Interesting. So it's a heating element (wire) which serves as a controlled heater to keep the close environment above dew point. If it's controlled, that is. Maybe it works as a dumb heater also?

OTOH, I've never considered somewhat more performant (not to say extreme) cooling of CMOS in order to eliminate noise. I'm surrounded by CMOS and CCDs every day, in film cameras and telecines (both array and linear chips). I knew, and saw, that temperature can influence noise, but apart from staying within 'normal' operating range, I never considered it would eliminate more noise if significantly cooled down. Probably because I knew noise was inherent to the way those things work and not much could be done to eliminate it altogether. Interesting. I'll try to experiment with a bit more extreme cooling on my equipment to see what yield would come out of it. I'm not that concerned about dew, because equipment is in controlled environment (dry and +-2C deg. stable environment), but nice to know.