Solar Mosaic tool/sequencer Output files Won't Stitch?

[admin]

Hi Chuck,

it's the mosaic alignment that needs to have happened in SharpCap for the 'Goto Sun' feature to include the offset, not the polar alignment. Basically, when you do the mosaic alignment SharpCap gets to work out the pointing error of the telescope, the rotation of the camera and the scale of the image (pixels per degree) - it can then use that data to do a correct 'Goto Sun' by cancelling the pointing errors that have been measured. So, yes, you can use ASIAir to polar align, just make sure you have done one mosaic alignment in SharpCap - after that the GOTO Sun should be central rather than just 'nearby'.

One of the 'fun' things about using a Quark is just the amount of adapters, extensions, etc that you need to get everything set up. I use a diagonal myself, with a 2" UV/IR cut filter on the front of the diagonal to reduce the energy getting to the quark (the sunlight is still quite out-of-focus as the front of the diagonal, so the filter receives the energy over a fairly wide area rather than a spot). I try to use the 2" connections for all the joins, but it's still hard getting everything firmly and securely attached!

Good luck with the new setup - as I say, try a mosaic with a very small number of frames per panel and focus on how the position for each panel works out to test things.

cheers,

Robin

[chuck666]

Just got done with my messes............... Not much fun today.
The new straight thru train worked after much work. I don't have much optical common sense compared to many other things. Haaaaaa
I did do your polar alignment routine. It's not as simple as the AIR; but I used to the AIR. I got a "good" rating with about 1 deg 30 asec error. It is extremely fuzzy as to the movements of the AltAZ controls. You do know that I am in GEM mode, Right?
At that point, I had to add the Quark and wait for warmup. Then I had 1 cloud for 30min. I use the tools GOto Sun command. And it wasn't there. But, you have explained that... TY. Now I really struggled to find focus today. The difference from with and without diagonal was way, way more than I expected. Eventally, I found it.... By then I had the mount controls all F'd up....... I did a flat like a good boy.. Started the mosaic and it complained that I was way away from where it thought it should be 6.x deg. Did I want to continue? I restarted and got the same message but went ahead. I did an alignment and checked the edges with the Gotos. About the same edge errors as previous. NOTE during all the above steps, I noticed that the tracking drift was as bad as before. So, the PA I did wasn't very good, I guess...... I proceeded with the mosaic, full 1000 images. ONE SUCCESS: no buffering!!!
But, the drifting was still enough to make the imaging points NOT like the planned graphic. I did get 8 ser files all on the disc. I will process them soon and see the results. What are others getting and what do you see?
Now, I have to ask/say something negative/////// This mosaic tool is wonderful in concept; but I have never seen anyone use it and NO users are commenting in this thread.... I just wonder why there is so little chatter. Perhaps Quark users never intent to do full disk? I'm NOT a quitter and I will continue tomorrow with an Air PA and see if it reduces the drifting.

Finally, I had a major sw error today with SC, I think. When I started it with the 317 update.. Everything seemed find. But when I tried to GOTO the sun from the tools menu, the mount didn't move. But the logs indicated it had? This is what happened several days ago and I thought was just odd ball glitch... Then I reinstalled a new beta and every thing was fine. TODAY. I did a repair install, same issue. Next I installed an older version, sameo. Then I reinstalled 317, sameo. I checked the mount continuously thru this. the mount responded to HC signals; but not to the onscreen mount control arrows. I went back to mount install window several times to see if anything happened. No change and then I said what happens if I use the arrows in the setup window. The mount moved and not only that, the mount controls now worked and I could find the sun and focus. Crazy.
Thought you should know.

A little venting today, sorry. But, your software is great and I really like it. And, I don't understand the drifting that I see with tracking being on.... No experience is an issue here.

Chuck, Clear Skies........

[chuck666]

Here is the outcome of the raw files saved as jpg..

9 files. They tell the story of why I'm confused as to what setting of mine is Wrong....

2 files in next email.

Chuck.

[chuck666]

The last two were separated.

We're so close. The composite one should have been it. But the panels were not acquired like the planned image????

Chuck

[admin]

Hi Chuck,

so close, and yet so far

Seeing the nearly complete mosaic helps though - I can see that it's not just a drift in one direction going on - it looks like the panels have all been rotated a bit from their intended positions, which may be a bit of poor alignment on top of the drift. I wonder if a drift happening during the alignment process might have a secondary effect of messing up the alignment calculations in this way (rotation not measured correctly). Hmm - in theory with a fourth measurement it might be possible to calculate the two drift rates as well as the current offset/rotation/scaling... The maths is probably scary though!

Have you worked out which direction the drift is in? If you just let the image drift a bit and note which way it moves, then move the RA then Dec axes and see if the drift lines up with the RA movement or the Dec movement.

Dec drift is an indication of polar alignment issues. RA drift is an indication of tracking speed being out. I think it would be worth checking that, just in case we are wrongly blaming polar alignment. Mind you, solar tracking speed is pretty close to sidereal - should only be about 2.5 arc minutes per hour difference. SharpCap will try to set the solar tracking rate on your mount, but maybe it can't?

From the description of the movement errors you were having, it sounds like the ASCOM driver was happily accepting commands from SharpCap (and not creating any errors), but somehow wasn't passing those commands on to the mount itself. You probably won't be able to diagnose anything about this from the SharpCap side, as from SharpCap's point of view, everything seemed to work. Sometimes you can get the ASCOM driver to make its own log, which might reveal something.

You can find a number of solar mosaic images in the first couple of pages of the 'Gallery' forum on here - for instance this one came out rather well : viewtopic.php?t=7118 and here's a 9 panel solar mosaic I did - viewtopic.php?p=36166#p36166

cheers,

Robin

[chuck666]

Let me just say a couple of things. And thank you for sticking with me. At 81, things are NOT as crisp as they once were. I started learning about telescoposy when covid hit. In the USA, our government sent us money to help with the problems it caused. Well, I really didn't need a lot extra money; so bought a telescope. Well the "black hole" never fills. HA.
Even though I have a Master in Engineering; I had little if any "Optical Common Sense". So things that are obvious to many of our friends, are not to me.

Today, I used the AIR to do a Polar alignment. I then I started SC. WHAT a difference. Far less misalignment in total in all aspects of this sequencer other pertinent areas of SC!!!

I just downloaded the files to my PC and haven't processed them. I bet you a pound sterling, ha, they will stitch............ Also, the edge Gotos were much closer to the correct edge and Goto Sun went to the SUN!!!!!!!!!!

Edit: I owe you a Quid, I guess. Got two halfs again. Attached. My mentor, Larry, is trying things in the background also. He says that a 200 overlap and that's what these files were also, fails? Is there anything peculiar with 200? Now, I noticed that I had lower contrast today. Larry said he though the seeing poor for his capture yesterday? I also tried to see if panels 7&8 would stitch and NO they wouldn't. 1&2 would though.
I may have the camera rotated differently, can't be sure. I been trying straight thru and diagonal. I'm back to diagonal config. What drift I saw seemed to be L&R. Watched your video, very nice. Is seeing a stitching factor in your analysis?

I haven't looked at the ser files; but all were quite close to the intended "circle" mark~~~~~~~~

So, I'll edit this in a bit with a FULL DISK view...........
Chuck

[admin]

Hi Chuck,

a little progress every time is better than none at all I guess!

Looking at your two halves, it looks to me as though there is enough overlap between them that they *ought* to have stitched together. Taking the sunspot group in the right half as a reference (in particular the two smaller spots just near the left edge of the group) - you can see that those spots and everything to the left is visible in the left half, and the spots and everything to the right is in the right half - there is complete overlap.

On the other hand, looking at the area just below/right of those two spots, there are clear bright regions in the right half which aren't visible there in the left half. That could be upsetting the stitching process, since the two have less similarity than they otherwise would.

Here's a question for you - how much processing are you doing before trying to stitch? Could the difference in those bright regions come from different processing of the panels before stitching, or is it in the video data captured? Probably worth looking back at the videos to see if you can spot the differences in that region in the raw data - those bright regions can come and go quite quickly, or another factor (variable thin cloud) could affect the image brightness.

Overall though, based on the visible overlap of the two halves and your report that there was less misalignment, it seems that the mosaic is now pretty much capturing the panels in the positions that it should and we are onto fine tuning.

There's nothing special about the overlap value of 200 (or any other) - it's just telling SharpCap how many pixels of each panel should overlap with the neighbouring panels. SharpCap treats that value as a minimum, so as long as it doesn't add more panels, SharpCap may find that a particular overlap can be 300 pixels and still meet the requirements for other overlaps and the 'safety margin' space around the edge. If you change the overlap value, you will see the planned panels move around (and possibly get more panels if you increase the overlap a lot). An increase overlap setting may help with stitching issues, as it will give more common areas between adjacent panels for the stitching application to work with - of course, if you get to the point that it increases the panel count from 8 to 12 then you end up paying the cost in extra data to process and extra time to run the mosaic.

cheers,

Robin

[chuck666]

Thank you, Dr. Robin...

To clarify, I only process raw ser images thru AutoStakkert to yield the tif files. Then to AP's panorama tool.

Can "poor" flats be the issue? I do do those prior to the Mosiac tool. I use the cereal box liner method with two sheets thick.. I have a flat cap in the delivery queue... It won't make it for today, though.
We have talked about this. But, to clarify, I reduce the gain setting from ideal 140 for the 432 camera to about 100? Is that enough? Then, strive for a centered 50% exposure? And I guess at the center or the sun to start. I always think my images look washed out. Today I am going to try and retune the Quark etalon setting. And do your centering trick, now that it should work?

Please comment.
Chuck

[chuck666]

Morning,
Looks like today is going to be cloud covered. It's Monsoon season here and the forecasts about clouds are "iffy" at best.
I received the "flat cap" from Daystar and watched their own howto yesterday. The cap is more opaque than 2 cereal box liners for sure. Their procedure narrow field flats, which is the view thru a Quark; is to reduce GAIN to "0" and adjust exposure to "90%"!. So, that's what I intend to do.
That will be after I use your "centering-Goto Sun" trick for the Sun's center... Side Note, I moved my setup back to my rear yard for my ease of use. I have less than an ideal location. I'm in suburbia and always shooting over my roof or the asphalt street next to my house. This time of year, 40 C is typical by Noon. I try to be done by 9 am and 30 C.

This is my environment and maybe I need wait 2 months when it's cooler and improved seeing? After setup, I plan on improving the Quark ban wide setting. As you know, the tuning knob will rotate 5 clicks CC or CCW. Currntly, my best setting is 3 CCW. I will iterate 2 and 4 to see if I can tell a difference. Any suggestions?????

A little more info in regards to the take 10 images of each panel and assess the quality of the shoot? Does redoing alignment reduce the variability or is it chance? And, in regards to the Gotos for each edge; how close is expected? I have never seen them goto the edge? And does repeating improve that or not?

Sorry, for the questions... I have found one YTer who did an howto and a lot of settings were NOT apparent. Apparently, NOT too many Quark users? It's such a great tool, that you have created. It is surprising????????

Clear Skies,
Chuck

[admin]

Hi Chuck,

the weather sounds like a real enemy that you have to fight against - fortunately it is still unusual for the temperature to get above 30 here on more than a dozen days or so each summer, and this year has been particularly cool so far. Still, it's been cloudy too, so no real win there...

I believe that the adjustment settings for the quark vary from device to device - the temperature alters the central frequency of the filter, and since the filters are not uniform by default, the adjustment is required to find the best setting for your particular quark. Apparently the temperature adjustment is an open loop thing (no feedback) - you turn the dial and the heating amount changes and the light turns orange for a set period of time before going green again. That means that differening conditions may mean different settings are even for your particular quark. I found a post/video somewhere of someone who had opened their quark up and traced the circuitry relating to the heater to find all this out...

I'd expect the GOTO edges to be close, but not exact - the key is that any error you see in the GOTO edge movements can be expected to also occur when running the mosaic, so you would want the safety margin round the edge to be bigger (maybe factor of 2) than the errors you see in GOTO edge checks.

cheers,

Robin