Two cameras, pixel size differs by 2, DSS wont stack - ?

[tooth_dr]

I've tried to process my data from my 1000d and 40d, and despite the same sensor, there is a difference in pixel size, so DSS rejects it.  I dont fancy buying registar, is it just a case of cropping 2 px off the bigger one?  If there any handy way of doing this, as I'll have 54 light, 30 bias and 30 flats to do?

 

3908 x 2602

3906 x 2602

[alexbb]

There might also be some small differences in the optical paths that might give you a slightly different scale and then you will need to use a smarter application to register your images.

However, if the scale is exactly the same, what you can do is to register and stack both sets of files from the 2 setups and then crop the larger stack by 2 pixels. You can throw the 2 stacks in DSS (the smaller original and the larger cropped), register and stack the stacks or uncheck the reference stack and register and stack (it will "stack" only the checked stack file) the stacks and then you can combine manually the stacks in another program.

[tooth_dr]

25 minutes ago, moise212 said:

There might also be some small differences in the optical paths that might give you a slightly different scale and then you will need to use a smarter application to register your images.

However, if the scale is exactly the same, what you can do is to register and stack both sets of files from the 2 setups and then crop the larger stack by 2 pixels. You can throw the 2 stacks in DSS (the smaller original and the larger cropped), register and stack the stacks or uncheck the reference stack and register and stack (it will "stack" only the checked stack file) the stacks and then you can combine manually the stacks in another program.

A lot of stacks in there lol!  Both were imaging at 522mm according to plate solving program, so hopefully the same imaging scale.

[tooth_dr]

Anyone? When I open in PS the raws from each camera has exact same pixel size 3888x2592, but in DSS this changes to 3908x2602 and 3906x2602?

 

Any ideas ?

 

 

[alexbb]

Stack the 2 sets individually, export them as .tiff, open them in PS, (maybe also convert them to a 16/32 bit integer representation), crop the larger one to the smaller's dimensions, throw them back into DSS and stack the 2 sets

[tooth_dr]

8 minutes ago, moise212 said:

Stack the 2 sets individually, export them as .tiff, open them in PS, (maybe also convert them to a 16/32 bit integer representation), crop the larger one to the smaller's dimensions, throw them back into DSS and stack the 2 sets

Thanks  Alex.  That was what suggested above.   I'm struggling to see how this works?  Will stacking the stacked sets of 54 each, be the same as if I had taken 108 subs from say one of the cameras and stacked them?  I cant see how that is the case?

[alexbb]

I don't expect significant differences. You have plenty of subs for the kappa sigma algorithm to be efficient.

Just choose "average" when stacking the 2 sets.

[tooth_dr]

Just now, moise212 said:

I don't expect significant differences. You have plenty of subs for the kappa sigma algorithm to be efficient.

But I'm stacking just 2 light frames at the final step?

[alexbb]

Just now, tooth_dr said:

But I'm stacking just 2 light frames at the final step?

Yes. These 2 will include all the other lights combined and their SNR is high. If they were composed of a different number of subs (different SNR), then you should have combined them at a different rate, but it's not the case

[tooth_dr]

Thanks Alex, I'll give that a go now!

[alexbb]

Welcomed! If the final result doesn't look good, you can import the 2 sets in DSS, set one as reference and uncheck it, then register&stack all checked (only one checked) with the "standard" composition (standard/intersection/mosaic). This will align your second set to the first one and output a final "stack" containing only the second set. You will be able to import it in PS and check if both were properly aligned (and have exactly the same scale).

[tooth_dr]

7 hours ago, moise212 said:

Welcomed! If the final result doesn't look good, you can import the 2 sets in DSS, set one as reference and uncheck it, then register&stack all checked (only one checked) with the "standard" composition (standard/intersection/mosaic). This will align your second set to the first one and output a final "stack" containing only the second set. You will be able to import it in PS and check if both were properly aligned (and have exactly the same scale).

 

It worked Alex.  The difference was superb, from 4 hours 30mins to 9 hours, really helped.  Thanks for the advice.