Lost star depending on imaging software? Low PC resources?

[rbote...@gmail.com]

Hello All

I have been using PHD2 for ages with no issues and one software suite (Voyager). Upon instruction, it acquires the guide stars (I use multi-star guiding) successfully and on it goes happily guiding. 

Recently I have been testing NINA with Advanced Sequencer and I am getting a lot of Lost Star - Low Mass (white screens) now.  Same hardware, no changes whatsoever and during clear nights (no passing clouds and guiding near zenith).   

I have a good mount which almost doesn't need guiding (1600GTO-AE) and the guidescope is a Baader Vario finder and guide camera is an Atik GP Mono.   I don't suspect the imaging hardware to be an issue as it works just fine with Voyager.

I am wondering if NINA uses more PC resources (which I think it does) than Voyager and this is perhaps making the guide camera skip frames and PHD identify this as Lost Stars?  The PC is a fanless Mele Quieter 2C (8GB RAM).  Do I need a better PC for NINA?

I take >4 second guide images and use Variable Exposure Delay.

Attached is the log from last night's session using NINA.

Thank you for any input!

Roberto

[Brian Valente]

Hi Roberto

PHD is unable to find a guidestar most of the time, so without that, guiding isn't going to get very far. 

I didn't see a calibration run or guiding assistant, although it doesn't report any calibration errors. 

Have you tried running PHD standalone with the same profile settings as NINA? This is just to confirm it's working standalone. If it does, your issue may be some additional settings in NINA's PHD configuration such as ROI

You aren't using any calibration frames, so I would start with adding a dark library. Also enable diagnostic image logging so you can see what it's choosing as a guidestar. It may be trying to guide on a clump of hot pixels. 

re: enough resources, I don't see any evidence in the debug log things are taking too long, You can pull up the task manager and watch for resource utilization such as cpu, memory, disk, etc.

Apart from that, my own opinion is that mini computer is underpowered. If you like the Mele consider a newer version with N100 chip or better. 

Brian

--
You received this message because you are subscribed to the Google Groups "Open PHD Guiding" group.
To unsubscribe from this group and stop receiving emails from it, send an email to open-phd-guidi...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/b9628427-a376-4a7d-b8b7-cd6f4bf62c1dn%40googlegroups.com.

--

Brian 

Brian Valente

astronomy portfolio https://www.astrobin.com/users/bvalente/ 

[Roberto Botero]

Thank you Brian. I thought I was using darks since I had taken some to create the library (admittedly some years ago). Same for the calibration, it has not changed in years as my setup is permanent and the AP mount just works. 

I'll check on running standalone. I still suspect the PC as NINA is sluggish. Works fine with Voyager though. I think a newer mini-PC is due then. 

Roberto

You received this message because you are subscribed to a topic in the Google Groups "Open PHD Guiding" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/open-phd-guiding/2MUEa_RK7gE/unsubscribe.
To unsubscribe from this group and all its topics, send an email to open-phd-guidi...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/CAJa45i60abbCUB6PpG8zANPYM%2BrZt3GeeEu-%3D81%2BKU%3D-uZSY5Q%40mail.gmail.com.

[Brian Valente]

calibration back in 2023? 

I'd say it's time to re-do the calibration as well ;)

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/671669fc.170a0220.39a6c5.74ca%40mx.google.com.

[Roberto Botero]

Last night I had very similar conditions to the night before (clear, rising Moon) but this time I imaged with Voyager. Same hardware, mini-PC, etc. I'm still to take darks again for PHD. 

Not a single star lost. Very strange! 

Logs for last night's session attached. Besides a star missing at the beginning of some images because PHD is trying to guide on a cached image, it all worked fine. No white screens, nothing. 

I'm getting a much better specd mini-PC tomorrow and will test over the next week. 

Thanks 

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/CAJa45i7X9sdW39c8D7JYbexEWJ8qLG-X16%3D_9LTWKcyi9-4ndg%40mail.gmail.com.

[Brian Valente]

Hi Roberto

Looks good. couple comments:

you might reduce the long variable delay to 7 seconds? 15 seconds seems like a long time, but then again you seem to be having good results. 

You might also look at reducing min moves. Dec seems a bit high, probably around 0.8", RA is just under 0.5"

>>>Besides a star missing at the beginning of some images because PHD is trying to guide on a cached image, 

not sure what you mean here. PHD does not use cached images to guide, that wouldn't make sense for this application. 

Brian

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/671755b3.170a0220.1986ff.b148%40mx.google.com.

[rbote...@gmail.com]

Hi Brian

Thanks again.  Yes, the problem is that last night didn't solve anything as there was nothing wrong before when I was using PHD with Voyager.  I have only found the Lost Star - Low Mass issue with NINA.   What I wanted to show is that under similar conditions, once software suite in conjunction with PHD was working for my hardware configuration and the other wasn't...which unfortunately is still the case.

I will look into the delay timings.  I am not sure how long I have had this for...maybe since the AP video on guiding with AE mounts!

As for my comment on the cached image, what I mean is that after Voyager finishes an image and instructs PHD to stop guiding, there's an image left in the PHD screen.  When guiding re-starts for the same object or another target, the (multi-star) guiding restarts with the same chosen spots as last target's.  This does not happen all the time but it does, resulting in lost stars (as none there) even when stars are visible in the image.

Roberto

[Brian Valente]

Roberto are you saying the logs below are from a Voyager run?

If so, that suggests there's a setting in NINA that may be causing your issues. Make sure you aren't using the ROI feature in NINA's PHD settings (check NINA documentation and/or discord for that)

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/5eebd092-fa22-4f9d-91c3-9b3820a01620n%40googlegroups.com.

[rbote...@gmail.com]

Hi Brian

Yes, exactly.  The first logs I posted were from PHD working with NINA.  The second set from a night after from its run with Voyager.  Similar weather conditions, exactly the same hardware.

I have 95% ROI set in NINA under equipment...I bet this is an issue then?  I don't think there's an option for this in Voyager.  Now that you mention this though, the AutoSelectFail FITS files in the PHD folder for the NINA session are all (black) and full size.

Roberto

[Brian Valente]

>>> Now that you mention this though, the AutoSelectFail FITS files in the PHD folder for the NINA session are all (black) and full size.

I haven't seen those, so I can't comment. maybe upload a few via file sharing service.

I would remove the ROI completely just to confirm.

Are you using the same profile between the two apps?

--
You received this message because you are subscribed to the Google Groups "Open PHD Guiding" group.
To unsubscribe from this group and stop receiving emails from it, send an email to open-phd-guidi...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/955aee7a-cd89-446b-b141-c7d58becfba5n%40googlegroups.com.

[rbote...@gmail.com]

Yes, same profile shared across apps.  Attached is a sample AutoSelectFail FITS file from the latest folder (20 October) which was when using NINA.  There is no folder for last night (when using Voyager).  Could this all be related to dropped (black) frames as I suspected originally?

[Brian Valente]

The frame you provided has no signal, not even noise. Do you have any auto select that aren't failed?

>>>Could this all be related to dropped (black) frames as I suspected originally?

I'm not sure what you mean by dropped frames, but this example is certainly a problem.

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/edeb8a26-53f7-4f3b-9123-f1d96351e8ddn%40googlegroups.com.

[Dale Ghent]

If configured with an ROI (by % of the frame) in NINA's PHD2 options, NINA passes that on as an argument to PHD2's find_star API call. The percentage is converted to the find_star's roi x,y origin pixel and length,width arguments after querying PHD2 for the guide camera's frame size.

While it's still not clear from the description, and if PHD2 is not throwing an error back at NINA after being passed find_star with an roi argument, my first thought is that there is perhaps some issue in the camera driver when it comes to doing a subframe exposure.

> To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/edeb8a26-53f7-4f3b-9123-f1d96351e8ddn%40googlegroups.com.
> <AutoSelectFail_2024-10-20_225347.fit>

[rbote...@gmail.com]

All the failed fits in that folder were black (and full frame - not 95% of pixel array).   Where would PHD save auto select files that are not failed?  Same folder?  I don't usually go looking for these as my system has always worked!

Thanks again for your help.

Roberto

[Brian Valente]

re: Dale

>>>there is perhaps some issue in the camera driver when it comes to doing a subframe exposure.

that is what i'm thinking as well

>>> Where would PHD save auto select files that are not failed? 

The diagnostic image logging feature is for any auto select frame

Are there any in there that aren't flagged as failed?

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/0feba850-2078-40da-bbc1-618f3b080ff5n%40googlegroups.com.

[rbote...@gmail.com]

I've never had any of that checked.  

I wonder how the folder for the 20th got filled....  And yes, they are all AutoSelectFail_2024-10-XX named.  53 of them.  There were obviously some successful ones it didn't save.

I will run a session tonight with NINA and will enable the above - both for lost-star as well as Auto-select.

As for the driver for the Atik camera, I can't comment.  It's an old (very old) GP Mono CCD but has never failed me.  The driver must be years old though.

Roberto

[Brian Valente]

Roberto you might check to see if there's an updated camera driver?

I did see a comment from Dale last year re: bringing NINA up to spec with current Atik SDKs.

There's also the atik ascom driver

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/7897a924-d674-4bcb-9ae5-a77e889aa181n%40googlegroups.com.

[rbote...@gmail.com]

Brian

I use the ASCOM driver.  Here's a snapshot of my Connect Equipment settings:

The driver is dated 2022 from this screenshot.  How can the driver be good with the Voyager app and not NINA?

[Brian Valente]

>>> How can the driver be good with the Voyager app and not NINA?

Well let's be clear: we are out of the PHD realm here, so i'm taking a stab in the dark. 

The ascom driver still relies on the underlying camera driver, so if there's a difference in those, they might load differently. 

the latest atik driver is dated 2024, so you might update to that 

https://www.atik-cameras.com/software-downloads/

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/264ce8fc-7ac3-472f-9217-f3028701ecc4n%40googlegroups.com.

[rbote...@gmail.com]

Yes, just saw a newer driver so downloading.  Will download and test with NINA and the frame saving settings on.

Thank you again!

[rbote...@gmail.com]

Driver installed (2024.7.29.46) and camera connects well.  20 minutes to dusk here.

[Brian Valente]

When nina launches PHD and has it connect the equipment, it will start looping

can you see any noise or differences between frames?

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/0a152f27-e8e0-4d12-ae61-6857728bd3bbn%40googlegroups.com.

[Roberto Botero]

Will confirm tonight but yes, there's an IR light in my ROR which is ON until imaging starts so I can see that.

You received this message because you are subscribed to a topic in the Google Groups "Open PHD Guiding" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/open-phd-guiding/2MUEa_RK7gE/unsubscribe.
To unsubscribe from this group and all its topics, send an email to open-phd-guidi...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/CAJa45i7tZ7otqgywB%3DWZU5JH%2BdWxBRsv4Rx7pTye97j9y1rDbg%40mail.gmail.com.

[Brian Valente]

🤞🤞

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/CA%2BqASPGfR2HCLB5kfZF-Mh1TWJt5LYHYLm1q5Kj59QVNTtKDvw%40mail.gmail.com.

[Roberto Botero]

Unfortunately, the driver didn't make any difference. Still Lost Stars almost from the get go. Switched to Voyager without restarting PHD and all working fine...

Attached logs and some sample FITS logged. Tried to image with NINA until about 19:50PM local. 

Roberto

-------- Original message --------

From: Brian Valente <bval...@gmail.com>

Date: 22/10/2024 18:06 (GMT+00:00)

To: open-phd...@googlegroups.com

Subject: Re: [open-phd-guiding] Lost star depending on imaging software? Low PC resources?

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/CAJa45i7icr4gS4B%3D_JA-y8F0MQrPrrJiPZb8SVjUbuKoym0uYA%40mail.gmail.com.

[Dale Ghent]

Not sure what to say. NINA has nothing to do with your guide camera. It doesn't control it, connect to it, or use it - that's PHD2's job. NINA just passes guide parameters and commands to PHD2 to follow via its API, and any errors there would be duly reported.

> The driver is dated 2022 from this screenshot. How can the driver be good with the Voyager app and not NINA?
>
> On Tuesday, October 22, 2024 at 5:25:59 PM UTC+1 bval...@gmail.com wrote:
> Roberto you might check to see if there's an updated camera driver?
>
> I did see a comment from Dale last year re: bringing NINA up to spec with current Atik SDKs.
>
> There's also the atik ascom driver
>
>
> On Tue, Oct 22, 2024 at 9:09 AM rbote...@gmail.com <rbote...@gmail.com> wrote:
> I've never had any of that checked.
>
>

> I wonder how the folder for the 20th got filled.... And yes, they are all AutoSelectFail_2024-10-XX named. 53 of them. There were obviously some successful ones it didn't save.
> I will run a session tonight with NINA and will enable the above - both for lost-star as well as Auto-select.
> As for the driver for the Atik camera, I can't comment. It's an old (very old) GP Mono CCD but has never failed me. The driver must be years old though.
>
> Roberto
>
>
>
> On Tuesday, October 22, 2024 at 4:58:59 PM UTC+1 bval...@gmail.com wrote:
> re: Dale
> >>>there is perhaps some issue in the camera driver when it comes to doing a subframe exposure.
>
> that is what i'm thinking as well
>
> >>> Where would PHD save auto select files that are not failed?
> The diagnostic image logging feature is for any auto select frame
>
>

> To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/67180485.050a0220.2ecc1b.1203%40mx.google.com.
> <PHD2_GuideLog_2024-10-22_174435.txt><PHD2_DebugLog_2024-10-22_174435.txt><AutoSelect_2024-10-22_193732.fit><event001_00232_2024-10-22_193740_StarLost.fit>

[Brian Valente]

these fits files look fine, plenty of signal

Are you sure these are from the NINA session? If so, it looks like you solved at least part of the problem

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/67180485.050a0220.2ecc1b.1203%40mx.google.com.

[Brian Valente]

The guidelog you provided shows pretty good guiding

there's no dark library, so I can see all the hot pixels ;)

On Tue, Oct 22, 2024 at 1:01 PM Roberto Botero <rbote...@gmail.com> wrote:

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/67180485.050a0220.2ecc1b.1203%40mx.google.com.

[Roberto Botero]

Hi Brian
Thanks again. Yes, all FITS I sent were from the NINA session. The guide logs extend beyond into the Voyager session as I didn't close PHD and simply reconnected to it.

To view this discussion on the web visit https://groups.google.com/d/msgid/open-phd-guiding/CAJa45i58G953atQ%2B6pOC8VRWkEoY-evK5S0DwSQCSW42wd59qw%40mail.gmail.com.

[Roberto Botero]

Here's a screenshot of last night's session and the lost star display from PHD2. 

[Bruce Waddington]

Are you trying to do dithering?  If so, how is that being done?

Bruce

[Brian Valente]

I also noticed you are running two copies of NINA simultaneously?

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/46a022a7-e3fc-4833-91fe-42744d3e22f0n%40googlegroups.com.

[Roberto Botero]

No dithering. Unless instructed by Target Scheduler. I haven't even got PHD running properly so not particularly worried about dithering!

Two instances are necessary for Target Scheduler to image with a synchronised tandem setup. 

Thanks

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/CAJa45i54iSBsN_qxq4GdFK7afDg29dKBw5qjQ2DY5MfWLj40uw%40mail.gmail.com.

[Brian Valente]

>>>Two instances are necessary for Target Scheduler to image with a synchronised tandem setup. 

that all makes sense, but it you have a lot of variables at play here. I would eliminate as many variables as possible and build back up.

Have you tried 

- using PHD2 standalone just to confirm it's working properly (I understand about Voyager, i'm talking only PHD2)

- using only one instance of NINA and doing a very simple test with PHD guiding using default NINA values, no target scheduler or anything else, just a simple target, slew/center, af, and start guiding

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/67192351.7b0a0220.375837.871d%40mx.google.com.

[rbote...@gmail.com]

Brian

Thanks again.  I will do so this weekend if I get a chance.  I still believe this is related to PC resources and NINA needing more than other imaging apps - at least for my current PC specs.  I suspect blank frames are being downloaded from the guide camera because of lack of bandwidth or something like that overwhelming the processor. 

Just for the record, Voyager also uses two instances to control my tandem setup so I have written Advanced Sequences in NINA to try to emulate what Voyager already does.  I think I am getting there but obviously need PHD to work...

Again thanks a lot for your input.

Roberto

[Brian Valente]

>>> I suspect blank frames are being downloaded from the guide camera because of lack of bandwidth or something like that overwhelming the processor. 

I don't think that's the case Roberto.The examples of failed fits you sent are readable files, have the correct dimensions, and have all the correct fits header information. They don't show problems with signal being corrupted or other problems: They completely lack any signal at all. I don't see how constrained resources could produce that, but again I really think you need to start by eliminating as many variables as you can and build up your tests until it fails and then you have some place to start looking. It's just guess work at this point

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/4b2390b3-351f-492e-ac5b-4bb01d894a8bn%40googlegroups.com.

[rbote...@gmail.com]

Well, PHD works as stand alone application.  Here's a screenshot of it looking at the twilight (not dark enough) and running 0.01s exposures:

[Bruce Waddington]

I'm not a NINA or Voyager user but I think your problems are consistent with something causing the scope to move while PHD2 is trying to guide.  That's why I was asking about dithering.  It might help if you uploaded a NINA log file that can be correlated with a matching debug log file from PHD2. 

Bruce

[rbote...@gmail.com]

And here's Sadr starting to show as a star:

[Brian Valente]

great, i'd say move on to a simple one install nina setup with a simple sequence to start guiding/stop guiding

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/d0936370-3a33-47a6-a262-4497e94157f9n%40googlegroups.com.

[rbote...@gmail.com]

Thank you Bruce.

Attached log files from yesterday's session for NINA (including Target Scheduler).

Roberto

[Roberto Botero]

And what do you know?! It's chosen a star and it's guiding (no blank images yet) with one instance of NINA and just running a subsection of my Target Scheduler's sequence:

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/63c25542-31b9-406d-a126-d5b8fc42005bn%40googlegroups.com.

[Roberto Botero]

Here it is 10 mins later still doing its thing!

[Brian Valente]

how about that ;)

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/67193670.050a0220.272203.a059%40mx.google.com.

[Roberto Botero]

What to do now? Try to run the two instances again?

This is baffling. 

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/CAJa45i5QYLe%3DFd6cT03PaQo7-EyQco_rT2thRmQjh0n_iA0VpA%40mail.gmail.com.

[Brian Valente]

That sounds like a good next step

Add one variable at a time, and evaluate results. 

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/CA%2BqASPEuAN6pizTNT49iZOTJ6PE2e9yvv_axzJAQTkgWT4g9ww%40mail.gmail.com.

[Roberto Botero]

Back to square 1!  😵‍💫

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/CAJa45i72vw1eNaBSUfh6vwNy%2BZc_M%2B%2BsC9hpOfXP%3DzZu%3DPATFw%40mail.gmail.com.

[Brian Valente]

So does running two copies of NINA cause the blank images?

Sounds like something that rhymes with camera connection contention

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/CA%2BqASPFe8rhPERyCuCpACKNKNjvjLQncvhyQLkswo_uZbCzEvw%40mail.gmail.com.

[Roberto Botero]

Yes. Something like that (whatever it means!)  It may be best if I test on a higher spec PC first and if it continues failing move the discussion to the NINA forum? As not a PHD2 issue really?
Thanks again Brian!

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/CAJa45i7adDXmAKs9S0R2D8pDZgZNXq3Daqh967T5RJi1DM3KQg%40mail.gmail.com.

[Brian Valente]

Yes, makes sense to move it there

Would be nice if you posted the resolution here to close things out

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/CA%2BqASPHvktPUF8Xwo2AEye74oB3djWHLyjdF6_tF6mgqcKL8ZQ%40mail.gmail.com.

[Roberto Botero]

If I survive Discord, I will.

Thank you again for your help (and patience) Brian!

Roberto

To view this discussion visit https://groups.google.com/d/msgid/open-phd-guiding/CAJa45i6ZrEVWKs-1yKf9VrAm-unfxRn0efBNsX-VMaBtpP3UkQ%40mail.gmail.com.

[rbote...@gmail.com]

I posted in the NINA Discord server but got no takers.   In any case, last night to make the most of a clearish night with a bright Moon, I tried with another guide camera.   I used my ASI174MM which is usually for solar imaging only.   NINA loaded PhD2 and guided with no issues.  Both with a single instance and two.
There must be an issue with NINA/PhD2 interaction with the older Atik GP (which is running with all latest drivers) and two instances.   Rather than going down a rabbit hole trying to find out what is causing this, I'm upgrading my guide camera...