Mosaic Help: Won't Register

[Guest]

Started a 4-Panel Mosaic containing LDN 1622, Barnard's Loop and M78 in N.I.N.A using the framing Wizard with a 20% overlap.

I was able to combine the top and bottom frames, but have not been able to combine them. Not sure if it's because of my plate solve tolerance, 20% overlap, preprocessing, star alignment settings, but I have just not been able to stitch this last part. TOP.xisfBOTTOM.xisf

[Knight of Clear Skies]

Is that showing the edge of each frame? Looking at the asterisms I'm not seeing any overlap, so it's possible something went wrong with the framing.

Edited November 30, 2021 by Knight of Clear Skies

[tomato]

Funnily enough I had the same problem on Saturday. I set up a 4 panel mosaic in NINA and ran the sequence. When I tried to combine them I discovered that the panels had overlapped by 10%  on the short side of the frame but there was a 10% gap on the long side instead of an overlap, so the software would only combine two panels. Fortunately I had some fill in data from a previous run that enabled the mosaic to be completed.

I have checked the framing tool tonight on the same target and it all looks good so this is a mystery. A six panel mosaic planned and run in NINA back in October worked fine.

[Guest]

8 minutes ago, Knight of Clear Skies said:

Is that showing the edge of each frame? Looing at the asterisms I'm not seeing any overlap, so it's possible something went wrong with the framing.

It shows the edge, it's the center that's problematic.. I haven't altered NINA's settings, so maybe the 20% overlap was too small

[Guest]

5 minutes ago, tomato said:

Funnily enough I had the same problem on Saturday. I set up a 4 panel mosaic in NINA and ran the sequence. When I tried to combine them I discovered that the panels had overlapped by 10%  on the short side of the frame but there was a 10% gap on the long side instead of an overlap, so the software would only combine two panels. Fortunately I had some fill in data from a previous run that enabled the mosaic to be completed.

I have checked the framing tool tonight on the same target and it all looks good so this is a mystery. A six panel mosaic planned and run in NINA back in October worked fine.

It's the weirdest thing.. My first mosaic, so I could certainly be missing something, but I've briefly watched some videos on it and it looked easy enough. The only thing I've changed is the number of vertical + horizontal panels. Maybe from the NINA sequence the "centering" settings conflicts with platesolving somehow? I've also read something about the platesolving tolerance being too high, but I haven't had a chance to test it

[tomato]

You could be on to something. My previous successful mosaic was with a different scope/camera combination, my current setup has a much bigger FOV.

[Knight of Clear Skies]

10 hours ago, LeftyAstro said:

It shows the edge, it's the center that's problematic.. I haven't altered NINA's settings, so maybe the 20% overlap was too small

Yes, looking at the stars there doesn't appear to be any overlap in the centre.

[ollypenrice]

Knight is perfectly correct. Here's the proof. (I overlaid your two panels onto my camera lens image of Orion using the incomparable Registar and there is a wide gap between them.)

Like 50% of problems in astrophotography, this probably comes from a software malfunction. (The other 50% come from USB errors. )  My own view is, Why introduce software to do simple things you can do manually and, therefore, reliably?  The more software we introduce, the more problems we will have. The 35 panel Orion telescopic mosaic which Tom O'Donoghue and I did was put together without plate solving or mosaic wizards, etc. Anyone doing your project manually would see straight away that the bright star 2/3 of the way down the left hand panel is absent from the right hand panel.

I'm in danger of going off on a rant so I'll stop there!!!

lly

 

 

[Knight of Clear Skies]

5 minutes ago, ollypenrice said:

Anyone doing your project manually would see straight away that the bright star 2/3 of the way down the left hand panel is absent from the right hand panel.

In theory, yes, but this was my attempt at a mosaic of the entire Virgo Cluster.

I couldn't see what I was doing as it was awkward to review two panels on my laptop screen. My attempts to work out or enter target coordinates went very wrong somewhere.

I tend to favour simple low-tech solutions where possible but debugging an auto-mosaic setup could well be time-effective in the long run.

[ollypenrice]

1 minute ago, Knight of Clear Skies said:

In theory, yes, but this was my attempt at a mosaic of the entire Virgo Cluster.

I couldn't see what I was doing as it was awkward to review two panels on my laptop screen. My attempts to work out or enter target coordinates went very wrong somewhere.

I tend to favour simple low-tech solutions where possible but debugging an auto-mosaic setup could well be time-effective in the long run.

Well, you may be right but this short thread contains two failures already, the OP's and Tomato's. At least look at the panels to see if  the overlap is credible (unless you've gone to bed, of course...)

Olly

[Knight of Clear Skies]

1 hour ago, ollypenrice said:

Well, you may be right but this short thread contains two failures already, the OP's and Tomato's. At least look at the panels to see if  the overlap is credible (unless you've gone to bed, of course...)

Olly

Indeed. Or try a large mosaic with single sub 30-second panes first to check how reliably it works.

[tomato]

My very first mosaic of M31 I did manually, but that’s a really easy target to check, given the shape and the presence of the satellite galaxies. Something like Sh2-240 is a bit more tricky, you can hardly see the nebula on a sub and the star field is a bit bland.

I guess the software has made me lazy, and keen to be capturing subs rather than setting up. If I lived in Atacama or New Mexico, my approach might be different.

[Guest]

On 30/11/2021 at 09:53, tomato said:

My very first mosaic of M31 I did manually, but that’s a really easy target to check, given the shape and the presence of the satellite galaxies. Something like Sh2-240 is a bit more tricky, you can hardly see the nebula on a sub and the star field is a bit bland.

I guess the software has made me lazy, and keen to be capturing subs rather than setting up. If I lived in Atacama or New Mexico, my approach might be different.

I've learned that people rely on automation for different reasons and that's okay. I typically don't look at my subs until days later unless something obvious happens during session, like PHD2 notifying that I lost a guide star, because I do homework while imaging. Anyways, just getting around to a second try. Did a test run of a several 5 minute exposures for each of the 4 panels to see that they correctly overlap and they did. Turns out, the issue was that N.I.N.A defaulted the camera rotation to 0 degrees instead of it's correct orientation (might have to do with using the nightly versions, but not sure). Lowered the pointing tolerance and accounted for the camera rotation =  gap gone. Hope I could help

MS