Diagnostic image logging

[Gary]

Hi, I noticed what I think may be an error, on the diagnostic image logging when I set "until this count is reached" to say 5 and tick the checkbox, click ok, then exit PHD2, when I reload PHD2 next the setting has reverted to off and to default value 1

[Bruce Waddington]

This is intentional.  That particular option is used only for a specific purpose and it doesn't make sense to assume that anyone would want to do that across many nights of guiding.  

Bruce

[Gary]

Correct me if I'm wrong, I was under the impression this setting stops an unlimited amount of exposures being saved when a diagnostic event selected is triggered? For example, if I lost guide star, the first 5 images might be useful I can see what the exposures was coming in like at that time, but lets say guide star was lost for an hour because of cloud, I don't want to save 1800 exposures. If my understanding is correct I'm not sure why that setting would need to default back to off.

I tried to set this setting because multiple times PHD2 has flooded my PC with exposures, but I want the diagnostic imaging enabled so I can at least see what was happening when there was an issue 

[Bruce Waddington]

I'm afraid your understanding is mistaken.  When you're having problems like this, it's always best to first check the user guide:

https://openphdguiding.org/man-dev/Advanced_settings.htm#Global_Tab

For convenience, I'm showing the relevant text in-line here:

• 'Enable diagnostic image logging' - used primarily for product support and diagnosis of problems dealing with the guide camera or PHD2 star-recognition and measurement.  Guide frame images are captured and logged in a FITs format subject to the filter/trigger controls in the group-box.  Images are saved in sub-folders of the PHD2 logging directory with the date and time encoded as part of the sub-folder name. For example, a folder named 'PHD2_CameraFrames_2022-03-03-184542' refers to a logging event that began on March 3, 2022 at 18:45:42 local time. Individual guide frames are saved with filenames that indicate the time the image was captured and the reason the frame was saved.  Since the guide frames are saved in a FITs format, the header will include other useful information such as exposure time.  Because the logging function is primarily used for trouble-shooting, the image sub-folders are automatically removed after 30 days.  If you wish to keep the images for your own purposes, you should either rename the sub-folders or copy/move them to a different directory.  When logging is triggered by one of the "events" - e.g. lost star or large errors - a group of images (an image set) will be saved, centered in time on the image that triggered the event. This provides a record  of guider images that will show what the guide star and guide frame looked like both before and after the unusual condition occurred.  The various triggering and filtering controls are described below and are also shown in the tool-tips for the controls:

• 'All lost star frames' - logs the image set for any lost-star events, regardless of the reason for the lost star (low SNR, mass-change, etc.)
• 'All auto-select star frames' - logs the image set for any frames used for auto-selection of the star, regardless of outcome.  Note that any failed attempts to auto-select a star will always result in a logged image, regardless of choices made in the user interface.
• 'When relative error exceeds' - logs the image set when the star deflection on the current frame exceeds the running-average error by the factor chosen in the adjacent spin control.  For example, if the average (RMS) error is 0.5 pixels and the current frame's error is 1.5 pixels, the relative error is 3.
• 'When absolute error exceeds' - logs the image set when the star deflection exceeds the number of pixels specified in the adjacent spin control.
• 'Until this count is reached'  - logs images until the count matches the value of the adjacent spin control.  The counter is reset to zero when the limit is reached.

So you can see that, in most cases, diagnostic image logging works on the basis of triggering events which then result in saving an image set - a small number of images that document the event.  An image set is normally comprised of the prior two images before the event, the image that triggered the event, then two more images following the event.  The option for  'Until this count is reached' doesn't use this triggering mechanism, it's just a convenience for saying "I want to capture a sequence of <n> images" for some undefined purpose.

The diagnostic image logging feature has been in place for a long time, but at some point we made a mistake that caused the auto-deletion mechanism to not work correctly.  That means these folders will continue to accumulate if you have diagnostic logging enabled and you are getting a lot of these kinds of events.  That problem has been fixed for the next dev release, but in the meantime you can safely remove all the folders you don't want.  We're definitely sorry for the inconvenience.  I will also update the User Guide to clarify the typical number of images that are being saved.

Regards,

Bruce