calculate real focal length based on astrometry.net blind solve

[Josh Burkard]

the configured focal length is essential for plate solving with local applications (ex. All Sky Plate Solver). when Blind Solving using Astrometry.net the real Pixel Scale will be returned. with this Pixel Scale and the camera Pixel Size, the real focal length could be calculated

for example:

[Double]$PixelPitch = 2.4 # µm / pixel
[Double]$PixelScale = 1.03 # arcsec/pixel 
[int]$FocalLength = [System.Math]::Round( ( ( 3600 * 360 / 2 / [math]::PI ) * [Math]::Atan($PixelPitch / (1000 * $PixelScale ))), 2 )
$FocalLength # returns 481 mm

then the real focal length should be set to the settings, so that after this, the fits-file contains the correct focal length and the fits-data can be used for local plate solving applications

[Dale Ghent]

This would be good for a separate “Find my focal length” kind of tool. This isn’t necessarily an astrometry.net specific thing either; we can get pixel scale from any supported solver. ASTAP will even gripe if the FL it calculates is sufficiently different from what was specified. Though, I wouldn’t integrate this kind of mechanism directly into the regular in-band plate solving logic because of one critical point of unreliability: DSLRs.

Unlike the typical astrocam SDK, both the Canon and Nikon SDKs do not provide pixel size for the camera, so users of those cameras must supply their own value based on whichever source is authoritative on this aspect. We’ve found that a good number of people get the pixel size for their camera wrong for some reason and this will of course throw off the focal length calculations. It’s annoying but it’s what we have to deal with, and committing to maintaining our own database of camera model:pixel size mappings is a business that I don’t think we want to get into.

[Stefan B]

Removing version: 1.11 Nightly (automated comment)