Metaguide vs PHD2 Generic equipment discussions · Himanshu Pandey · ... · 29 · 1419 · 0

I use the Ekos guider and it works better for me. Calibration seems to be more flexible and faster, the interface is more straightforward, and I think I get better results.

I used to switch between PHD2 and the Ekos internal guider, and with the same settings for calibration and guiding parameters, same results in guiding.

Or I should say I got similarly bad results with my strainwafe mount (WD17). Just dreadful amounts of periodic error. Had many occasions where PHD2 PPEC was not able to lock into the major period of the RA oscillation, so ended up with a good starting value from PHD2 frequency analysis, used this as a starting point in Ekos GPG, and used the resulting period in PHD2 without adaption. 

Then after the last KStars update, which changed the way Ekos calculates the RMS error so it’s in line with the way PHD2 calculates it, same results still. Until I accidentally deleted my equipment profile, and had to make a new one from scratch, haven’t seen guiding deviations above 0.4” since. Cannot find what changed. 

More on topic, is the use of single guide star, single guide star with lucky imaging or multi star. With a low powered mini PC, and a SW mount that needs frequent correction, multi star guiding is notably slower. Depending on seeing conditions I got the best results with either single star with frequent corrections, or multi star with only 5 stars and a 50% lower correction frequency. I assume single star with lucky imaging could give the same results as multi star, as long as the algorithm is fast enough. But there’s no way to use lucky imaging in guiding on Linux yet.  Would be interesting to try out. 

Last five night, and probably the next five nights as well, seeing has been good, even excellent for my location. So I’ll stick with single star guiding since it gives me good results right now.

Jochen Schambach:

Tobiasz:

Himanshu Pandey:
Thanks, I currently use phd2 and understand why it is so popular.   It has the benefit and care of many people over many years and has been purposely made to be easy for beginners. 

I don't believe metaguide is intended to control at a high frequency but rather deal with the seeing induced problems in a different way.  At some point they implemented multi star guiding in PHD2 to deal with atmospheric seeing to do the same thing.

I think metaguide only looks at one star but has the benefit of a much higher sampling rate and doesn't have to wait for the next multi-second exposure to make a change and can step in sooner if needed.

*
Yes, MetaGuide uses one star and averages out the centroid via videobased "lucky guiding". Me personally I guide with a 5iii585M at around 5-10fps with guide pulses between 0.75 and 1.5 seconds. This means if I guide every second metaguide calculates the centroid from a stacked average of 9 - 10 frames which works really well (for me). "Nervous" mounts like strain waves without encoders or cheaper mounts do benefit from this approach. 

Talking about guiding tools is one thing, I would recommend testing them on the same target and several nights and see for yourself. I did it too and definetly do benefit from aggressive guiding (much better FWHMs and better eccentricity). I thought I am just living in "bad seeing" but I was just not guiding well enough with PHD2. 

Some say it is complicated but Frank @Freestar8n updated the documentation and has his own forum around MetaGuide if you need help. 

Happy guiding!

Regards
Tobi

Hi,
Is this the same thing Like Zfilter in phd2?
zfilter works Fine in some conditions.
sometimes I have Great Results in dec with Pool polar aligment to fight backlash

*
Hi, sorry for the late response. 

To be honest I do not know what the zfilter is or how it works, so I need an explanation here. 

The algorithm part in MetaGuide is only the calculation of the centroid. It uses a stacked & aligned image and searches the "hot spot"(=brightest part) of your guide star. This way most of the seeing/abberation induced deformations in the outer parts of the stars can be ignored (because they are not bright enough) and you are guiding on the "real" star core. 

You have direct control how the guide star is "presented" to the algorithm via the gain, exposure time, "stack time" and focus. The last session my seeing was between 1.8-2.5", but the "AFWHM"(= [Stacked &] Aligned full width half maximum) of my guide star was between 1.2-1.6". This makes your guidestar pretty pinpoint and it enables tight guiding. 

Guide settings are up to you. With the video view you can "live" test your guide settings and can change them instantly without affecting the calibration and without a "learning phase" of the tool. Guide period, guide rate, aggression are settings known from other tools. 

In general, I would say the algorithm is only one part of the equation and knowing your mount is the other. 

sometimes I have Great Results in dec with Pool polar aligment to fight backlash

In Metaguide you can manually slew your mount by small amounts and find out your backlash value and type it in the backlash compensation field in the settings. Everytime your mount changes directions the exact amount of backlash comp. in milliseconds will be applied to the guide pulse. Because of the live view slewing around to find out your backlash is a pretty quick process. 

Frank did a presentation of Metaguide on the Astro Imaging Channel a few years ago. There he also explains the guiding part and how it works: Youtube

Hope this helped a bit!

CS
Tobi