I have been struggling with my guide camera switching to 8-bit video mode with no intervention on my part. It looks to me that PHD2 still thinks it's getting 16-bit data so even a saturated star would be too low to be detected. I found one point where this occurred in the PHD2 debug log and found the following entry in the kstars log:I didn't see anything to tell me why the change occurred. I appreciate any help I can get.

The relevant log files are attached.

I noticed that when a new version of KStars gets installed sometimes the default configuration values file get used for a device. That is to say, the configuration is stored an XML files in the .indi directory and next to the .xml files there also are .xml.default files. Those .xml.default files contain a config setting for 8 bit instead of 16 bit and so the 16 bit setting gets overwritten. I usually save my camera config via the INDI Configuration Panel and then edit the resulting .xml.default file and make sure that that ALSO contains 16-bit.


HTH, Wouter

Wouter,

I thought I had saved the 16-bit configuration, and reviewing the config file confirms it. The default configuration is also 16-bit, but it seems to want to start at 8 (strange).

Very strange indeed! Perhaps this is something in the new ZWO driver that was included in the latest KStars update.

So the "Switching to 8-bit video" happens only for ASI120 cameras that fail to exposure in 16-bit. Then it automatically switches to 8-bit and stays that one. This was done recently (2020-02-17) because the default is 16bit now, but once 16-bit fails for ASI120 cameras, it switches to 8-bit automatically since users usually do not know why it fails.

There is no evidence at the time that this is happening and PHD2 doesn't know the bit depth is 1/2 so it can no longer find a star. My solution has been to disconnect PHD2, reset the driver to 16 bit, reconnect and go back to guiding. After two (maybe 3) rounds of this, I had no disruptions and the camera worked at 16 bits the rest of the session.

If I don't connect the camera in INDI but use PHD2's native driver, what do I lose on the guiding tab in EKOS. Is it only the guider image?

In my experience you don't loose anything. You can still connect the EKOS guide module to Phd2 and have it pass the imaging to it. As a side-effect this lets EKOS perform dithering while controlling Phd2.

Hi,
I have a similar issue. I just didn't want to start a new topic for it, but now you brought it up it's good to contribute:

I've had two occasions in which I had my nice and shiny data the next morning and found out that the ASI1600mm recorded everything in 8 bit, while the day before it was data as expected at 16 bit depth. In the mean time I did not change this.
It's happened some days later again (and again totally unexpected) I happened to catch it after a quick check of one of the FITS. I hate to make it part of my work-flow to check if all settings keep being set as I set them before.

I've made a screen capture of it. Of course I had no logging on, because who expects a failure, right?


-Really- tiring that Stellarmate is so unpredictable/ unstable!

The imaging camera settings can, somehow, start up in 8-bit instead of 16-bit in my experience too. The difference is, when it's changed back to 16-bit, it stays there.

As Jasem said above, for my guide cam, dropping to 8-bit is a "feature". After reading some of the discussion on the PHD2 forum, I'll just set my guide camera to 8-bit and leave it there.

Noticing your gain, it's nice we can specify gain in the capture module again.

Yes, they stay at 16 bit when I set them at that at the beginning of the night,

But why would it lose this setting the next time I start my telescope for a night of imaging? It's busy enough setting up everything. I should be able to trust that once I've set a setting, it's -set-.
Or at least if "Then it automatically switches to 8-bit and stays that one." happens, it should display a popup or something.
I don't know if it's displayed in the message box below (with the date/ time/ [type of message]). With me I can't make the text box larger (say, 5 or 6 lines) without either fully covering the screen or going back to zero lines. So it can pass by really fast and I'm not going to scroll through it all.

The switch to 8bit is limited only to ASI120 cameras. All other cameras will not fall back to 8bit when capture fails. If any camera switches back to 8bit by itself, then this is a bug that needs to be resolved. Is this random? can it be reproduced somewhere? Any insights appreciated.