4.1 bugs
4.1 bugs
These may not be new bugs for 4.1, but that is where I noticed them. This was the last beta, not the one that just dropped today (2022/11/28).
1) The ADC tool and the Focus Assistant use the same ROI box, which is not what I want. For ADC I want the entire planet, but for focusing I have been told to restrict the focus ROI to within the disk. But changing the box size in one context changes it in the other.
2) In the Automatic Refocus tool, it asks you to select the focus ROI . . . but you can't any longer (I used to be able to). You cannot grab the focus box and move it or the edges or corners to resize it. It is fixed. And since it is now wrong because you changed the box for the ADC tool, the only option is to cancel the Automatic refocus, which leaves the focus assistant up, and then you can change the box size and move it and choose automatic refocus again.
3) WinJUPOS compatible filenames are incorrect. This is really bad for trying to make planetary derotations and animations. Here's an example from last night. Focus on the SER files, which were shot 3 minutes apiece, 1s between each, in batches of 5. The first capture started at 7:25:42 PM, or 00:25.7 UTC. You can see immediately that the filename is wrong, because it also says 0025_7. It should be the time from the center of the video, which is 1.5 minutes later, or 0027_2. If all of the filenames were from the start time instead of the center of the video that would be annoying but could be dealt with, but that isn't what happens, either. The filename, and the start (file created) and end (file modified) times for the first 4 videos are, for example:
0025_7 7:25:42 7:28:42
0028_7 7:28:43 7:31:43
0033_2 7:31:44 7:34:45
0034_7 7:34:47 7:37:47
The time spacings according to the filenames are 3 minutes, 4.5 minutes, 1.5 minutes. But in reality they are of course all 3 minutes.
1) The ADC tool and the Focus Assistant use the same ROI box, which is not what I want. For ADC I want the entire planet, but for focusing I have been told to restrict the focus ROI to within the disk. But changing the box size in one context changes it in the other.
2) In the Automatic Refocus tool, it asks you to select the focus ROI . . . but you can't any longer (I used to be able to). You cannot grab the focus box and move it or the edges or corners to resize it. It is fixed. And since it is now wrong because you changed the box for the ADC tool, the only option is to cancel the Automatic refocus, which leaves the focus assistant up, and then you can change the box size and move it and choose automatic refocus again.
3) WinJUPOS compatible filenames are incorrect. This is really bad for trying to make planetary derotations and animations. Here's an example from last night. Focus on the SER files, which were shot 3 minutes apiece, 1s between each, in batches of 5. The first capture started at 7:25:42 PM, or 00:25.7 UTC. You can see immediately that the filename is wrong, because it also says 0025_7. It should be the time from the center of the video, which is 1.5 minutes later, or 0027_2. If all of the filenames were from the start time instead of the center of the video that would be annoying but could be dealt with, but that isn't what happens, either. The filename, and the start (file created) and end (file modified) times for the first 4 videos are, for example:
0025_7 7:25:42 7:28:42
0028_7 7:28:43 7:31:43
0033_2 7:31:44 7:34:45
0034_7 7:34:47 7:37:47
The time spacings according to the filenames are 3 minutes, 4.5 minutes, 1.5 minutes. But in reality they are of course all 3 minutes.
- admin
- Site Admin
- Posts: 13358
- Joined: Sat Feb 11, 2017 3:52 pm
- Location: Vale of the White Horse, UK
- Contact:
Re: 4.1 bugs
Hi,
Ok, just spent some time looking at these
The Winjupos thing is annoying - the code to rename the file exists, but for some reason it is now being run after the information that is needed to do the rename has been cleaned up, so nothing happens. It used to run in the right order and I probably changed something else to fix another bug and this was the side effect. I will fix it for 4.0 and 4.1 for the next updates.
The movement of the ROI area while refocusing works when you run the refocus step straight from the sequencer, but not when it is launched from the menu or the focus area button - I should be able to fix that.
Separate ROI areas for the different tools is a bigger thing - will think about that longer term.
cheers,
Robin
Ok, just spent some time looking at these
The Winjupos thing is annoying - the code to rename the file exists, but for some reason it is now being run after the information that is needed to do the rename has been cleaned up, so nothing happens. It used to run in the right order and I probably changed something else to fix another bug and this was the side effect. I will fix it for 4.0 and 4.1 for the next updates.
The movement of the ROI area while refocusing works when you run the refocus step straight from the sequencer, but not when it is launched from the menu or the focus area button - I should be able to fix that.
Separate ROI areas for the different tools is a bigger thing - will think about that longer term.
cheers,
Robin
Re: 4.1 bugs
Thanks, Robin.
That separate focus and ADC box I have a workaround for (just try to remember to shrink the box back down after tuning the ADC before closing the tool) but it really should be separate.
That separate focus and ADC box I have a workaround for (just try to remember to shrink the box back down after tuning the ADC before closing the tool) but it really should be separate.
Re: 4.1 bugs
Had a few problems withe the Dec 5 beta last night. Entering Feature Tracking once, then trying to change ROI with the FT dialog box open both hung SharpCap. At one point FT calibration took Mars completely off the screen; canceled but the mount tracking seemed to have gotten mangled, as Mars would not set still; it either drifted rapidly in +RA or -RA. I had to turn off the mount snd re-align. Reports were submitted online for the hangups and I stopped trying to test the updates to FT at that point.
Seeing was all but perfect, essentially frozen. Edge detection with noise reduction = 2 still did not seem to be working. ROI was larger than the planet though; Fourier Detail method was working well so I did not mess with the edge detection method much.
Seeing was all but perfect, essentially frozen. Edge detection with noise reduction = 2 still did not seem to be working. ROI was larger than the planet though; Fourier Detail method was working well so I did not mess with the edge detection method much.
- admin
- Site Admin
- Posts: 13358
- Joined: Sat Feb 11, 2017 3:52 pm
- Location: Vale of the White Horse, UK
- Contact:
Re: 4.1 bugs
Hi,
did you send any bug reports in? I can see a few have come in for that version related to feature tracking, so if that was you then I can attack the hang/crash problem from the bug reports.
cheers,
Robin
did you send any bug reports in? I can see a few have come in for that version related to feature tracking, so if that was you then I can attack the hang/crash problem from the bug reports.
cheers,
Robin
Re: 4.1 bugs
Yes, there should be 2 from me I believe.
Re: 4.1 bugs
Attempted feature tracking again last night; crashed SharpCap again. Should be another report submitted.
And a new one:
When choosing Automatic Refocus from the menus, SharpCap hangs. Works fine if I choose "Refocus" after having opened up the focus assistant instead.
And a new one:
When choosing Automatic Refocus from the menus, SharpCap hangs. Works fine if I choose "Refocus" after having opened up the focus assistant instead.
- admin
- Site Admin
- Posts: 13358
- Joined: Sat Feb 11, 2017 3:52 pm
- Location: Vale of the White Horse, UK
- Contact:
Re: 4.1 bugs
Hi,
Ok, good to know that the feature tracking bugs are your issues. I have a fix for that, so hold on until Monday.
I will check the refocus thing to see if I can make it happen here.
cheers,
Robin
Ok, good to know that the feature tracking bugs are your issues. I have a fix for that, so hold on until Monday.
I will check the refocus thing to see if I can make it happen here.
cheers,
Robin
Re: 4.1 bugs
Great.
By the way, regarding WinJUPOS filenames. It seems that this has indeed been fixed. However, there is still a problem with the filenames that SharpCap generates. SharpCap generates filenames in this format:
2022-12-10-0123_3.ser
Note the underscore instead of the hyphen before the tenth-of-a-second position. I don't believe that this is standard WinJUPOS format, and indeed, when you import stacks with these filenames into WinJUPOS, the tenth-of-a-second position is lost and the resulting time becomes 0123.0 instead of 0123.3. This then has to be corrected manually, and if you forget, the derotation will be off.
By the way, regarding WinJUPOS filenames. It seems that this has indeed been fixed. However, there is still a problem with the filenames that SharpCap generates. SharpCap generates filenames in this format:
2022-12-10-0123_3.ser
Note the underscore instead of the hyphen before the tenth-of-a-second position. I don't believe that this is standard WinJUPOS format, and indeed, when you import stacks with these filenames into WinJUPOS, the tenth-of-a-second position is lost and the resulting time becomes 0123.0 instead of 0123.3. This then has to be corrected manually, and if you forget, the derotation will be off.
- admin
- Site Admin
- Posts: 13358
- Joined: Sat Feb 11, 2017 3:52 pm
- Location: Vale of the White Horse, UK
- Contact:
Re: 4.1 bugs
Hi,
Thanks for spotting that!
just had a dig through the WinJupos documentation, and that seems to suggest that the separation between the minutes and the tenths of minutes should be a '.', and that the dot itself is optional. I will test this and fix as appropriate.
cheers,
Robin
Thanks for spotting that!
just had a dig through the WinJupos documentation, and that seems to suggest that the separation between the minutes and the tenths of minutes should be a '.', and that the dot itself is optional. I will test this and fix as appropriate.
cheers,
Robin