Any comment I hear about the Huygens images always seems to include “It’s so small – how did it cost so much?” (Planetary.org).
While this article is concerned with the Huygens probe, it applies equally to any space probe. Hopefully this should explain why a billion (considerably less in the case of Huygens) of your tax dollars (or Euros) result in images no larger than those of a decent webcam.
The tolerances these chips need to meet are considerably higher than anything within on or near earth.
This chip needs to be low noise (static), low light, low power, and low mass. They also need to be highly reliable and highly durable (10 years of cycling between +200 & -200ºC and withstand launch reentry and landing).
They need to be custom made to meet the first four requirements. Meeting the fifth and sixth requirements is a matter of selecting the best one of many.
For starters, these are class 0 chips, meaning they can have no column defects and no dead pixels. Yield rate for these is on the order of 1%. ESA (European Space Agency ) needs to make a number of these chips and select the best one. If they want to pick the best of 10 then they need to have 1000 chips made.
ESA was building this probe in the late 80s to early 90s. Nowadays it would not cost anywhere near as much as it did.
If you double the resolution then you quadruple the number of pixels and the chance of a defect. You also quadruple the power consumption and the weight.
This is all a moot point when compared with the one real factor: upload speed.
Huygens, or any probe for that matter, can only upload data at a fixed rate. Given the probe has a fixed length of time in which to transmit the data, you have an absolute amount of information you can upload. A certain amount of the bandwidth is devoted to the other instruments. This leaves you with a finite number of pixels you can transmit. There is no data redundancy with the pictures so you can double that number. Divide this number by the number of photos you want to take and you arrive at the maximum resolution of your imager.
Ian Anderson –
Custom Optical Systems.com