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

[chuck666]

Morning,
Attached screen shot. You should see it when you start the mosaic applet?

Thank you for the method on histogram. I reviewed Daystar's video and they were saying No gain and 90%?

Doesn't your mosaic sequencer open looking like this, too?

Just curious?

Also, the processed half panel.

Later,
Chuck

[chuck666]

Robin,
Post #2 today.

I tried the centering routine this morning between cloud groups. Both times I tried, I got the same result. First, I selected the mosaic tool with the bottom edge in view and focused. It acknoledged finding the edge. I started alignment and it finds 3 points and displays the "next" button. I then select "goto sun", it brings up warning, I continue. It starts slewing and misses the sun and says, both times... "No Bright Areas Found"????? After the second time, I clicked the down mount arrow about 5 times and there is the sun...
For reference, this is what happens on startup... My manual polar alignment was off by about that same amount!
It's seems like it is NOT using the alignment data; but, using startup data.

Any ideas or mistakes on my part? I went back to post #17 and reread your howto. In it there is this parened note (correcting for any mount pointing offset, etc) That seems like what's happening? What are daytime imagers expected to do? I could, I think, manually move the tripod or the mount head controls to get closer to sun center? But, not all Quark users will be able to do that; I would guess?

It seems like I'm being a pest now and that's NOT my style?

[admin]

Hi Chuck,

no worries - these features are complex and the fact that they work for most people doesn't mean that they are 100% correct - it just takes some time to get to the bottom of why they don't work sometimes...

I think we are getting somewhere on the cut off buttons... here's what I see

Screenshot 2024-07-06 151910.jpg

Two things are noticeable here - mine is not cut off and yours has much bigger font for the 'N' and 'S' on the sun. I think that this means that you have set a bigger font size in Windows and that might be causing the issue. I will set my font size bigger here and see if I can repro the issue.

As to the 'Goto Sun' issue - I will need to test here - I have checked in my simulation code and it seems that it is doing what it should, but maybe something is different in real life. I will look at that one.

cheers,

Robin

[chuck666]

Thanks on Pest Control, LOL
I do everything PC from my recliner in my LR.. That's were my PC is and I use my SONY 75" 4K TV as its monitor. I'm a widower and bachelor and no one mindssssssssssssss. HA

I just checked and I had the font scale at 250%. I just switched it to 200% and restarted and almost all of the graphic was there. So that is the issue. From 12 feet away I need to go back to 250 to be comfortable. You can correct the software to handle this issue????

I tried something today to help my centering the sun issue when using the Quark. I did something wrong in my plan and didn't get the out come I wanted. I tried using a WL mask and NO Quark, I did a goto sun in the tool menu, and then centered the sun. Next I swithch out the WL and installed the Quark. I tried to center the sun in the on screen view using the manual mount controls like polar aligning. Struggled for 30 minutes and couldn't quite get there. By now it was 90F and wanted to try another mosaic, so quit that effort. Went to live view and tried to guess where center was and then did a flat capture using my cereal liners. I got something. Did the mosaic and it wanted 10 panels today. So, 12 minutes later had ser files.
Then came to my emails.
It was 118 F yesterday. So, try to finish solar before 8am. Ha.
Thanks for the continuing help.

Edit: attached my Master flat. Does that look typical? Its like a gradient file. I Don't see Quark feature differences. And the partial mosaic has a light center? Where is your gain setting? DayStar recommends "0" for the Flat Cap?

Edit: This is the 10 panel mosaic attempt.
I was noticing that the "run mosaic" tool was weird. But, I didn't stop it. Somehow the N and S poles on the graphic were showing as mirrored image?
Also, the captures were not as shown on the graphic; the cross hairs of the panel centers were different? As the mosaic capturing continued; at about the second half it started doing the first ones over or slightly different. It was chaoic. I let it finish and the reviewed the ser files and still processed them to tifs and reviewed them. They looked like a "drunk" just guessed where to take panels. They were ok images and even with lousy flats were usuable. So, I tried to stitch them AP...
They stitched but there was so much overlap in spots and the bottom third was totally missing. There were on images of that area.
I think that in my foray into trying get a centered sun for taking flats; I confused SC as to where it was. But, I really have NO clue. I'm calling it a fluke for now.
And try again tomorrow. It is mostly obvious to me that I can get a "whole disk" eventually.................
The colorize outcome of the stitch.....

Chuck

[admin]

Hi Chuck,

OK, firstly I will look into the partially vanished button on a 4K monitor - I tried on a normal monitor at 125% and 150% font size (hoping that 125% on a normal monitor might show the issue if it takes 250% on 4K), but it all looks fine for me. It might be a Windows version thing too, possibly - I have Win11.

The flat looks a little bright to the right and dark to the left, but actually the mosaic has come out without any obvious patterns of brightness that might be caused by a bad flat, so I think it has done a good job there.

No need to worry about the strange orientation of the 'N' and 'S' in the display after aligning. That display is trying to show the sun as it appears in the camera, and if you are using a quark with a diagonal then the fact that you have a diagonal mirror in the telescope will make everything appear mirrored, so your 'N' and 'S' markes will be mirror writing. That particular feature of the software is more important for the moon mosaic tool, where it should make the view of the moon on camera have the same orientation as the preview in the mosaic tool.

I'm not sure about the mosaic run that didn't go as planned - if you were watching the screen during the run, you might notice a magenta 'plus' symbol moving around on the mosaic tool sun preview - that plus shows where SharpCap thinks the mount is currently pointing on the sun and the main image with crosshair should roughly match that. It sounds as though things went wrong as you moved down to the later panels, but hard to be sure of why. Did you get a chance to use the 'Goto Top/Bottom/Right/Left' edge buttons to check the alignment before running the mosaic? That's always a good procedure if you are having any trouble.

cheers,

Robin

[chuck666]

Morning,
I am using Win 11 on my main PC. And, Win 10 Pro on the miniPC and Win 10 Pro on my laptop. The main PC is set to 250% and the laptop is 125%. I used the laptop this morning and was trying to polar align before the sun rose. I had issues as I learned some of its nuiances. I thought a couple of screens had graphics cut off on the left; but, I was not concentrating on that... I will retry tomorrow now that I have some experience.
I thought I understood from previous computer experiences; that there was sw code that handled those graphic variations between PC and users?
To be sure, I can't program a lick. The last time I did code was 1962 in Freshman college and it was Fortran F. haaaaaaaaaaaaaaaaaa.
I can setup to me in dual monitor mode here. The 4K and a std 32" monitor. I'll do that later....

EDIT: Well my 32" computer monitor is also 4K. That's the photo editing standared 32@4K......... I forgot. Of course, I got the cut off graphics. I did change the setting to 225% enlargement and the whole words are visable.... That's good.

I think you are correct about the flats. I also re-tuned the Quark and tried a different setting and really helped the contrast. I also ordered a Flat Cap from Daystar. NO eta yet.

I thought I mentioned the Plus sign and said it was significantly different than where the center mark on each panel being captured. In fact the last 2 panels captured weren't even on the sun. My guess: The software starts the next capture where the last location that mount tracking had left off. I don't have any idea if it adjusted for an offset amount? That last location was wrong by the amount of drift that occurred during the buffering of the images NOT yet transferred to the memory. I think. In my setup it is always wandering downward and goes off the edge eventually. MY overlap setting was 200 pixels. If I go more, I will have more panels.... Maybe that is OK?

I did try the Gotos for the 4 edges. None were NOT on the edges, all were about the same amount outside the disk......

My mentor is now trying this mosaic tool with his Quark setup. He can't get a stitch either. We are now thinking that our polar alignments are enough wrong that is causing poor panel imaging due to tracking drift. I tend to agree. But, since that centering trick of yours didn't work for me. I am not sure?

That's why we are both trying to find a better polar alignment routine. He and I won't leave our scopes out all night unattended. He's working polar drift control. I'm trying to get an early morning PA before the sun rises?

So, I have no images today. In my messing with polar alignment; I mixed up the mount control such SC wouldn't move the mount anymore. Even though it said/thought it had. I did all the logically things like shutting down, restarting, etc. Finally, I decided to download the latest version ogj SC and try that. The clean install cleared everything up.

I am a pain, sorry.

Chuck

[admin]

Hi Chuck,

I'm sure that really we only do this astro-imaging thing because it's a challenge - if it was easy then it wouldn't be nearly so interesting... So, it's no surprise that it throws up all sorts of issues.

I'm surprised that the pink cross is not on the solar disk for the final frames. Even if the mount has drifted, SharpCap doesn't know about that, so it would still think the mount is in the right location and show the pink cross where it has calculated it should go. I wonder if it might be worth you recording the screen during a mosaic run (there's nice,free, software for this - I use OBS Studio) and then sharing the video with me with some notes about when to look to see issues 1,2 and 3...

Polar alignment for daytime is a pain - I'm lucky enough to have an observatory, so don't have to worry too much, but if you need to setup each time then you have a couple of options...

1) If you can move the whole setup without disassembly then polar align the night before, mark the tripod leg positions and return the whole setup to that same location the next day

2) If your GOTO movements are accurate and the mount home position is accurate, then GOTO the sun and bring it into view using the Alt/Az polar alignment adjustments. The theory here is that an accurate GOTO on a polar aligned mount would put the sun in view. So, if the GOTO is accurate and you adjust Alt/az to put the sun in view, logically you must have polar aligned the mount

You can, of course (in theory) drift align on the sun, but drift alignment is tedious at the best of times

cheers,

Robin

[chuck666]

Thanks for the sad truth, haaaaaaaaaaaaaaaaa.

I won't PA the night before since my scope is exposed in the front yard for theivery. At least at this time of 115 deg days. But, I can PA late in the early morning and continue on after sun rise. What I have't accomplished in one trial was using SC for that task. But, I think I better understand it now. Any tips would be appreciated.

I been doing #2 but using the the mount control arrows to bring into view. Let me try the AAz controls. DUH.

New item. I have been buffering 820 of the 1000 frames with each mosaic panel. The scope continues to drift until the buffering is finished. I learned from by fellow solar mentor that he has NO buffering because he has a "data SSD" drive in his miniPC. I will be getting one to install shortly. That should stop the drift from going off disk and allow the "overlap" setting to cover any drift.
DOES THAT MAKE SENSE?

Finally, you indicated that you were researching the "GOTO SUN" button's lack of centering. That would help the day time PA? Any progress?

Thanks, Chuck

[admin]

Hi Chuck,

yes, an SSD for saving your data should help lots - ideally an internal one will be fastest. The very best are M2 PCIe ones, but SATA SSDs will still beat hard disks by a long way. With the SSD, the 'dead time' while writing out frames should be cut, making the mosaic much faster and less likely to drift. You can test this theory in advance of the SSD arriving by setting SharpCap to just capture 10 frames at each position (which should save quickly, even with a hard disk). I have had a request to continue moving to the next step of the mosaic while frames are being written to make best use of time, but that is actually quite hard to achieve, so it hasn't happened yet.

I did look at the GOTO sun movement and checked that it will use the alignment offset information once an alignment has been done. In the latest update I added an extra log entry to confirm that is happening and record the offset data, so if you try 4.1.12371, you should see in the log

Code: Select all

Slewing to Sun, using offset of <some RA/Dec offset> calculated from alignment data

when using that button.

cheers,

Robin

[chuck666]

Robin,
Good morning here.
Yesterday, I installed a 2 TB M.2 in my miniPC in its open slot. It is now the capture data drive D:.....
And, I happened to have updated to 371.

To confirm, I have to use SC PA to use that feature? Aligning with my AIR won't provide the offset? For me, using the AIR is easier at this time? I then would switch over to SC to continue Solar imaging work?

Hopefully you still are monitoring?

My Flat Cap comes tomorrow for better flats???????

I also reconfigured my imaging train to "straight thru" and removed the diagonal. That makes it 7 inches longer or 32" in total with the Quark?

Lots of things to test this morning.

Chuck