19-08-2021, 04:29 PM
I'm not 100% sure about this, but if you can scan with a spot that's exactly 1 pixel by 1 line and has uniform light distribution (or sensitivity distribution if it's a camera) then that's an ideal aperture. There's still the zero order roll-off which would need to be corrected with an aperture corrector. This provides HF boost but with linear phase, unlike simple filters. (A practical aperture corrector is a transversal filter with symmetric taps. Can be analogue or digital.)
If you had a very small scanning aperture then there would be no such roll-off but sensitivity would be poor.
Practical cameras and FSS need greater aperture correction since the spot doesn't have an ideal rectangular profile. A vertical aperture corrector requires a pair of 1 line delays. They used to be an expensive add-on in the old days.
A lot of this was understood in 1934, in Mertz & Gray's seminal paper: https://archive.org/details/bstj13-3-464
Stiil worth reading but the maths is quite tricky. There once was a time when I understood all of this properly. Now it's just an overall feel for what's happening.
If you had a very small scanning aperture then there would be no such roll-off but sensitivity would be poor.
Practical cameras and FSS need greater aperture correction since the spot doesn't have an ideal rectangular profile. A vertical aperture corrector requires a pair of 1 line delays. They used to be an expensive add-on in the old days.
A lot of this was understood in 1934, in Mertz & Gray's seminal paper: https://archive.org/details/bstj13-3-464
Stiil worth reading but the maths is quite tricky. There once was a time when I understood all of this properly. Now it's just an overall feel for what's happening.
www.borinsky.co.uk Jeffrey Borinsky www.becg.tv
