration techniques reasonably well worked out. The cameras could be operated with light flood on for optimum calibration, or light flood off for low-light-level sensativity. Inflight calibration between encounters greatly helped maintain decalibrated image quality, though a lot of the cal data has annoying levels of bit errors and missing lines etc, and I had poor luck using the calibration target images to decalibrated flat-field shading from Voyager images (they had an aluminum plate they could get sunlight on by reorienting the spacecraft and then point cameras and other instrumets at). Image stability and pointing was better than Mariner 10's but still less than ideal. Stability was improved after Saturn for the extended mission by reprogramming the attitude control system. The vidicon cameras still had a problem with image distortion, and the images are literally "stretched" toward bright objects by a pixel or two due to electron-beam deflection by the charge pattern on the vidicon surface. This made truely precise geometric measurements in navigation images and the like quite difficult. CCD utterly don't have that problem!
but: "The sad thing about the optical calibration issues that appear to be cropping up on so many current/future missions is that 35 years ago, they had the thing done and dusted"... Uh.. done and dusted?.. no way!
I take the point regarding the steep learning curve on past missions - but it doesn't change the fact that nowadays the problems are known yet dumb mistakes keep getting made. I have every sympathy if it happens once, but after the Hubble mirror fiasco (and it's expensive recovery, for which all credit to those involved) you'd think that the word 'focus' would be at the forefront of everyone's mind...
Bob Shaw: "... but it doesn't change the fact that nowadays the problems are known yet dumb mistakes keep getting made. "
AMEN.
I think it was a BAD mistake not to have a coarse focus adjustment on the deep impact camera.
A big "Uh... I thought we knew about this..." problem... Stardust had a massive problem with condensate from outgassed crud fogging the optics. The earth-flyby image of the moon was 90% fog and 10% image.... massively blurred. Repeated heating of the camera to degass the optics got most of it, but I'm not at all sure they ever did get all of it and the Wildt comet flyby pics I think are somewhat degraded by it. NEAR's optics were significantly fogged by hydrazine byproducts during the loss-of-control event at the first arrival burn attempt that nearly lost the mission (not the camera's fault), and Cassini has had significant fogging problems between Jupiter and Saturn that I think still has some residual that makes it impossible to search for low brightness outgassing plumes at Enceladus.
Fogged optics seem to be the other problem-of-the-decade, besides out of focus cameras.
The Deep Impact defocus problem has distinct similarities to the Hubble problem. Tester malfunction, and not enough independent testing to catch the error. Granted, the Deep Space camera was assembled out of focus so that as the structure degassed, it would drift into focus, but images of real objects at or near infinity are remarkably good targets for a final check on optics... even if you only measure the defocus independently. (granted, it's not easy to get images of objects at infinity from a cleanroom)
Sadly, the noise added to the Deep Impact images will probably render multispectral imaging of the nucleus useless due to noise amplification during the deconvolution process. There may well be no usable color information with the original s/n of the camera, but all we may now see is the brownish overall color described in postings from the DPS meeting.
Anyway.... there's essentially no such thing as "Rocket Science"...
It's ROCKET ENGINEERING.
But 99% of reporters doen't even know the difference between a Scientist, an Engineer and a Technician.
I think the MRI, as with the impactor camera ( they were exact copies I believe) were sans-filters, the Cometary equiv of Navcam.
Doug
Oo - my bad, thought it was a straight copy. Perhaps it was just a copy of the electronics and optics - but had the filter wheel dropped in for the flyby.
Doug
I've heard nothing to suggest that they didn't get color-filter images through MRI. I may recheck that recent article in "Space Science Reviews", which mentiones among other things what kinds of geological color observations they had hoped to make.
In addition, if they can get what appear to be color variations in the HRI pics, then MRI could be used perhaps to verify that they aren't artifacts. If they are not artifacts, HRI might be able to tell us more precise positions of regional boundaries given the MRI "ground truth."
The article on Deep Impact imaging of geological features on Tempel's unimpacted surface is at http://www.beltonspace.com/bsei_web_page_g000000.pdf . Specific references to color imaging are on pages 2, 3, 6 and 9. The impression I get is that they could still get quite a bit of useful color data -- if, of course, there are any significant color differences to be found.
Well - I saw a reasonable colour image at the BAA conference - it showed a general brown colour, but there were different shades across the surface, some brighter areas etc.
Doug
Bruce Moomaw: ".... -- if, of course, there are any significant color differences to be found."
That's the rub. Small bodies tend to have vanishingly small color variations and it's very easy to go from noisy but usable data to unusable data on small scale features.
While overall color or spectral properties tell you about bulk composition, variations tell about chemical variations, or to a lesser extent physical property variations...they tell you about processes that made the surface, distinct from what you're told by feature morphology.
At risk of reopening what appears to be a dead thread (and mods may want to move this post for that reason), my search for LORRI optical characteristics led to this thread where the original question about Ralph and LORRI focal lengths was never answered. That information has been published elsewhere but the thread needs closure for the sake of other seekers who arrive here, and I have a follow-on question that is not answered anywhere else, so this thread may yet be useful.
The optical characteristics of the Ralph visible/NIR/IR camera are described in this paper, principally in Table 2 on page 6:
http://www.boulder.swri.edu/pkb/ssr/ssr-ralph.pdf
(TL;DR:
an unobscurred, off-axis, three-mirror anastigmat design;
Telescope Aperture: 75 mm
Focal Length: 657.5 mm
f/#: 8.7)
The optical design of the LORRI camera is described in these slightly variant resources, in Table 6 on page 10 and in section 3.1.2 Optical Design on page 12:
http://lanl.arxiv.org/ftp/arxiv/papers/0709/0709.4278.pdf
http://www.boulder.swri.edu/pkb/ssr/ssr-lorri.pdf
(TL;DR from the abstract:
a narrow angle (field of view=0.29°), high resolution (4.95 µrad pixels), Ritchey-Chrétien telescope with a 20.8 cm diameter primary mirror, a focal length of 263 cm, and a three lens field-flattening assembly)
I gathered that the LORRI primary and secondary mirrors are made from low thermal expansion silicon-impregnated silicon carbide. The field-flattening lenses are of fused silica. This design was based on monochromatic imaging (color via filters) therefore the refractive components are not achromatic (they can't be, being all of the same refractive index). This goes a long way in explaining some of the transmissive properties of the system. Yet two details don't seem to be mentioned anywhere:
1. What reflective coating was used on the mirrors (aluminum? something more exotic?)? Was the reflective coating hardened or overcoated in any way (such as silicon monoxide commonly applied to terrestrial mirrors)? Is deep space tarnishing even an issue?
2. Were the optical components coated in any manner? In a monochromatic environment, I suppose that spectral multicoating may be meaningless, but how durable are fused silica surfaces? And aren't internal reflections (ghosts) still a design problem for a multiple element field group?
The papers are otherwise helpful about nearly all other questions one might have about these camera systems. But even if coatings were not needed, that design point was not clear to me in the papers. Thanks for any... erm... illumination on this.
Thanks, all. I had read this on p. 12 and somehow inferred that baffles were the primary mitigation for ghost suppression. It makes sense that AR coatings were still involved.
Just came across a 1/2021 (open access) paper that goes into a bit of detail about various LORRI low light level imaging biases and methodology to remove them.
https://iopscience.iop.org/article/10.3847/1538-4357/abc881
Also, NASA PR page about result https://www.nasa.gov/feature/new-horizons-spacecraft-answers-question-how-dark-is-space
Powered by Invision Power Board (http://www.invisionboard.com)
© Invision Power Services (http://www.invisionpower.com)