Spaceship based telescopes

[Anthony]

High tech is wrong; accurate is more like 8-10. Human 20:20 vision is about 1/3,500; at 250km that's a resolution of 70 meters. Dropping that down to 15cm is 500x better, or 9 levels of telescopic vision.

[David L Pulver]

Originally Posted by Anthony

High tech is wrong; accurate is more like 8-10. Human 20:20 vision is about 1/3,500; at 250km that's a resolution of 70 meters. Dropping that down to 15cm is 500x better, or 9 levels of telescopic vision.

One of the biggest problems I faced working vehicle design rules has been trying to scale the optical and infrared sensors to GURPS values. Let me know if this makes sense:

* HIGH TECH will let you have a 10x stabilized binoculars are about 4 lbs. at TL8.

* As far as I can ascertain, the various parts of optics tend to scale fairly well with the square of what GURPS simplifies as "magnification."

* This would suggest that a 100x system, 10x larger, would be 400 lbs. and a 1000x system, 100x larger, would be 40,000 lbs. or more.

* However, various sources and HIGH TECH's suggestion for Telescopic Vision implies that a Hubble or KH-11 style system is closer to 4,000 to 8000x. That would be 160, 000 to 640,000 lbs.

Allowing for telescope having a smaller field than binoculars (half weight, say) that's still 80,000 to 320,000 lbs. That's much larger than the maybe 4,000 to 16,000 lbs. the actual system (depending on versions) should weigh.

So, which is wrong:

(a) "scale by the square of magnification"
(b) the estimate that Hubble / KH-11 is about 4,000 to 8,000x magnification
by GURPS standards
(c) the base weight of the 10x system.
(d) it's right, but when you blow a billion dollars on a telescope, you should be able to get multiple levels of "super very expensive"

Also, it looks like if you scale DOWN you get a super-light 1x magnification system. At TL8 what would be a realistic weight for a 1x "human eyeball" camera?





What's the solution?

[Ulzgoroth]

Originally Posted by David L Pulver

One of the biggest problems I faced working vehicle design rules has been trying to scale the optical and infrared sensors to GURPS values. Let me know if this makes sense:

* HIGH TECH will let you have a 10x stabilized binoculars are about 4 lbs. at TL8.

* As far as I can ascertain, the various parts of optics tend to scale fairly well with the square of what GURPS simplifies as "magnification."

* This would suggest that a 100x system, 10x larger, would be 400 lbs. and a 1000x system, 100x larger, would be 40,000 lbs. or more.

* However, various sources and HIGH TECH's suggestion for Telescopic Vision implies that a Hubble or KH-11 style system is closer to 4,000 to 8000x. That would be 160, 000 to 640,000 lbs.

Allowing for telescope having a smaller field than binoculars (half weight, say) that's still 80,000 to 320,000 lbs. That's much larger than the maybe 4,000 to 16,000 lbs. the actual system (depending on versions) should weigh.

So, which is wrong:

(a) "scale by the square of magnification"
(b) the estimate that Hubble / KH-11 is about 4,000 to 8,000x magnification
by GURPS standards
(c) the base weight of the 10x system.
(d) it's right, but when you blow a billion dollars on a telescope, you should be able to get multiple levels of "super very expensive"

Also, it looks like if you scale DOWN you get a super-light 1x magnification system. At TL8 what would be a realistic weight for a 1x "human eyeball" camera?





What's the solution?

How much of the weight of a 4 pound set of binoculars is actually optics, and how much is a rugged, ergonomic casing?

The Hubble can't cut corners on optics (though it might be able to be more weight-efficient on them by using some alternate design that's either too costly or otherwise unsuitable for binoculars, I don't know the details of space telescope design), but its casing is neither rugged nor ergonomic.

[Kraydak]

Originally Posted by Ulzgoroth

How much of the weight of a 4 pound set of binoculars is actually optics, and how much is a rugged, ergonomic casing?

The Hubble can't cut corners on optics (though it might be able to be more weight-efficient on them by using some alternate design that's either too costly or otherwise unsuitable for binoculars, I don't know the details of space telescope design), but its casing is neither rugged nor ergonomic.

Actually, spacecraft have a really good claim for being rugged. Remember: they are really really expensive, and really really hard to do maintenance on (Hubble got really lucky). Any time the engineers can fit spares for a mission-critical piece, they do, and everything that can be engineered conservatively is. Spacecraft also have to deal with violent vibrations (launch) and extreme environments (space).

[Anthony]

Originally Posted by David L Pulver

* However, various sources and HIGH TECH's suggestion for Telescopic Vision implies that a Hubble or KH-11 style system is closer to 4,000 to 8000x. That would be 160, 000 to 640,000 lbs.

Diffraction-limited optics have a resolution of (wavelength) * 1.22 / (diameter), meaning you need a minimum lens diameter of (wavelength)*1.22/(resolution). Human 20:20 vision is defined as 1 minute of arc resolution (1/3438 radians) at a wavelength of 400-700 nm, so for 1x you need (700 nm) * (3438) * (1.22) = ~2.9mm optics. Most good telescopes are diffraction-limited, so the Hubble has about 800x human visual resolution.

You can, of course, magnify by more than 800x; it will just be a bit blurry. Also, the diffraction limit is the size of the blurry spot a point-like object becomes; if you know you're looking at point-like objects (which will often be true in astronomy) you can do a bit more processing to find the center of the blurry spot, and maybe even figure out that two blurry spots are overlapping. This will not typically be very useful outside of astronomy.

Originally Posted by David L Pulver

So, which is wrong:
(a) "scale by the square of magnification"

Really, 'scale lens diameter by magnification. This tends to result in an actual scaling factor between 2 and 3, as a larger lens requires more support.

Originally Posted by David L Pulver

(b) the estimate that Hubble / KH-11 is about 4,000 to 8,000x magnification

This one is wrong (well, it might magnify by that much to make the images easier to work with, but it doesn't have that much resolution).

Originally Posted by David L Pulver

(c) the base weight of the 10x system.

Binoculars are not a good model for sensors, as they're limited by the structure of your eye. Also, field of view matters, a telescope with a 1 degree field of view is a lot less weight than a wide angle lens.

Originally Posted by David L Pulver

Also, it looks like if you scale DOWN you get a super-light 1x magnification system. At TL8 what would be a realistic weight for a 1x "human eyeball" camera?

The equivalent of human visual resolution in a 60 degree arc (the human eye only has that resolution in about a three degree arc, but cameras cannot easily emulate that) is about 13 megapixels, so your average cell phone camera is slightly worse than human eyeball.

[Ulzgoroth]

Originally Posted by Anthony

Also, the diffraction limit is the size of the blurry spot a point-like object becomes; if you know you're looking at point-like objects (which will often be true in astronomy) you can do a bit more processing to find the center of the blurry spot, and maybe even figure out that two blurry spots are overlapping. This will not typically be very useful outside of astronomy.

Actually, I think it's also being used in fluorescence microscopy these days, where the point-like objects you are looking to localize are individual fluorescent molecules.

[Anthony]

Originally Posted by Ulzgoroth

Actually, I think it's also being used in fluorescence microscopy these days, where the point-like objects you are looking to localize are individual fluorescent molecules.

True. It's also useful in radar -- aircraft are reasonably likely to be a blurry blip, but you can usefully aim at the center of the blur as long as they're in midair.

A lot of the problem with statistics is that figures like 'magnification' aren't actually meaningful -- you can magnify by as much as you want. What you want is resolution.

This also applies to night vision gear. Typical NVG only has around 100x the raw light-gathering power of a dark-adjusted human eye, but it might well turn the gain up to well over 100x, because if you do, the user can use it at night without waiting half an hour for his eyes to adjust.

[David L Pulver]

Originally Posted by Anthony

True. It's also useful in radar -- aircraft are reasonably likely to be a blurry blip, but you can usefully aim at the center of the blur as long as they're in midair.

A lot of the problem with statistics is that figures like 'magnification' aren't actually meaningful -- you can magnify by as much as you want. What you want is resolution.

The magnification number is just a colorful but inaccurate way of tracking bonuses to the roll; the real number of import is the bonus.

A 800x magnification = +8 (or +16 with narrow focus). If that represents a hubble/KH-11, the question is whether that is sufficient?

The range penalty for 250 miles or so is -32. If you have an analyst with skill 14 and a +16 bonus and perhaps a +2 time spent bonus you end up with effective skill 32.

If the range penalty for 250 miles is -32, this gives a 0 chance of success.

Anthony, this suggests that 800x cannot - in GURPS terms - adequately represents a telescope of hubble sized (KH-11) pointing down. What am I missing?

[Anthony]

Originally Posted by David L Pulver

The range penalty for 250 miles or so is -32. If you have an analyst with skill 14 and a +16 bonus and perhaps a +2 time spent bonus you end up with effective skill 32.

If the range penalty for 250 miles is -32, this gives a 0 chance of success.

Anthony, this suggests that 800x cannot - in GURPS terms - adequately represents a telescope of hubble sized (KH-11) pointing down. What am I missing?

+10 for in plain sight, to start with, and typically target size, spy sat pictures usually look at buildings, vehicles, and crowds, not individual humans. The equivalent situation with naked eye observation is an analyst looking at 35mm pictures taken from a plane at 500 meters -- range penalty -15, skill 14, +2 time bonus, modified skill 1.

[David L Pulver]

Originally Posted by Anthony

Diffraction-limited optics have a resolution of (wavelength) * 1.22 / (diameter), meaning you need a minimum lens diameter of (wavelength)*1.22/(resolution). Human 20:20 vision is defined as 1 minute of arc resolution (1/3438 radians) at a wavelength of 400-700 nm, so for 1x you need (700 nm) * (3438) * (1.22) = ~2.9mm optics. Most good telescopes are diffraction-limited, so the Hubble has about 800x human visual resolution.

Really, 'scale lens diameter by magnification. This tends to result in an actual scaling factor between 2 and 3, as a larger lens requires more support.

Binoculars are not a good model for sensors, as they're limited by the structure of your eye. Also, field of view matters, a telescope with a 1 degree field of view is a lot less weight than a wide angle lens.

The equivalent of human visual resolution in a 60 degree arc (the human eye only has that resolution in about a three degree arc, but cameras cannot easily emulate that) is about 13 megapixels, so your average cell phone camera is slightly worse than human eyeball.

When you say "scale lens diameter by magnification" I am not sure what you mean. Do you mean that a 800x magnification sensor should be 800x, rather than 800 x 800 times, heavier than a 1x sensor?

If a given optical sensor of 1x magnification is "weight x", what would you recommend a 10x or 1,000x sensor be?





"Field of view" is tricky as GURPS often uses it to include the degree you can pan or scan back and forth. For example, GURPS routinely gives optical scopes a "tunnel vision" field of view (about 60 degrees) even though the scope field of view is only a few degrees.

What do you think the weight increase for a sensor should be between "human eye equivalent" (240 degree), "typical vision goggles" (GURPS assigns 120 degree) and "typical scope" (GURPS assigns 60 degree)?