Is my setup simply mismatched (for PHD2)?

165 views
Skip to first unread message

Stefano O

unread,
Jun 3, 2021, 9:11:05 AM6/3/21
to Open PHD Guiding
Hello,
given the difficulties I am finding in attaining good guiding at all, I wonder if my setup is hopelessly unbalanced, or so enough to make it really hard to get good guiding unless the sky is super steady (which is almost never in my case).

-- main OTA & camera:
Meade (LX-850 AFC) 10", focal length ~2000mm, F/8
focal reducer 0.63
Atik Titan camera with pixel size 7.4 microns

according to astronomy.tools calculator, the setup amounts to a resolution of 1.19 arcsec per pixel

-- guider:
iOptron's CEM70g iGuider, having the following specs
aperture 30mm, Focal length 120mm,  Imaging sensor 1/3 in CMOS, Pixel size 3.75μm,  Resolution 1280X960

again according to astronomy.tools, this gives a 6.45 arcsec/pixel

Given that a single pixel of the guiding camera that PHD2 sees translate into a whopping 5 or 6 pixels of the image I'm trying to take.
Or, that avg seeing conditions of 2 or 3 arcsec mean that the guider still sees it all in 1 pixel.
If I hope to get pinpoint stars say of a cluster one day,  I have to push MinMo's to 0.1/0.15...
Is this too much to ask for PHD2 or in general in order to avoid just always simply chasing seeing?
If so, what could I do best, considering iGuider is embedded in the iOptron's mount? Binning it 2x2?

Thanks
-s

mj.w...@gmail.com

unread,
Jun 3, 2021, 10:50:17 AM6/3/21
to Open PHD Guiding

bw_msgboard

unread,
Jun 3, 2021, 12:14:19 PM6/3/21
to open-phd...@googlegroups.com

If you want any kind of informed advice, we will need to see some data.  Try following this procedure:

 

https://www.dropbox.com/s/os1thorvswmzaul/How%20to%20create%20a%20baseline%20for%20guiding%20results%20using%20OpenPHD2.pdf?dl=0  

 

 

You don’t need to build a new profile but you should reset all the guiding parameters to their default values.  We need to see the Guiding Assistant output and some reasonably long guiding sessions where you aren’t fiddling around with anything.

 

Bruce

 


--
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/a0803c6e-705c-4266-93eb-e3b84d32a967n%40googlegroups.com.

S O

unread,
Jun 3, 2021, 1:27:41 PM6/3/21
to open-phd...@googlegroups.com
Thank you Bruce -- I'll follow up.
I am aware of the directions in asking questions -- I thought the problem I posed was well defined and self contained, but it seems not to be so after all.

I'll update after I have a chance to practice the recommendations.
Thanks again
-s

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/hAWd0S0JpJI/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/84BA003507A141379D994C501B8765E3%40HomeDesktop.

bval...@gmail.com

unread,
Jun 3, 2021, 1:49:32 PM6/3/21
to Open PHD Guiding
Hi Stefano

Generally speaking i think there are a couple "mismatched" things in your setup:

at 2000mm focal length, i would only be using an off axis guider. Differential flexure with a separate guidescope at those kinds of focal lengths are going to make your life miserable

very generally speaking, i would not go past 4x difference between guiding image scale and imaging image scale for the reasons you mentioned. But the OAG recommendation trumps this, since you would be guiding and imaging at the same focal length and very close to the same image scale when using OAG

Stefano O

unread,
Jun 3, 2021, 2:28:51 PM6/3/21
to Open PHD Guiding
Bruce,
thanks again.
Indeed I wrote my post looking for some pieces of evidence in that direction, after having read a number of things around. I had seen recommendations to stay below 4, max 5x, as you say. And people suggesting OAG.

Some other people though seem to rather forcefully warn about "not falling for OAG", so to speak -- not sure why yet: greater difficulty of finding a good guiding star? If so, I suppose that in my case, adopting OAG for sure would mean also changing the camera to something that allows a much larger FOV so that I can hope to pick a bright enough guiding star?  The Atik Titan chip is really very small and makes it difficult to pick a bright star most of the time.

One last further question based on your feedback: is differential flexure still an issue in my mount, given that the embedded guide scope and camera are housed in the mount's "top" (Dec) plate, the same that also directly hosts the OTA's dovetail? Essentially they are both in the same "block" (and I give it for granted that the OTA is rigid wrt to the mount's head once it's locked in its place there!).
Thanks
-s

Brian Valente

unread,
Jun 3, 2021, 2:39:49 PM6/3/21
to Open PHD Guiding
This is Brian, not Bruce ;)

>>>Some other people though seem to rather forcefully warn about "not falling for OAG", so to speak -- not sure why yet:

Maybe they can tell you more specifically why. I think that general comment is bogus internet lore or from people who are less comfortable with OAG, i don't know for sure.


I don't really understand your comments about your setup, perhaps you can post a couple pics of it. If you have a separate guidescope, you will have flexure, the only question is how much. 



>>>adopting OAG for sure would mean also changing the camera to something that allows a much larger FOV so that I can hope to pick a bright enough guiding star?  The Atik Titan chip is really very small and makes it difficult to pick a bright star most of the time.

Usually OAGs are limited by the pickoff mirror, not the camera sensor size. In the past I fell for that notion, but found larger chip guidecameras (like ASI174mm) just resulted in a good portion of the sensor being outside the pickoff mirror.

guidecameras are so sensitive nowadays it's rarely the issue. Good focus and correct camera settings are important too


--
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.


--
Brian 



Brian Valente

bw_msgboard

unread,
Jun 3, 2021, 2:43:38 PM6/3/21
to open-phd...@googlegroups.com

Hi Stefano.  This is all interesting and factual but it likely has nothing to do with your problem.  It’s a good example of why we want to see live data.  The hallmark of differential flexure is that the guiding numbers are good but the stars in your main images are elongated.  You haven’t described your problem that way so there’s no way to know whether this generic advice is relevant.  If you send the data, we will be able to tell you something specific to your problem.

 

Regards,

Bruce (not Brian V)

 


From: open-phd...@googlegroups.com [mailto:open-phd...@googlegroups.com] On Behalf Of Stefano O


Sent: Thursday, June 03, 2021 11:29 AM
To: Open PHD Guiding

--

Bryan

unread,
Jun 3, 2021, 8:28:56 PM6/3/21
to Open PHD Guiding
Stefano

This was posted a few years ago.  While it does not address all the aspects of guiding that impact the image, perhaps it helps answer some of your questions.


Bryan

Stefano O

unread,
Jun 4, 2021, 4:19:12 AM6/4/21
to Open PHD Guiding
Brian and Bruce, sorry for mixing up your names!
Thanks to you and Bryan for all the feedback this far.  I don't get to use the scope very often due to lack of time and weather being usually not ideal, but as soon as I get a chance, I'll practice your recommendations and get back with the data.

Brian, sorry I didn't explain the setup well. My CEM70g has a "Vixen/Losmandy-D dual saddle" to receive and lock the main OTA's dovetail. Within the same saddle you find the iGuider miniscope and camera. See attached picture please, and notice the "iGuider" cap at the left - that's the backend of the embedded guider scope+camera system.

Thanks much again
PS: I'll look up the docs to try and understand how PHD2 can actually issue any sub-pixel guiding corrections... but if you know the explanation is not there, the question is: how can PHD2 possibly compute any useful correction when, as in my case of 6.45 arcsec/pixel resolution in the guiding camera, even seeing conditions that smear a star image over 2 or even 3 arcsecs mean that the guiding star remains within the same pixel of the guider's camera all the time?
CEM70g_saddle.jpg

Brian Valente

unread,
Jun 4, 2021, 9:42:11 AM6/4/21
to Open PHD Guiding
>>>  but if you know the explanation is not there, the question is: how can PHD2 possibly compute any useful correction when, as in my case of 6.45 arcsec/pixel resolution in the guiding camera, even seeing conditions that smear a star image over 2 or even 3 arcsecs mean that the guiding star remains within the same pixel of the guider's camera all the time?

You need to be looking at the guiding resolution in arcsec, not pixels, and judge from there.PHD is able to guide to very small pixel fractions, but ultimately your image scale in arcsec is what matters when translating to how it will impact your imaging

I made this point to you several messages ago regarding OAG - i think your image scale on your guidescope is too coarse for your imaging ota. 



George Shoup

unread,
Jun 4, 2021, 10:07:51 AM6/4/21
to open-phd...@googlegroups.com
Stefano:

I asked the same question about sub pixel corrections about a year ago.
One of the experts sent me a technical paper that explained PHD looks at a group of pixels surrounding a star, not just a single pixel.
As the light intensity changes across the group, PHD responds.  So the star does not have to "move" entirely to a new pixel before PHD knows that a correction is required. 

If I can find the paper I will post it.

George
 





S O

unread,
Jun 4, 2021, 11:25:51 AM6/4/21
to open-phd...@googlegroups.com
Thanks much George.
At any rate, for now I'm still not clear on the matter, as I think back to a course I've taken on photographing comets, where the main trick is to keep the exposure time short enough that essentially the comet (or each of its tiles in the image) doesn't move more than the resolution of a pixel of the camera and scope setup. This way you can do scientific estimations on the comet itself, as you are getting a meaningful sample in (each) pixel.

So if I assume all the photons fall in a single bucket (as at least theoretically in my case even with 2-3 arcsec seeing smear), how I could possibly assess subpixel adjustments is still puzzling to me -- because while I heed what Brian says above, ultimately the guider is still based on the  pixel-based reading of the guiding star, isn't it? The translation to arcsec logically seems to happen after the pixel reading -- but again, if all falls within one pixel, it would always translates to 0 arcsec of correction, as the guiding star will appear to never move.

Thanks
-s

bw_msgboard

unread,
Jun 4, 2021, 11:40:36 AM6/4/21
to open-phd...@googlegroups.com

Stars aren’t imaged as point sources because of diffraction.  A star appears as an Airy disk with a surrounding pattern of diffraction rings.  Even with a very coarse image scale like you have, the Airy disk typically spans more than one pixel unless your guide camera has very large pixels.  If you use the Star Profile tool, you can see how big your guide stars are.  Calculation of the star position uses a centroid algorithm that reacts to brightness fluctuations in the region covered by the Airy disk.  Centroiding is why the star position accuracy can be fractions of a pixel – it’s a floating point numerical calculation.  Our experience is that the centroid accuracy is typically around 0.05 pixels for multi-star guiding, and 0.1 – 0.15px for single-star.  The reference material George sent you explains the details of the various algorithms – it’s a math problem, not something that will be intuitively obvious unless you are familiar with the physics of light and optics.

 

Bruce

 


S O

unread,
Jun 4, 2021, 11:48:41 AM6/4/21
to open-phd...@googlegroups.com
Cool Bruce, it does have to span more than one pixel though. Which is what I was wondering about...
The rest was intuitively within reach.
Interesting to know the accuracy you report.
I hope George finds the paper, it'll be yet one more must-read in my (long) list ;)

Thanks

George

unread,
Jun 4, 2021, 2:48:24 PM6/4/21
to Open PHD Guiding

I did post the links to a new post

Stefano O

unread,
Jun 4, 2021, 5:06:10 PM6/4/21
to Open PHD Guiding
I see it, thanks !
Reply all
Reply to author
Forward
0 new messages