Why are the LROC images so lousy?

Originally posted by ngchunter
GeoEye is google's top of the line satellite (not even the "average" satellite), and the resolution it attains is IDENTICAL to LRO's.

However GeoEye's altitude is 684 km vs 50 km for the LRO's final orbit. Can't really compare the two. It's like saying a cheap point and shoot camera is as good as the best professional hardware, if you can get close enough to the subject. GeoEye's angular resolution is much better.

Just thought Id throw this in..


NASA's LROC Satellite will offer bird's eye view of moon

A lunar satellite NASA plans to launch in May will send back the highest resolution photographs ever taken of the moon's surface, providing scientists -- and the public -- with a virtual view that's close to the one found by Astronaut Neil Armstrong in 1969. Scientists say the 10,000 by 1,000 pixel resolution of images from the Lunar Reconnaissance Orbiter Camera(LROC), will roughly equal one foot for every pixel. The satellite will orbit about 31 miles above the moon. "I believe we'll have better public imagery of some parts of the moon than we do of some parts of the earth," said Dan Stanzione, director of Arizona State University's Fulton High Performance Computing Initiative. Top 10 Ways to Increase IT ROI Without Adding Staff: Download now LROC is part of NASA's Lunar Precursor Robotic Program, and is the first spacecraft to be built as part of NASA's plans to return to the Moon." Arizona State has partnered with NASA and The Johnson Space Center to compile a digital archive of thousands of new images from the LROC as well as images from past Apollo mission flight films for the world to view. The Apollo film archive project started in June 2007 and is expected to be completed this summer.

Link

Originally posted by nablator

Originally posted by ngchunter
GeoEye is google's top of the line satellite (not even the "average" satellite), and the resolution it attains is IDENTICAL to LRO's.

However GeoEye's altitude is 684 km vs 50 km for the LRO's final orbit. Can't really compare the two.

Umm, yes you can when what you're talking about is image quality. The spatial resolution is identical. I guess you didn't notice, but the person I was responding to was claiming that google's satellite images have better spatial resolution than LRO. That's absolutely wrong. Whether it could get better photos if it were closer (or whether it would blur such photos or not) is irrelevant. Incidently, its focal ratio is f/12 (as opposed to about f/3.5 for LRO), so if you did put it as close as LRO is to the moon and kept everything else the same, the images would come out as either far too underexposed or extremely blurred.

[edit on 22-7-2009 by ngchunter]

Originally posted by nablator

Originally posted by ngchunter
GeoEye is google's top of the line satellite (not even the "average" satellite), and the resolution it attains is IDENTICAL to LRO's.

However GeoEye's altitude is 684 km vs 50 km for the LRO's final orbit. Can't really compare the two. It's like saying a cheap point and shoot camera is as good as the best professional hardware, if you can get close enough to the subject. GeoEye's angular resolution is much better.

I have to agree with ngchunter on this one. I don't dispute the differences you mentioned, but when you talk about how those differences translate to how many pixels per meter resolution you get on the ground, I think that's a fair comparison, so yes I think we CAN compare the two. You have a point that they arrive at their respective resolutions by very different means, but the end result is very close with respect to resolution.

And clearly a cheap point and shoot camera IS capable of getting results as good as a professional camera if the professional camera is far enough away and the cheap camera is a lot closer to the subject.

reply to post by ngchunter

I guess you didn't notice, but the person I was responding to was claiming that google's satellite images have better spatial resolution than LRO.

No, I wasn't exactly saying that.

Like I said, on the GE satellite imagery I can still see LM sized objects, way, way better.

You can get as technical on me as you want, but this is what I see with my own eyes.

Regardless of the LROC's specification and potential.

Originally posted by Point of No Return
No, I wasn't exactly saying that.

Right, you were using incorrect or esoteric terminology to describe spatial resolution.

Like I said, on the GE satellite imagery I can still see LM sized objects, way, way better.

If that were true it would be a function of spatial resolution. Google's satellite imagery is best provided by a satellite with spatial resolution matching that of LRO exactly. Ignoring the specifications in favor of subjective judgement won't change the cold hard facts. Personally I see no substantive differences in the spatial resolution of Tehran from LRO's best image at half meter resolution, though such a thing is hard to judge without a LM in the picture. You seem to be ignoring the fact that the LM pictures taken so far have not been at LRO's maximum resolution.


[edit on 23-7-2009 by ngchunter]

reply to post by ngchunter

You seem to be ignoring the fact that the LM pictures taken so far have not been at LRO's maximum resolution.

You seem to be missing the point. I was comparing to the current LROC images, you know, the ones this thread is about.

The thread is not about future LROC images, the thread's premise is that the current pictures aren't bad.

When I look at GE satillite imagery, 4 meter sized objects are visible very clearly.

Unlike the tiny spot in the LROC image that represents the LM.

That is the difference I see with my own eyes.

Ignoring the specifications in favor of subjective judgement won't change the cold hard facts.

Spewing technical information and specifications, won't change what I'm seeing with my own eyes.

When I look at the LROC Appollo 14 image, and open it full screen, the LM represents 4 pixels, when I use GE, using satillite imagery, I can look at a car, and have it covering 50 some pixels, and still clearly see it's a car, granted, not completely in focus, but if I could see the LM in the same quality, it would be much, much better than the LROC image.






[edit on 23-7-2009 by Point of No Return]

Originally posted by Point of No Return
reply to post by ngchunter

 

You seem to be missing the point. I was comparing to the current LROC images, you know, the ones this thread is about.

The thread is not about future LROC images, the thread's premise is that the current pictures aren't bad.

Do you really honestly think the difference between a 1 meter resolution and .5 meter resolution makes an image "bad"?

Spewing technical information and specifications, won't change what I'm seeing with my own eyes.

When I look at the LROC Appollo 14 image, and open it full screen, the LM represents 4 pixels, when I use GE, using satillite imagery, I can look at a car, and have it covering 50 some pixels, and still clearly see it's a car,

You're way oversampling that car. It still looks like a car to you because a car is conspicuous and a familiar shape to you. How familiar or conspicuous is a LM descent stage with a missing ascent stage viewed from overhead? Compared to a car, not very.

granted, not completely in focus,

Oh it was in focus, that I guarantee you, but you oversampled it by "zooming in" too much.

but if I could see the LM in the same quality, it would be much, much better than the LROC image.

The difference between a 1 meter resolution image and a .5 meter resolution image will not make the descent stage appear to be 50 pixels at a proper, native size, nor would it do the same to a car.

[edit on 23-7-2009 by ngchunter]

reply to post by ngchunter

Do you really honestly think the difference between a 1 meter resolution and .5 meter resolution makes an image "bad"?

Once more, for the last time, my point of view is that in the LROC image, the LM is just a tiny spot.

When I use GE I can see objects of the same size, only much bigger and still see the shape and characteristics.

You're way oversampling that car. It still looks like a car to you because a car is conspicuous and a familiar shape to you. How familiar or conspicuous is a LM descent stage with a missing ascent stage viewed from overhead? Compared to a car, not very.

Like I said, I can still see it's a car, very clearly, the windshield etc. It goes for all objects. I'm not talking about some blob of wich I know it is a car, I can clearly see it.

Oh it was in focus, that I guarantee you, but you oversampled it by "zooming in" too much.

Well, forgive me for my semantics mistake, I'll just say it wasn't completely sharp, yet, with the same quality, you would be able to recognize the LM.

The difference between a 1 meter resolution image and a .5 meter resolution image will not make the descent stage appear to be 50 pixels at a proper, native size, nor would it do the same to a car.

Lol, I know what I'm seeing on my screen, so it doesn't matter what you say.

Originally posted by Point of No Return
reply to post by ngchunter

 

Once more, for the last time, my point of view is that in the LROC image, the LM is just a tiny spot.

Then open it in an image editor and blow it up way past the size it's supposed to be; that's NO different than what you did in google earth.

I'm not talking about some blob of wich I know it is a car, I can clearly see it.

That's because a car has familiar features you've know about all your life that you instinctively look for. Your windshield point is a perfect example of exactly what I was saying. It can apply to just about anything in an earth focused image because the objects you see on the street are, of course, familiar to you. It's similar to a form of apophenia, only there was true data acting as a basic stimulus, you just oversampled it.

Lol, I know what I'm seeing on my screen, so it doesn't matter what you say.

You already admitted what you were seeing wasn't sharp, but you mischaracterized it as a focus problem. There is literally no resolving difference between oversampling the best LRO images and Google's satellite images, that's a cold hard fact irrespective of subjective analysis, the only true difference is the familiarity of the environment and its objects.

reply to post by ngchunter

That's because a car has familiar features you've know about all your life that you instinctively look for. Your windshield point is a perfect example of exactly what I was saying. It can apply to just about anything in an earth focused image because the objects you see on the street are, of course, familiar to you. It's similar to a form of apophenia, only there was true data acting as a basic stimulus, you just oversampled it.

Jeah, so? We also know what the LM looks like, so that argument doesn't hold up.

You already admitted what you were seeing wasn't sharp, but you mischaracterized it as a focus problem. There is literally no resolving difference between oversampling the best LRO images and Google's satellite images, that's a cold hard fact irrespective of subjective analysis, the only true difference is the familiarity of the environment and its objects.

Jeah, obviously I'm not very well informed on the technical details of camera's and such, plus English is not my native language, wich makes it a bit harder to find the right words.

But, even if I don't "zoom in" too far, so that the picture is still sharp, it's still better.

Maybe I should add that I'm looking at a 50" screen, wich makes differences more obvious.

Then open it in an image editor and blow it up way past the size it's supposed to be; that's NO different than what you did in google earth.

I was thinking about that, but I don't know how to do that. Maybe you could help?

I would like to see if that would come out the same, cause to me the LROC images also look a bit grainier compared to a GE satellite image with the same scale.

[edit on 23-7-2009 by Point of No Return]

Originally posted by Point of No Return
reply to post by ngchunter

 

Jeah, so? We also know what the LM looks like, so that argument doesn't hold up.

The difference is it's not an everyday object, so you're not psychologically trained to pick it out from an oversampled mess of semi-random pixels. You can do that instinctively with an oversampled photo of a car because a car is an intimately familiar object set against a highly contrasted but familiar background, the LM is not, not in the way that you're suggesting.

But, even if I don't "zoom in" too far, so that the picture is still sharp, it's still better.

If you don't zoom in too far, the picture of the LM's landing sites are very sharp as well. If someone handed you a photo of a street with cars on it you'd easily recongize it even if the cars were only 5 pixels long, if someone hands you a picture of a street without cars on it, you'd still recognize it and probably ask "where are the cars"? If someone randomly handed you a photo of an Apollo landing site of the same resolution with no LM on it, would you say "where's the LM"? Probably not, very few people would.

Only astronauts, mission planners, or geeks who've simmed the landing a dozen times or more would know the environment and the signature of the LM that intimately to pick it out the way you and I pick out cars on google. The first thing I did when I got home the night these images came out was download the massive uncropped apollo 17 photo and just manually found the LM using nothing but the experience I had of simming the landing. The lighting conditions were quite different, but it was still right where I expected it to be and stuck out like a sore thumb, as did Apollo 14 when I tried that one. (didn't have time or harddrive space to try all the others)

I was thinking about that, but I don't know how to do that. Maybe you could help?

I can't open the original source files on the machine I'm on at the moment, but if you can and want to try it yourself, here's a free program that's good for it. Just use the "resample" function set to bicubic sampling or something along those lines.
www.gimp.org...

I would like to see if that would come out the same, cause to me the LROC images also look a bit grainier compared to a GE satellite image with the same scale.

Given the fact that google always gets their satellite pictures at very high sun elevations, you shouldn't be surprised to see grainier images of lunar locations taken with very low sun elevations. Darker lighting leads to lower signal to noise ratios.

[edit on 23-7-2009 by ngchunter]

You people need to understand that the lens on this baby is of a fixed focal length, it doesn't zoom at all, the ONLY thing dictating the resoloution of the images is the distance between the probe and the surface of the moon.

That is no basis upon which to judge or criticize the images we are getting, it's a simple matter of camera to subject distance.

There also seems to be a lot of people who don't realize that these are 100% crops from images that are much, much larger, the pictures being posted of the landers are just minute parts of much larger images

It's no different to taking a photo of an entire tree with your digital camera and then complaining that when reviewing the images at 100% on your computer the particular leaf you are interested in is tiny and has only a few pixels dedicated to it. That is no measure of the quality of the camera at all and is simply a measure of how far away the camera was from the leaf.

People need to realise that (metaphorically speaking) the landers are just individual leaves in a photo of an entire tree. Nasa didn't set out to take photos of the apollo hardware, that was just an inevitable bonus, that's why the orbit and indeed the lens is not optimized for imaging apollo hardware, it is optimized for imaging potential landing sites.

Nasa are imaging things that are important but perhaps mundane to some (landing sites, images at a resoloution that help to understand and record the geography of the place) and people are complaining that they should be imaging things that are interesting but ultimately a waste of time and money (hardware that we all already know is there)

And anyone who thinks that nasa are out to prove a few conspiracy theorists wrong need to well and truly GET OVER THEMSELVES.

reply to post by weedwhacker

The government isn't going to need a satellite above the Lunar surface to see details as small as newsprint, now are they???

Actually.... I could find several uses for a camera capable of that. Look for direct evidence of ice in craters, look for small outgassing sites, minute measurements of the lunar surface that may be used as landing areas (look for cracks, or unstable areas).... The list goes on and on.

But hey. This is me. I'm not a NASA employee. Which is why all that's on Mars is a bunch of junk we sent up there, no colonies, ect.

I think the problem with the LROC pictures IS the lack of color, and hence a real lack of definition between the surface and the stuff we left on the moon. Black and white may work for certain things, but there's also stuff black and white is completely inadequate for..

Originally posted by wylekat

I think the problem with the LROC pictures IS the lack of color, and hence a real lack of definition between the surface and the stuff we left on the moon.

I disagree, B&W is excellent for seeing detail. Ever seen the beautiful work by Ansel Adams, for instance?

Anyway, there just isn't much color there, not unless you're way up close. Look at this:
Antares composite

The gold foil-covered landing legs, really too thin to see with the LROC. The upper surface of the descent stage that is sitting there is likely just bare aluminum. (If I get a chance I'll pop in to the NASM here in DC and look at the one on display, there).

Various scattered equipment left lying around, mostly monochromatic, at least in the resolutions we're talking about. Our best bet for picking out ANYTHING is by seeing the long shadows, thus the best shots are during early lunar 'morning' or late 'afternoon'.

(link)

[edit on 27 July 2009 by weedwhacker]

Originally posted by wylekat
Actually.... I could find several uses for a camera capable of that. Look for direct evidence of ice in craters,

How are you going to do that with a high focal length optical camera when the ice is covered in permanent darkness?

look for small outgassing sites,

That only works if the outgassing is frequent enough to detect with very sporadic pictures, and I don't see why a resolution > .5m would be necessary either. You probably have a slight advantage with a lower resolution but wider field of view.

minute measurements of the lunar surface that may be used as landing areas (look for cracks, or unstable areas)....

Again, is a crack less the 50cm wide really going to pose an issue that couldn't be solved by the commander on final approach?

Originally posted by wylekat
reply to post by weedwhacker

 

The government isn't going to need a satellite above the Lunar surface to see details as small as newsprint, now are they???

Actually.... I could find several uses for a camera capable of that. Look for direct evidence of ice in craters, look for small outgassing sites, minute measurements of the lunar surface that may be used as landing areas (look for cracks, or unstable areas).... The list goes on and on.

But hey. This is me. I'm not a NASA employee. Which is why all that's on Mars is a bunch of junk we sent up there, no colonies, ect.

But those things you mentioned are not the mission of LRO. LRO is simply mapping the areas that NASA will eventually go to during the next manned missions.

There is no reason for NASA to build a probe with additional capabilities -- and additional expense -- beyond what is required for the mission. If NASA had a huge or unlimited budget, then I would agree with you. With a huge budget, it would be nice for NASA to build their tools with capabilities beyond what is required by the mission. However, they don't have such a budget.

Furthermore, even if the equipment on LRO WAS capable, spending time looking for ice in craters and outgassing would detract from the primary mission -- which is mapping.

That's like wondering why the people who build weather satellites don't go to the extra expense of equipping their weather satellites with spectrometers and other specific equipment to help geologists find minerals.

[edit on 7/27/2009 by Soylent Green Is People]

reply to post by ngchunter


I found a handy way to get details out of images that are in shadow (not all of them). Not to mention the rig I am thinking of would actually come with a light....

My ideas are for the notion astronauts are unable to make it up there. I am also a little tired of hearing about 'amazing resolution' and seeing blurry photos better suited for vacation slides than much else. I think an orbiter with a camera capable of 1 mm/ pixel would be the lazy way to do lunar research, but then again, I *AM* lazy. We haven't sent someone to the moon in 30something years- so if we're going to go make cute pics, let's get close up, mind boggling pics. I wanna see the dust particles on the rocks, dangit! :-D

reply to post by wylekat

Furthermore, even if the equipment on LRO WAS capable, spending time looking for ice in craters and outgassing would detract from the primary mission -- which is mapping.

Just out of curiosity... how many times do they need to map the moon? This is like the multiple dozen-eth time, isn't it? It's a bit like mapping out a trip between my house and Disney World over and over and over again, using Google Maps' satellite function.