Silly Astrometric Solution Question - There has to be a reason.

Edward Smith

I was playing around with a set of LRGB data just seeing what happened when a set of pretty lousy data came out when run through WBPP.  I noticed at the end that it ran the Astrometric Solution separately for each channel (ie, 4 times).  Now the data set was fairly small and the Astrometric Solution was only a small piece of the processing time, but it had my scratching my head about why you would run each data set through the process separately rather than just doing it once and copying the data over to the other images.  All I could think of was that there may be very small areas at the edges that might be different in each of the channels and the time spent extrapolating the solution to the other 3 channels was equivalent to the time spent simply solving the the image again.  I guess it could also be that since this is simply a batch process and it is performing the same actions on what the program treats as 4 different data sets that it isn't worth the complexity to do something different when you get to that last step.

BlockHead

Astrometric solutions are computed for masterlights.

To apply your logic, you would assume that WBPP knows you are making more than one master set as a set of RGB images that as a triad will make a single color image. This is true that in this case you really only need to do the solution for one master and "copy" it over to the others that are registered.

However,

1. The general solution is to always calculate an astrometric solution for any image. They are unique. Copying one to another is an approximation which can be OK for our purposed but it is not a general solution.

2. Creating more than one masterlight frame is not a general usage of WBPP. It doesn't know what you are doing with the masterlights. It is just a pipeline. 

-the Blockhead