Problem using SplitCFA - wrong image dimension?

[cmassa]

I am learning to process my newly acquired Ha images which were taken with a DSLR using a 12nm clip in Ha filter. It was recommended that I use SplitCFA on my calibrated files but I get this error:

File 67 of 67
Open E:/PI/Heart - Ha/calibrated/light/heart nebula - Ha-1_AT65_34.0C_Light_900sec_ISO800_2014-10-08_220844_HA_frame5_c.fit
Wrong image dimension: odd width [5202] or height [3465].
Close E:/PI/Heart - Ha/calibrated/light/heart nebula - Ha-1_AT65_34.0C_Light_900sec_ISO800_2014-10-08_220844_HA_frame5_c.fit

I get the same error for all files I try. The files were taken with a Canon T2i, converted to FITS by PI and then calibrated with master darks, flats, bias using the BPP.

I get the same result using the uncalibrated FITS or raw Canon CR2 files.

Any ideas?
Thanks

[lensman57]

Hi,

I will be posting a question with regards to a topic not far from your question but meanwhile this is the way that I do the image integration to avoid getting the noise from B and G channels in the R ( Ha ) Channel. I use BPP script.
Set the BPP to calibrate only and add all the usual files. Set the debayer to SuperPixel and carry on as per normal. Then once BPP has finished its stuff close BPP and go the directory that you specified in BPP. Now using Batch Channel extraction  on the Debayerd files, I think they have a suffix of d,  extract the Red channel only and dispose of the other two. The files will have a suffix of red at the end their names. Now open Staralign and register the RED files and then use image integration to combine them. You should  end up with a master Mono Ha file and you should have nothing to do with CFA configuration rather than setting it to SuperPixel for debayering. My prolem is whether the flats need to treated differently for this process as the B and G channels affect the exposure length of the flat and cause over correction.

Regards,

A.G