PixInsight Forum (historical)
PixInsight => General => Off-topic => Topic started by: Nigel Ball on 2010 October 11 11:45:00
-
I've recent;y ventured into Mosaic territory O:)
I've got reasonable results using the Star Alignment and Pixel Math as per Juan excellent video tutorial. However, I have a question /comment.
From what I have experienced so far the Frame that is added to the source overlays/masks the source image where there is alignment. There is therefore a straight edge. Now straight edges are easy to see - great for the PixelMath operations but how about an option to have a jagged edge and or blending?
-
Hi Nigel,
Thanks for your nice words! Keep a few things in mind regarding mosaics:
- The manual frame adaptation procedure I described in the tutorial is no longer needed. StarAlignment has now an automatic frame adaptation feature that is much more accurate and efficient.
- You must always build your mosaics with linear images. Among other important reasons to work with linear images, the automatic frame adaptation algorithm won't work optimally with nonlinear images.
- For the frame adaptation routine to yield a seamless mosaic, individual mosaic frames must be accurately calibrated. In particular, flat fielding must be perfect, and if you have gradients, they must be fixed with DBE before building the mosaic. Irregular illumination will cause visible seams.
how about an option to have a jagged edge and or blending?
It isn't needed. The frame adaptation routine already applies the computed correction to the frame being added at each mosaic step. Or am I misunderstanding/missing something?
-
Thanks Juan
Indeed I had a slight gradient on the edge of one of the frames. I had quite an overlap so cropping the frames down slightly removed this problem.
I tried the automatic Frame Calibration as well as the Pixel Math method. The latter is more satisfying ;)
Anyway I've got my first mosaic ;D
-
Or am I misunderstanding/missing something?
No - I was trying to understand how the mosaic function works in terms of how much of each frame overlap ends up in the final mosaic. It appears to me that the target frame is laid on top of the reference frame at the appropriate angle and position - is this correct? Rather than blending the two