Topics

Autoguiding questions ...

Khalid Baheyeldin
 

I am using PHD2 with OnStep 3.16.

I started to image before the calibration was done, and saw something unusual

See this crop up

http://i.imgur.com/T2EsNS9.png

The fact that the star did not return back on the same line, is that a concern?

As a result, the stars are a bit oblong with 300 second exposures

http://i.imgur.com/iwRL9yd.png

If it matters, these are the raw polar alignment errors from INDI:

Az -192" Alt -480"

Khalid Baheyeldin
 

And RMS is considerably higher than usual.

RA 1.36
DE 1.22
Total 1.85

Khalid Baheyeldin
 

And here is the calibration data from a later attempt the same night

http://i.imgur.com/yeyTesb.png

Only difference is that this last one was binning 2, and the earlier one was 1.

RMS is horrible

http://i.imgur.com/4OBQIXy.png

Khalid Baheyeldin
 

Now nearer the pole, pointing at M101

Calibration (using binning 1)
http://i.imgur.com/Wjla7AA.png

And RMS is way better

http://i.imgur.com/AfkgDW8.png

But nearer the pole is always better. So I have no idea if there are other factors,
such as seeing (the forecast said average 3/5, so should not be the source for
all this).

All the above is exactly what I complain about when I say that autoguiding (and any
other additional components in a given system) increase complexity.

Khalid Baheyeldin
 

And just like that, autoguiding improved dramatically.

RMS
RA: 0.59
DE: 0.44
Total: 0.74

No idea why?



This is on M13 with 300 seconds @ ISO 200. Here is a cropped image.

Voodoo ...

Alex F
 

Could there be a relationship between the guiding south only in Dec in that last graph?

My guiding is always better overall when guiding in one direction in Dec with deliberate unbalancing in the opposite direction as Dec trends to drift similar to keeping RA unbalanced to the East.

Perhaps reversals in Dec direction when guiding is causing instability in the whole system as weight shifts around or tension is released in the worm.

These have been my thoughts anyway and for now keeping both RA and DEC slightly unbalanced along with a deliberate slight misalignment in my polar alignment to keep DEC trending in one direction seem to bring me down sub 1" RMS. I also have to let it settle a good 30s or more between dithering and a new exposure.

There's a great PDF here that helped me a lot
https://drive.google.com/file/d/1qkRRBqBpHHsKHBXioOE6-upq8R4WK5i0/view

Howard Dutton
 

This is with an SCT right...

Does it have mirror locks?
OAG or separate guide 'scope?

Khalid Baheyeldin
 

On Thu, May 21, 2020 at 08:24 AM, Howard Dutton wrote:
This is with an SCT right...
Yes. It is a C8 with a 0.63X reducer.

Does it have mirror locks?
No.

OAG or separate guide 'scope?
Separate guide scope with a QHY II L Mono.
Same I have been using since autumn 2019.

What bugs me is the erratic nature of guiding. One night it is good, another it is horrible.
Even in the same night, like last night, it could vary from horrible to great.

Khalid Baheyeldin
 

Alex,

Thanks for the tips and link.

Yes, I suspect balancing is an issue (if not THE issue).
So for RA, east heavy is better.
What about DEC? Make it heavy in what direction?

For the polar misalignment, do you always be a certain distance from the celestial pole?
How much in Alt and how much in Az?

hemantkulkarni.flickr@...
 

The issues with star not in same line, is the scope polar aligned correctly? Do you use the polar alignment feature in Phd2?

-Hemant

Khalid Baheyeldin
 

On Thu, May 21, 2020 at 05:42 PM, <hemantkulkarni.flickr@...> wrote:
Do you use the polar alignment feature in Phd2?
No.

Someone mentioned it before, but I don't know how tedious it is.
Do you use it?

Alex F
 

From the PDF I linked I follow this advice about DEC:
If you have southward drift:
- If the scope is pointing South of the zenith, imbalance the scope so that the North (lower) end is heavy.
- If pointing North of the Zenith, again imbalance so that the North (higher) end is heavy.
If your polar alignment is causing a gradual northward drift instead, then you reverse the imbalance:
- Pointing South of zenith, South/higher end heavy.
- Pointing North of zenith, South/lower end heavy.
As for misalignment I tend to ensure it's good but not spot-on, something like 5 minutes out on my ALT usually. My ALT has such a coarse thread on the adjusment bolts and it likes to overshoot whenever I try to fine tune so I leave this where it ends up. AZ is easy to get under 60" though. I'll be honest that I really don't know if having one or the other misaligned is preferable but the logic of keeping it out of the backlash by forcing corrections in one direction is sound like with the East heavy RA. I've turned to using a rough polar scope alignment followed by PHD2 drift alignment to get me there which also lets me see my drift direction during the process. Then following the above logic I move my dovetail bar about 1cm in the corresponding direction. I've heard some people stick a weights on too but I don't like the idea of adding even more weight to the system.

After this If min-move is set high enough in DEC then it will only guide in the one direction, You can force the direction in the settings as well by changing it from Auto to North or South, but find this messes up my spiral dithers which I keep large because I'm using a DSLR.

Dave Schwartz
 

It is tedious. It involves watching the trendline of the error in the graph with output disabled. You then adjust the mount mechanically try to steer the trendline as near to horizontal as possible through a number of iterations in two locations.

If something takes as much 'splaining as this https://sites.google.com/site/openphdguiding/phd2-drift-alignment I've got to think the OnStep 'Refine Polar Alignment' feature has to be easier. After all, this is the OnStep group.

On 2020-05-21 6:07 p.m., Khalid Baheyeldin wrote:
On Thu, May 21, 2020 at 05:42 PM, <hemantkulkarni.flickr@...> wrote:

Do you use the polar alignment feature in Phd2?

No.

Someone mentioned it before, but I don't know how tedious it is.
Do you use it?

Khalid Baheyeldin
 

Thanks for the tips.

Tonight it is far far better (at least for the time being).



It is not yet astronomical dark (~ 20 minutes more), and I am capturing the Leo Triplet.

What I did was balance the scope better to make sure it is east heavy.

OnStep 3-star alignment results are: Az +96" Alt -1856
Very different from yesterday.

I can't ever explain why they are different from one night to the next despite the scope never being moved in between ...

Alex F
 

The only real question is how did the images turn out? If you're happy with them you can be chasing down guiding problems forever with only small improvements to your final image quality which is what I caught myself doing. We're looking for good images in the end and the stats are largely irrelevant past a certain point.

This is nearly 4h of guiding on M51 from a few weeks ago, it was 1.4" RMS total, my graph is definitely not smooth and you can see the struggles with dithering which means I have to set a 40s pause time after each dither for things to settle again. But the finished image is something I'm pretty happy with (minus the blown out stars - in fact I'll probably re-process this one...)

Those numbers look great and the peaks are a bit tighter as well. That section where the DEC changes direction with several North corrections followed by a large jump look like backlash or stiction to me which then repeats as it guides South a little later which is where avoiding those DEC reversals shows its importance as look how much smoother the trend of changes is afterwards.

The RA spikes a bit though, for me the partial cure was switching to the Predictive PEC algorithm which did a good job of bringing it tighter. You'd probably be looking at an RMS value of 0.6-0.75 if that was brought under control. In this instance at least...

On Thu, May 21, 2020 at 07:38 PM, Khalid Baheyeldin wrote:
I can't ever explain why they are different from one night to the next despite the scope never being moved in between ...
I've put this down to the tolerances in the mounts we tend to use. I imagine at the higher end with better machined parts and tighter tolerances that things just remain more consistent.

It's like the advice in that link says: think about every possible point of movement and eliminate problems from the root (swarf in the drive/worn washer/worm end float) to the branches (looseness of any fixings/coupling of finder) to the leaves of the system (focuser slop/cable drag/fine-tuning balance) and then look at the guiding settings last of all. I'm imaging on ground that is mostly clay which settles, shrinks and cracks as it dries out and swells and sinks after a bit of rain so even when I put my tripod back in the same markers as before it's always a different level.

Once I was watching the CCTV camera on my scope for security reasons during imaging and a cat jumped onto the plastic box housing my power supply which bumped the tripod legs a little and there was a jump in the guiding. Stuff like that hammers home the sheer precision we're looking at. I dread to think what a nearby train line or highway would look like on my guiding graph!

I'm looking at building a Todmorden Pier perhaps when I get hold of a bigger mount and a larger OTA at some point down the line. I was inspired by this micro-observatory idea as well so I can eliminate some of that night to night variability that comes from taking everything down, aligning, balancing, routing wires etc.

Dave Schwartz
 

Last night I found that turning on the PHD2 'RA only' flag in the Advanced settings, Global tab really helped my DEC guiding. Often it seemed like a DEC dither was opening up the backlash and it resulted in some fairly large excursions as the drift (there's always drift) began to be noticed and then the backlash had to be taken up (with a resulting overcompensation). It seemed to take about 50 exposures until it settled down again. After turning off DEC dithering, the DEC RMS was pretty consistently under 0.4". I'm not sure what effect that will have yet on the images from the DSLR.

RA RMS stayed over double that because of the constantly running motor and gear train. You could also see the jittery effects of quick seeing changes (on the order of scintillation timescales) as well as larger, longer-lasting excursions due to slower, large-scale twists and turns happening in the atmosphere.

One other thing I leaned is to be careful playing around with the guiding parameter dropdowns below the graph... I had clicked DEC guide mode dropdown just to see the options it had. I closed it up and then moved the mouse to a browser window open on a second monitor without clicking on it. I noticed that my mouse scroll wasn't working so I clicked in the browser and it worked again however I did not notice that my scrolling had cycled to a random position in the DEC guide mode list, in this case, 'south'. It wasn't until the graph showed the DEC drifting off the graph with no corrections being sent that I wondered what the heck was going on and eventually noticed the change that had been made. Ruined at least one sub due to that. Probably works that way for many other dropdowns too so word to the wise.

On 2020-05-22 1:14 p.m., Alex F wrote:
The only real question is how did the images turn out? If you're happy with them you can be chasing down guiding problems forever with only small improvements to your final image quality which is what I caught myself doing. We're looking for good images in the end and the stats are largely irrelevant past a certain point.

This is nearly 4h of guiding on M51 from a few weeks ago <https://drive.google.com/open?id=1onPujZSrbNsafVsTcR3CouLJuOKci7zX>, it was 1.4" RMS total, my graph is definitely not smooth and you can see the struggles with dithering which means I have to set a 40s pause time after each dither for things to settle again. But the finished image is something I'm pretty happy with <https://www.astrobin.com/full/k6cx6z/B/> (minus the blown out stars - in fact I'll probably re-process this one...)

Those numbers look great and the peaks are a bit tighter as well. That section where the DEC changes direction with several North corrections followed by a large jump look like backlash or stiction to me which then repeats as it guides South a little later which is where avoiding those DEC reversals shows its importance as look how much smoother the trend of changes is afterwards.

The RA spikes a bit though, for me the partial cure was switching to the Predictive PEC algorithm which did a good job of bringing it tighter. You'd probably be looking at an RMS value of 0.6-0.75 if that was brought under control. In this instance at least...

On Thu, May 21, 2020 at 07:38 PM, Khalid Baheyeldin wrote:

I can't ever explain why they are different from one night to the
next despite the scope never being moved in between ...

I've put this down to the tolerances in the mounts we tend to use. I imagine at the higher end with better machined parts and tighter tolerances that things just remain more consistent.

It's like the advice in that link says: think about every possible point of movement and eliminate problems from the root (swarf in the drive/worn washer/worm end float) to the branches (looseness of any fixings/coupling of finder) to the leaves of the system (focuser slop/cable drag/fine-tuning balance) and then look at the guiding settings last of all. I'm imaging on ground that is mostly clay which settles, shrinks and cracks as it dries out and swells and sinks after a bit of rain so even when I put my tripod back in the same markers as before it's always a different level.

Once I was watching the CCTV camera on my scope for security reasons during imaging and a cat jumped onto the plastic box housing my power supply which bumped the tripod legs a little and there was a jump in the guiding. Stuff like that hammers home the sheer precision we're looking at. I dread to think what a nearby train line or highway would look like on my guiding graph!

I'm looking at building a Todmorden Pier <https://www.google.com/search?q=todmorden+pier&sxsrf=ALeKk02kVOmTskiPG2i1TVXmZmH2BjvVDA:1590162686200&source=lnms&tbm=isch&sa=X&ved=2ahUKEwja2K7Q6cfpAhVCRBUIHXJtAJcQ_AUoAXoECAwQAw&biw=1920&bih=938> perhaps when I get hold of a bigger mount and a larger OTA at some point down the line. I was inspired by this micro-observatory <https://stargazerslounge.com/topic/353296-carl-johan-ns-diy-observatory/> idea as well so I can eliminate some of that night to night variability that comes from taking everything down, aligning, balancing, routing wires etc.

Alex F
 

On Fri, May 22, 2020 at 10:58 AM, Dave Schwartz wrote:
Last night I found that turning on the PHD2 'RA only' flag in the
Advanced settings, Global tab really helped my DEC guiding.
I've read that RA only dithering can lead to 'walking noise' in one direction on the final image stack as fixed pattern noise gets smeared back and forth over the same image area again and again which is especially prevalent in DSLR images. Found an example here. Worth a try though as it's largely down to the sensor being used and keeping out of that DEC backlash eliminates a lot of problems.

RA RMS stayed over double that because of the constantly running motor and gear train. You could also see the jittery effects of quick seeing changes (on the order of scintillation timescales) as well as larger, longer-lasting excursions due to slower, large-scale twists and turns happening in the atmosphere
The last clear night I had last week and had such bad seeing from a strong jet stream I didn't even set the telescope up in the end. Must be nice in to be on top of Mauna Kea or Las Campanas sometimes! Maybe some adaptive optics would be worth it if I suddenly found an Anglo-saxon hoard in the garden...

Khalid Baheyeldin
 

Alex, Dave, thanks for the comments ...

All this makes me lean more and more towards the idea that autoguiding, though having definite
benefits, adds to the overall complexity (more moving parts, more knobs and dials, and so on).

One turn off is that the "understanding PHD2" document that was linked to earlier is 20 pages.

Since I never stack, I can turn off dithering and eliminate one big source of variability. But I did not
find an obvious On/Off option for that.

You can find my guide logs for the 20th and 21st here.

They were two very different nights. The former had turbulence and wind. The second was significantly
calmer. I don't recall seeing an airline trail staying in a large swath of the sky (from north to south west)
for such a long time.

Balancing was different from one night to the other too.

PEC is off in all cases, since past experiments found that autoguiding compensates for the +/- 2"
of PEC that I have (or was it +/- 4"? Can't remember, but it is quite low).

If anyone can glean anything useful from them, let me know.

Khalid Baheyeldin
 

Some examples
First, from the first night (May 20th), which was more turbulent and balancing was not so good.

M16 The Eagle Nebula 300 seconds 05-21T04:10



From the second (better) night (May 21st)

This is M66 180 seconds at 05-21T22:18


Leo Triplet 300 seconds, 05-21T23:07



Part of the Markarian Chain @ 300 seconds 05-22T00:11


NGC4565 The Needle @ 300 seconds 05-22T01:08


NGC4631 The Whale @ 300 seconds 05-22T01:52


M13 @ 300 seconds 05-22T02:45

 

Khalid,

Looking at these single frames you've taken, I really think you should
consider taking 20 or 30 images and stacking them. Deep Sky Stacker is
free, and the learning curve is not that difficult. There are a
multitude of programs to use to bring out the wonderful colors and
details, too. Processing can be a little intimidating at first, but if I
can do it, I'm sure it would be a breeze for you! You could always send
the stack TIF file to me, and I'd be happy to work my magic! (-:

Rick
Tucson

____________________________________________________________
Sponsored by https://www.newser.com/?utm_source=part&utm_medium=uol&utm_campaign=rss_taglines_more

Drummer Knew Iconic Album Was Pretty Good
http://thirdpartyoffers.juno.com/TGL3141/5ecc9c09e45861c097e49st01vuc1
Golf Coverage Prompts New Trump Attacks
http://thirdpartyoffers.juno.com/TGL3141/5ecc9c0ab7c01c097e49st01vuc2
Air Force Scraps Pilot Height Rule
http://thirdpartyoffers.juno.com/TGL3141/5ecc9c0a287eb1c097e49st01vuc3