Help me understand overscan coordinates

[joelshort]

I have a new camera and I'm using the overscan regions enabled.  I feel like it should be easier or more obvious but I'm just not getting the numbers involved. 

My camera has the following pixel dimensions:
Overscan - 4630 x 3694
Active pixels - 4524 x 3624  (overscan disabled)

I'm trying to calibrate some flats (4630x3624) with a master bias (4630x3694), and I'm running into a problem regarding the overscan region.  When I try to calibrate the flats I get an error message "Inconsistent overscan geometry."  So my settings are not correct.  I just don't understand the description of the boxes in the overscan settings.  Any ideas?

[joelshort]

So originally I was trying to calibrate flats with a Master Bias.  Apparently there is some issue with the master bias.  If I load up the flats in ImageCalibration and just have the overscan parameter enabled (no master bias) the process completes and I'm left with files that have been cropped to 4524 x 3624. 

I don't get it (clearly).  Both the flats and master bias are the same pixel dimension at 4630 x 3694.

[RickS]

Nothing stands out that I can see.  My KAF-16803 camera has an overscan area to the right of the image and the parameters are:

Image region:0 0 4096 4096
Source: 4107 0 9 4096
Target: 0 0 4096 4096

So long as all frames are captured with overscan enabled everything just works as expected.

[joelshort]

Thanks for the reply. 

All images (lights, darks, flats, bias) were taken with overscan enabled.  But whenever I try to calibrate any subs with a master (bias, flat, dark) I get the "Inconsistent overscan geometry" message. 

I don't know if the problem is related to my overscan settings, or my masters.  But the masters have the same pixel dimension as all subs.  This is truly maddening. 

[joelshort]

So I'm showing my ignorance here...I didn't realize the overscan calibration needed to be done on each individual sub prior to combining for the masters.  So I was trying to combine the raw bias into the master bias, and then run the overscan during flat calibration with a master bias.  I haven't done any actual calibrating yet but I believe this was my problem. 

[RickS]

I didn't realize the overscan calibration needed to be done on each individual sub prior to combining for the masters

Glad you figured it out.