Hi folks,
SGPro 2.5.1.14 is an official release! This release contains an overhaul of the auto focus and image history routine used to find stars, some better memory management and a bunch of bug fixes.
This release officially marks the end of support for versions of SGPro that predate 2.5.1.14.
All help files have been updated to reflect changes in this release (CHM file distributed with SGPro, Web Help and the PDF Downloadable Version)
If you use PHD2, this release requires the use of a of PHD2 2.6 or better. You can find the required PHD2 version here:
http://openphdguiding.org/downloads/
This release requires the use of a of ASCOM 6.2 or better. You can find the required ASCOM version here:
http://ascom-standards.org/Downloads/Index.htm
As always, you can download the official SGPro 2.5.1 release here:
http://www.mainsequencesoftware.com/Releases
Release Highlights
- Refactor of methods used to find stars in images. While slower, this method is much more judicious in its approach (it is now shape based instead of intensity based), WAY less likely to find false stars and can now deal with very out of focus stars from scopes having a central obstruction (the old AF routine had trouble with these "donut stars"). Some of you with very large format cameras (DSLRs and some CCDs) and under-powered computers may find it difficult to use this new methodology. We have tried to make it as fast as possible, but shape detection is computationally expensive any way you slice it. Users of the most popular CCD chip for astronomy, the KAF-8300, when binned at 2x2, should be good on just about any machine. This new code does not alter the way in which auto focus works (data points, step size, etc), just the way in which stars are found.
- Near complete refactor of the help file detailing how auto focus works and how to set it up. It is distributed with every copy of SGPro and can be viewed here as well.
- Dozens of important bug fixes (details below)
Important Notes
- If you are upgrading from SGPro 2.4 or prior: SGPro 2.5 contains a refactored settings storage system. While most of it will be transparent to you, there are some key settings that will appear to be “unset” (list is above) when opening sequences or equipment profiles created with SGPro 2.4 or earlier. In order to ease the transition into SGPro 2.5 and avoid invalidation of your current sequences we have provided an optional data migration tool. You can alter your sequences and equipment profiles by hand (to repair them and prep them for use in SGPro 2.5) or you can have the provided Data Migration tool do it for you (Tools->Sequence & Profile Migration). This tool will take previously global settings and apply them to your older sequences and equipment profiles. After running this tool, your old equipment profiles and sequences should work seamlessly in SGPro 2.5. More information here: http://mainsequencesoftware.com/Content/SGPHelp/SGPro25SequenceMigration.html
-
Important note for the new "Minimum star size" field with respect to targets and equipment profiles! This field is unique in that it is the only one in auto focus that can vary by target (the others apply to the entire sequence). In its current form it is not super-intuitive, so we will do our best to explain it here:
- The value that shows up in this field when you open the auto focus setting dialog will be based on which target you have selected in the sequencing window. It is easier (and more intuitive) to set via the focuser's docking module.
- For equipment profiles, the value saved in the profile will be used for the creation of any new target. If you change that value (in the sequence file) for a target, it will have no effect on other targets in the sequence.
- If you experience great difficulty with the new star finding routines because of its speed on your gear, you can still use 2.5.0.23 for a bit (we will keep it available for download). This version of SGPro uses this less reliable, but very fast, intensity-based centroid detection algorithm. It is important to note that many of the bug fixes below are applicable to 2.5.0.23, but will not be ported over (meaning they are only fixed in 2.5.1). Apologies for a sticky situation there...
- SGPro 2.5.1 has essentially removed "image caching" features that were designed to make review of the past 5 images very fast. This also meant that SGPro carried the last 5 images in memory. This is not a big issue for most cameras... Now that large format cameras are more commercially available at reasonable prices, we are starting to see issues that force us to take notice of our memory usage. They manifest themselves in cryptic errors like "System.OutOfMemoryException" and other strange issues (like just crashing). This does not remove any features... it just means that image review of recent images will always load a file from the disk (and take a little longer).
Change Log
2.5.1.14 Changes:
- SGPro in-sequence memory demand has been reduced by about 60% at the cost of slower review of past images (small price to pay). This may help a great deal for users that experience System.OutOfMemory type errors.
- Stabilized user based abort of centering
- Added better error checking for failure to change to the specified plate solve filter during auto center process
- Fixed a bug where the auto guider might attempt to reconnect equipment upon resuming a sequence.
- Fixed a bug where SGPro was no longer checking the PHD2 version for validity.
- Fixed a bug that might not clear the "Star Lost" flag after successful recovery of the guider (and cause the camera to immediately abort the new exposure).
- Guider error (in pixels) above the specified user threshold will no longer restart the frame if the sequence is already waiting to recover from "lost star" or "lost lock"
- Location copy / paste from Stellarium should now be language agnostic.
2.5.1.13 Changes:
- Internal build for memory testing
2.5.1.11 / 2.5.1.12 Changes:
- HFR calculation (auto focus and image history) will now automatically resize large images to make the process faster (at what expense is still TBD). KAF-8300 images at 2x2 binning will remain unaffected.
- Auto focus now uses significantly less memory during the processing of a frame.
- Near complete overhaul of the auto focus help documentation including a "how to set AF up" section (found in the local help file and here http://mainsequencesoftware.com/Content/SGPHelp/UnderstandingAutoFocus.html).
- Implementation of auto focus "Minimum star size" for profiles. The value used in the profile will be used as the default value for new targets.
- Fixed a memory leak that might potentially cause sequences with large numbers of frames to capture to lock up.
- New targets now use the text "Target N" instead of "Target Set N". I don't even know what a "Target Set" is...
- Fixed a bug that would not properly rotate camera after a flip if the user wants to ensure that a particular angle is always honored.
- Fixed image memory leak that could occur when navigating through images with the (green) forward and back arrows.
- Fixed a bug that would not always fetch all of a DSLRs ISO capabilities.
- Increased the display accuracy of RA / Dec seconds to 2 decimal places.
- Added RA/Dec Copy/Paste for Stellarium. Copy location in Stellarium and paste into the target settings RA field and SGPro will find, split and populate the location text properly.
2.5.1.10 Changes:
- Fixed an issue with DCRAW not working on (probably most) machines
- Hopefully stabilized "fast" DSLR exposures (if not, hopefully trace logging will lead to issues we are seeing)
- Fixed an issue that could attempt to run recovery even if it was off (the opposite of this was also true)
- Attempt to stabilize usage of the image history tool during execution of a sequence
2.5.1.9 Changes:
- Small tweaks to AF to improve performance
- Reworked the way in which DSLRs use DCRAW for RAW image extraction
- Fixed a bug where end of sequence actions from unsafe shutdown might not run properly
- Fixed a bug where losing the guide star might trigger an unnecessary sequence recovery (or failure) if PHD2 self-corrected after the message was sent.
2.5.1.8 Changes:
- If your camera supports it, the new auto focus method will now allow for fast image download. As part of the beta, we will ascertain the new AF routine's tolerance for nosier data (this is not a commitment to add it back as an option... just a commitment to try it out and see what happens).
- Increased the maximum star size (still a function of minimum star size). Both sizes are logged if you want to see them.
- Updated UI layout to support minimum pixel diameter (and remove old nebulosity rejection slider). Still not supported in the equipment profile manager.
2.5.1.7 Changes:
- Auto focus crop has been re-implemented for the new auto focus method (still just uses part of a full frame)
- Cleaned up the auto focus docking module GUI (the AF settings dialog is still not fully reorganized).
- Added the Alnitak flip shutter device.
- Modified camera angle adjustment in meridian flip to be a bit smarter. It will now query the running target or the user (in the case of a manual flip) to ascertain if it should attempt to validate and correct camera angle during the process.
- Additional trace logging for unexplained Canon / DCRAW crash
- Fixed an issue where an unsafe conditions shutdown could fail to execute end of sequence actions of the sequence was waiting on some type of user input.
- Fixed a bug where SGPro could crash if observatory option "Slave on Sequence start" was selected and the sequence was not connected to the proper equipment.
2.5.1.6 Changes:
- Fixed an issue where the AF minimum star size would not scale properly when AF binning was changed.
- Possibly fixed an issue that could cause file naming patterns to erroneously use the name of the selected target instead of the name of the running target (should be a pretty rare issue). Still looking at this...
- Fixed a bug that would allow the flip wait timer to continue after flip (or flip failure).
2.5.1.5 Changes:
- More changes to auto focus. Specifically:
- Replaced the "minimum star size" slider with a numeric entry field (the slider was always meant to be temporary). This will, among other things, allow you to be precise about this number.
- This new numeric entry box might be more permanent-"ish" (it is ugly as all get-out, but serves a purpose right now; aesthetics later)... depending on how the rest of the AF beta goes. This field is also smarter about binning... allowing you to set it for 1x1 frames and then scaling itself as you adjust to different binning modes for AF frames. This will hopefully prep the field for entry into equipment profiles as well (not currently implemented).
- More star rejection logic to help clean up false positives (specifically for data starved images).
- Minor changes that do not affect metrics, but should help with performance.
- Now opening the help file in the way not recommended by Microsoft (hopefully, will cure focus and Z-Order issues)... we'll see.
2.5.1.4 Changes:
- More changes to new auto focus routine (stabilization)
2.5.1.3 Changes:
- More improvements to the new auto focus routine
- SkySafari control now properly converts epoch to / from JNow
- Fixed an issue that could cause sequence image persistence (save) failure if the file naming components contained illegal file name characters. These characters are now auto detected and replaced with the "_" character (in the file name only).
- If the user chooses to associate SGPro with FITS files (this is not done automatically), double clicking a FITS file should now open SGPro and then automatically open that FITS file.
- Attempt to fix an error that could erroneously report that an image did not download in a timely manner.
- Added some code to try and prevent a very rare case in which sequence file saves might be incomplete.
2.5.1.1 Changes:
- More changes for the new auto focus method
- Added a minimum star size (in pixels) slider. This can be adjusted using the slider below "Nebulosity Rejection".
- The minimum star size is a "per-target" setting. You will likely want to have, for the same set of equipment, the same minimum star size (no guarantee here... just a rule of thumb). If this is true, we will find a way to work this into the equipment profiles later.
- Nebulosity Rejection is still disabled (not needed)
- General improvements on star detection technique
- Note: With this new method, you should not be dissuaded if it only selects a few stars from what looks like a field of eligible stars (the new routine, assuming proper minimum star size, is far more resistant to false star detection).
- AFPacks now take a screen shot of your AF graph automatically and save it with the FITS files that belong to that run.
- Fixed a bug that, when turning on the option to always finish calibration frames from the options dialog, might not actually allow this to happen.
2.5.1.0 Changes:
- Major refactor for Auto focus routines used to find stars