Reprocessing Historical Images, Looking for REALLY big challenges? Options

[GregM]

.

[Phil Stooke]

A few more words about the Surveyor panoramas. The first step was to find prints of the original pans for scanning. I often work at LPI in Houston, so on one trip I looked through the Surveyor stuff. They have all or most of the original frames archived as photo negatives, box after box of them - 80,000 total. But indexing is complex and messy, and besides I didn't have time to work that way. But they had dozens of folders of prints. I searched every folder. A few were mis-labelled, which didn't help. (I added a correct annotation but they are still filed with the incorrect labelling).

Among thousands of prints of individual frames or small mosaics were lots of 'regional-scale' mosaics, but most were compiled only for indexing/coverage plotting purposes and are 'aesthetically challenged' beyond belief. Ray Batson said they were compiled as the pics were coming off the printers to check for gaps in coverage etc., and they could barely keep up.

The useful ones are called "improved mosaics", made for science rather than mission operations. The tonal variations were reduced by processing (in the dark room, mainly) to give better-looking results. The collection includes a lot of mosaic fragments (sectors) with different lighting, but most in local projections which would take a lot of work to fit together. I needed the rectangular versions made at JPL which fit together into cylindrical pans. At LPI I found enough sections to do Surveyor 1, Surveyor 3, and Surveyor 7. These were 8 by 10 inch prints, one print per section, but sharp, and scanned well to give fairly detailed images. Full pans are about 10000 to 15000 pixels wide, each made of ten of these sections. LPI didn't have these for Surveyors 5 and 6. JPL didn't seem to have retained them - maybe in their archives which are a bit tricky to use like this.

Later I went to LPL in Tucson. They had large format prints of Surveyor pans, including Surveyors 3 and 5 which I scanned. They have the best Surveyor 3 material, including excellent prints of the horizon. Finally I spent a week at Flagstaff. USGS in Flagstaff had Surveyor 6 and a set of prints of a Surveyor 7 pan - but not in cylindrical format - which is different from the one I have done, taken near sunset. I'll post some details later. So between these locations I found the full data set that I needed. There is a LOT more to be done with the Surveyors, if anyone feels like a challenge!

Phil

[dvandorn]

A few more words about the Surveyor panoramas.  The first step was to find prints of the original pans for scanning.  I often work at LPI in Houston, so on one trip I looked through the Surveyor stuff.  They have all or most of the original frames archived as photo negatives, box after box of them - 80,000 total...

As I understand it, the original television signals were not captured and recorded, the only permanent record of the images are these negatives and any prints made from them, correct?

Let's see -- the Surveyor camera had a low-res 200-line mode and a high-res 600-line mode, if memory serves. It seems to me that what *really* needs to be done is for the negatives (yes, all 80,000+ of them) to be scanned directly, using a scanner that scans each individual line of the television image and generates a "line" of pixels that most closely represents that line of the original television signal.

You could then assemble those lines into digital images that can be both stored far longer and more securely than the negatives and can be manipulated to create highest-possible-quality images and panoramas. (Since you're scanning negatives, you'd just reverse the grayscale somewhere in the process to create positive images.)

In addition, since the mirror position on the camera was probably slightly variable, you might even be able to generate super-resolution images by overlaying images taken of the same scene in the same lighting conditions. At least three of the Surveyors survived long enough to image the same patches of real estate several times at very similar lighting angles over two or three lunar days.

It would not be a cheap process -- it would have to be a government-funded project, I imagine. Maybe it could be sold on the basis of preservation of historical data. But I can well imagine being able to design a scanning system that would accurately scan each TV line and create a high-quality digital version of the image. It would require precise negative positioning control and the width and "targeting" of the scanning beam would have to be custom-designed and precisely controlled, but given ten or fifteen million dollars, a staff of 10 or 15 people, and a couple of years, I bet it could be done quite effectively.

-the other Doug