PEC help

[codemonkey]

I've been trying to record the PE of my EQ8, generate a PEC curve, and apply it. I'm using PHD, EQMOD and PECPrep. Basic process has been:


Record guide log in PHD (guide output disabled)
Stop guiding / hit timestamp button in EQMOD as close to simultaneously as I could
Park mount to defined position
Load guide log into PECPrep with appropriate details, including the application of the EQMOD timestamp
Change worm cycle length to 199 in PECPrep
Manipulate PEC by disabling the first, second, and last worm cycles, all of which seemed to deviate a reasonable amount from the other 6 cycles captured, especially the first
Save PEC
Unpark mount
Open/play PEC in EQMOD - Gain is defaulting to 0x - is this right?


As soon as I apply PEC, the guide star seems to move a fair amount, as might be expected, but then it continues to move, as though the PEC is worsening the PE rather than improving it. Do I need to let it run for a full worm cycle before it'll sync up or something?

Was disabling the last cycle in PECPrep a bad idea, maybe causing it to be out of sync with the timestamp? Maybe I should drop its weighting to 0 rather than disabling it.

I've noticed that playback of the PEC in EQMOD always seems to start from the beginning of the curve (at least that's what the line slowly moving across the PEC curve display seems to suggest) -- if this is the case how can it ever be in sync with the worm?

[glend]

Maybe this is a dumb question, but if you can guide why bother? What are you gaining?

[codemonkey]

Not a dumb question :-)

I'm often imaging at 0.74"/px these days, which means I have a very low tolerance before things affect the final image. For guiding to correct an issue it has to see it first, and when guiding corrects an issue it has the potential to over-correct, causing even more issues.

Couple the above with the fact that I'm often imaging in less than perfect seeing, where the stars naturally bounce around a bit, it's easy for guiding to make things worse than it has to be. PE + bad seeing = guiding can't possibly be doing as good as it could be. I can't do much about the seeing, but I can do something about the PE.

In an ideal world I'd prefer to work unguided, and I'll do that when I can. With PEC and good PA, I'm sure I can do that with my smaller scope. I've heard of people in the UK using my same mount and running at high resolution for subs of up to 30mins unguided before trailing became a problem, so if I can do that with my big scope, I might do that as well. At the very least I want to minimise issues before they hit my sensor, not afterwards.

[Eden]

What method are you using to do your polar alignment these days, Lee?

[codemonkey]

Drift.. I'm well under an arc minute on both axis. In addition to that, I have a permanent set up, so unless the pier moves (possible, haven't had it long enough to be sure) it should be consistently good now.

[Shiraz]

pretty much matches my experience :P. Gain should be 1 however.

I also noticed that the PE curve display always dropped back to the start - it doesn't do that in the simulator, so maybe something is haywire when reading out the ring counters on the mount. I also found that the little bar jumped back to the start before it could cover the whole of the worm cycle when I moved the phase slider.

I wonder if anybody out there has actually got PE working on the EQ8

[codemonkey]

Ah, I thought the gain of 0 sounded a bit suspicious. Thanks for confirming that Ray :-)

Sounds like this might be a bit of a dead end when done directly via EQMOD then eh? Maybe I need to look more at PEMPro, see if I can get that to play with my QHY5L-II and EQMOD. Shame.

[Amaranthus]

Did you try VS-AutoPEC?

[Shiraz]

Hope you don't mind me butting in Lee.
Barry, I tried autopec - didn't work and had to reload EQMOD and phd2 to get everything working again. For a starter, autopec gathered data for 3x EQ6 periods rather than 3x EQ8 periods (I think that I had defined the EQ8 at that stage). it then generated a curve that looked different to that from pecprep and tracking failed altogether - even with pec off. Looks to me like there may be a bit of a bug in the translation of mount parameters or something. Pity really, since it is a very good process in principle. I guess that this is the sort of problem that might be difficult for the code developers to devote much time to when they very generously do so on a voluntary basis.

[Amaranthus]

I couldn't get AutoPEC working on my AZ-EQ6 either Ray, it kept fighting rather than helping the PHD2 guiding. I am still working through what the problem might be. But it has definitely worked for some people. Note that Chris recently updated the software - it can now train on up to 9 worm cycles, which should in theory underpin a better curve. More tinkering ahead...

[Eden]

Barry and I were discussing this a few weeks ago when I asked for his input on why the Synscan PEC routine was behaving strangely. He suggested, among other things, using the AutoPEC in EQMod.

After recording 5 worm cycles (which takes a fair while), it kicked in and worked beautifully. I watched my RMS drop down to somewhere between 0.16 and 0.20 (at 900mm, in near-IR), compared to the 0.30 - 0.40 beforehand. This was with a PA of within 1-2'.

If the seeing up your way is really that bad, trying guiding in the near-IR. It can only improve the situation.

[codemonkey]

Not at all mate, appreciate it in fact.

In response to the original question, I haven't tried it with my EQ8, I did briefly try it with my old NEQ6 and it appeared to immediately worsen results as well. Also given that the first couple of cycles in particular consistently seem to be misleading when evaluating with PECPrep, I'm disinclined to use a fully automated system, especially one that only allows you to capture 5 worm cycles... wait, apparently that's 9 now? I should probably update :-)



Ah man, I wish I could get close to that. Last night I was getting mostly 0.75 - 0.85, with spikes out to 2". Mind you, it did look particularly poor last night. All the same, that is much better than with the NEQ6, I could never get that below 1.2".

Starting to think maybe I should give up on imaging at this resolution and down size. My primary interest is galaxies, but since I've had this scope I've managed only once to get images with stars having an FWHM ~3, usually it's upwards of 4 and very soft due to seeing/guiding. I suspect something around 1200mm might be more usable for me, and might still give me a decent amount of targets.

Near-IR guiding is a great idea, not something I've tried yet. Having said that though, if it really is just seeing, then even if I get better guiding, I'm still not getting the resolution to justify imaging at this resolution so maybe there's no point.