First Mosaic

ReaPerMan (Paul) · #1 08-12-2009, 09:29 PM

Out in the cold until 0400 this morning ... Just about to give up and the air cleared. I took a number of shots but I have had a go at trying to link them together in gimp. A little bit rough in places and could do with some more processing to bring out the detail and clear up the joins. Should jump on my daughter's computer and use CS4 instead. Anyway here is a Jpg to keep the size down. C9.25 with Celestron Neximage CCD and registrax 5

troypiggo (Troy) · #2 08-12-2009, 09:48 PM

Mate, download hugin. It's a free panorama/mosaic stitching software, very easy to use, and it'll fix/handle those seams no worries at all.

ReaPerMan (Paul) · #3 08-12-2009, 10:25 PM

Thanks Troy, Nice little program free too! Stitched together but it wants to know all those details about the camera Lens FOV etc not sure how that relates to the Neximage CCD. Anyway heres a quick restitch. Thanks very much.

Paul

mithrandir (Andrew) · #4 09-12-2009, 12:23 AM

Quote:

Originally Posted by ReaPerMan

Thanks Troy, Nice little program free too! Stitched together but it wants to know all those details about the camera Lens FOV etc not sure how that relates to the Neximage CCD. Anyway heres a quick restitch. Thanks very much.

Paul

For a NexImage hung off a C9.5, FOV is 0.49 arcsec/pixel. Sensor is 640 x 480 pixels.

Or does it want focal length (2350mm) and f ratio (f10)? It wants to know what spherical distortion to compensate for.

Getting those right makes a dramatic difference to the final image.

mswhin63 (Malcolm) · #5 09-12-2009, 02:54 AM

Try IMerge it is quick and also auto adjust picture brightness and colour. It is quite simple to use. Although had a go at Hugin, very compicated but it expect to be better than IMerge once i learn how to use it.

ReaPerMan (Paul) · #6 09-12-2009, 03:19 AM

Thanks for that Here is the latest run with those figures put in. I agree a much flatter result.

troypiggo (Troy) · #7 09-12-2009, 03:23 AM

It uses the FOV (degrees) to correct for lens distortions as Andrew mentioned. If you know your scope and camera details, you can work out the horizontal field of view. I think if you get it pretty close, it'll be sufficient. Looking good based on that shot you just posted.

ReaPerMan (Paul) · #8 09-12-2009, 03:41 AM

Thanks for all the help guys. I can see the Hugin is very owerful and can do some remarkable things. I think ive got that pretty close now but as Malcom suggest Imerge is damm simple i produced the same result in less than a minute. Working on the kiss principle i may stick with the simple option for now.

mithrandir (Andrew) · #9 09-12-2009, 09:00 AM

Quote:

Originally Posted by ReaPerMan

Thanks for all the help guys. I can see the Hugin is very owerful and can do some remarkable things. I think ive got that pretty close now but as Malcom suggest Imerge is damm simple i produced the same result in less than a minute. Working on the kiss principle i may stick with the simple option for now.

Your iMerge and Hugin versions are pretty similar.

iMerge comes from an astrophotographer and probably assumes a long focal length which does not need correction.

Hugin was designed for stitching frames taken with ordinary camera (ie relatively short focal length) lenses, so has to know what the lens was to correct for spherical distortion. The curved frame edges in your original come from it assuming the focal length is around the default full frame 35mm camera lens of 50mm and compensating for it.

Digital photos using the camera's own lens have the necessary information in the EXIF data, so you don't have to supply it. If you are using a DSLR on a telescope, either at prime focus or through eyepiece projection, the data will be wrong and will distort the mosaic so you have to override it.

CCDs don't supply the FITS or EXIF data by themselves, which is why you have to fill it in. I can't see any reason why it could not be added by the capture software, but the user would have to feed the figures to the program.

I had the same trouble the first time I tried stitching the moon with AutoPano. Using my NexImage with it's small FOV wasn't too bad. With my a200 it was terrible.

mswhin63 (Malcolm) · #10 09-12-2009, 10:00 AM

The real test would be to change the size of the images to match the same and compare the results.

Both look great though, well done.