PHD2 Predictive PEC Algorithm

[Startrek]

Over the past 2 weeks on various nights with varying conditions I’ve been comparing running PHD2 with the conventional Hysterisis algorithm and the Predictive PEC algorithm with my mount not PEC trained by any other software I use like EQMOD.
I’m running PHD2 Multistar guiding too
At this stage I’m slightly favouring the PPEC algorithm for 2 reasons -
1/ Total guiding error appears to be slightly better by between 0.05 to 0.08 arc sec across 2 to 4 hours of imaging
2/ Guide corrections in Ra and Dec appear to be smoother ( especially Ra ) and more responsive with less erratic pulses which shows on the guide graph ( haven’t check guide logs yet )
3/ There also appears to be less excursions in Ra like there was using the Hysterisis algorithm ( possibly a slight reduction in periodic error )

Has anyone done any comparisons in PHD2 with both algorithms as I’d be interested to hear their comments and results

Thanks

[JimsShed]

Thanks for the info Martin. I wasn't aware of the Predictive PEC option, so I will have to give it a go.
Regarding the Multi-star guiding, I do have that switched on these days but not noticing a material difference for my location and gear. For me, both work well.

[Startrek]

Jim,
Thanks for your comments
Just remember the Predictive PEC algorithm requires at least 2 worm cycles of your mount to evaluate your mounts PE , the guiding operates like hysterisis in the interim until PPEC kicks in
My EQ6-R pro mount requires 8mins for a worm cycle so I usually wait to review the guiding performance after 20 mins or so
Cheers
Martin

[Pepper]

Martin, I have indeed found both my mounts guide better using ppec in phd2.
And I’ll also agree with Jim that I haven’t really noticed any difference with multi star guiding.

[Startrek]

Thanks Pepper
Glad PPEC works for you
I didn’t go down the path of PEC training using EQMOD and AutoPEC as it affects 2 things , auto restore calibration and dithering ( I use both ) The PHD2 PPEC works well and allows me to use auto restore calibration and dithering
Multistar guiding wasn’t designed to improve guiding but add another layer of redundancy to maintaining a good guide Star to guide on
I too haven’t noticed any improvement in guiding using Multistar guiding, I just use it now as default and a bit of reassurance
Cheers

[peter_4059]

Martin, I think the intent of PHD2 multi star guiding is to average the position of star centroid over many stars to minimise local seeing errors.

[Startrek]

Peter,
Thanks
Correct , in simple terms maintaining a good guide star to guide on
Martin

[peter_4059]

No....averaging the centroid of multiple stars to cancel the seeing

[multiweb]

Exactamento :thumbsup:

[Pepper]

Ppec is working very well for me now.
Give my ra worm a tune up and after a couple of cycles to get itself together, my guiding has been best it’s ever been. You can see it working when the corrections are shown.
Flat as a tack ever since

[Startrek]

Steve,
Your guiding numbers look excellent
May I suggest you change the scale on your graph to x 200 and y 8”
as the current graph lines look really flat due to the scaling
The graph isn’t the be all and end all of monitoring your guiding but with an upscale of your graph at least you will see any minor or medium excursions in Ra or Dec.Obviously the log files are the best way to analyse your mounts performance but with an upscale of your graph it will provide a more detailed view of your guiding
Cheers
Martin

[Pepper]

Yeah I know about the scale. I’ve been using it like that since the days of my eq5 so it’s what I’m used to. Can’t get my brain around change.
Pretty happy with how it guides now. As long as I don’t see big spikes like I used to I’m pretty chill about it.

[Addos]

used predictive pec quite a few times. as others have noted there was a slight improvement, but not significant over the default.

ascom based pec training made a far bigger difference, in my opinion.