PME Commissioned ... finally!!!

[Paul Haese]

I have my mount pretty close to the pole but I would like it better. Currently I am using Tpoint but doing the alignments with a reticle on the spare scope on the mount. This generally is ok but proves to be not consistent for PA purposes. I move the mount one way then the next Tpoint run it tells me to go the other. To be honest I would like to see a nice tutorial on how to setup Tpoint, automapper and pinpoint with sky6. I am sure that I am making a mess of this all the time and it would certainly help with my guiding.

Besides being able to do a 20 point run and then adjust after the equipment has done the work has got to be better. So who is up for a lengthy tut? BTW I have the catalogue too.

[marc4darkskies (Marcus)]

Originally Posted by Paul Haese

I have my mount pretty close to the pole but I would like it better. Currently I am using Tpoint but doing the alignments with a reticle on the spare scope on the mount. This generally is ok but proves to be not consistent for PA purposes. I move the mount one way then the next Tpoint run it tells me to go the other. To be honest I would like to see a nice tutorial on how to setup Tpoint, automapper and pinpoint with sky6. I am sure that I am making a mess of this all the time and it would certainly help with my guiding.

Besides being able to do a 20 point run and then adjust after the equipment has done the work has got to be better. So who is up for a lengthy tut? BTW I have the catalogue too.

Are you using CCDSoft Paul? If so no lengthy tut is necessary and you don't need Pinpoint! AutomapperII is a breeze to use and their documentation got me going. http://www.newastro.com/downloads/automap2/automap2.pdf

It's worth spending an evening getting familiar with how it hangs together. Once it's set up doing another model requires no further setup.

Before adjusting polar alignment I now do a 50 point all sky model (30 to 85 degrees alt). It only takes 30 minutes! You might get away with 30 points but I wouldn't go less than that.

Before you adjust the alignment, you MUST (IMO) add additional terms to the resulting TPoint model to optimise its accuracy (and hence the PA recommendation). This isn't rocket science either. Go into the TPoint model you've created, open the fit data for the model and click the More Terms ... button. Then do Suggest Terms and Start. If the term at the top of the list improves your parameters (RMS and PSD) by 10% or more, click Add to Model. Repeat this procedure until your parameters stop improving. This is fairly quick & dirty (as opposed to analysing different combinations of model parameters) but it works.

The beauty of the PME is that TPoint tells you how much to twist the knobs. But I always prefer to confirm the alignment by doing another 50 point model.

Because of problems with my Laptop I only did a 100 star final model this time. Some argue that it's a case of diminishing returns to do more than that but I will when I get a new laptop and TheSkyX working.

[Paul Haese]

Marcus I am using MaximDL. I suppose I could use CCDsoft, perhaps I will take a look at it.

I am already using the extra terms and got the pointing down to 93 arc seconds at one point but now it is up near 125 after my last try at PA. I really need to sit down a bit more and read this stuff inside and out I suppose and then do it in the field.

[gregbradley]

I had the same problem Paul. Doing the t-point model manually isn't really feasible as it takes too long. And then as you say the recommendations can sometimes make the autoguiding errors worse.

I did a polar alignment using Pempros Polar Alignment wizard (the best software I have used so far for that). It got the PA very very close all by itself.

Then Precision PEC is possible (it requires a close PA to work).

Those 2 actions are getting me quite small guiding errors with 3 second guide exposures, callibrating the autoguider each time before I do an image run. But I do sometimes get slightly eggy stars so I am off a bit. it can vary by simply using a different guide star.
My typical guide errors are at the low end .05 and at the high end about 1.0, average is about .25. So pretty good really.

Automapper II does not come with a help file. It is on a separate link at the New Astro website. This tells you how to setup both the Sky6 server settings and same for CCDsoft. I am assuming that is why the plate solving failed when I used it last (wasn't setup properly).

I hope to use it next spare night (its hard to spend a clear night on fiddling with your gear when it is already working fairly well).

I am not sure if Automapper works with Maxim, probably not. Perhaps AAG Tpoint automapper works with Maxim as it requires Pinpoint which is a Maxim product.

Greg.

[marc4darkskies (Marcus)]

Originally Posted by gregbradley

I had the same problem Paul. Doing the t-point model manually isn't really feasible as it takes too long. And then as you say the recommendations can sometimes make the autoguiding errors worse.
...
Greg.

Beg to differ there Greg. Getting TPoint to recommend harmonic terms is perfectly feasible provided you have enough data points (50 is good). Took me a minute!!! No, I didn't spend ages optimising it, but that's not necessary. Just stop adding them when you don't get a 10% improvement. By the way, when I did this in TheSky6 I got almost identical accuracy to TheSkyX's Supermodel function with very similar polar alignment results. You need to optimise your model (by adding harmonic terms) to achieve a low sigma for your PA terms (a very low 1.5 in my case). If you have a high sigma for these terms the recommendation will not be reliable.

[Paul Haese]

Right then. I was trying to use Maxim, AAG and pinpoint (I made an error before when I said automapper II).

The pin point program has run out and the AAG was free and well maxim came to me from some other place (er I mean I ahh bought it).

So maybe I will go and get CCDsoft (must remember to load the drivers for the QSI if I do so it will work) and automapper II and see how this works. Do I still need to keep the GSC (star catalogue) or does automapper II use the sky6 catalogue?

[marc4darkskies (Marcus)]

Originally Posted by Paul Haese

Right then. I was trying to use Maxim, AAG and pinpoint (I made an error before when I said automapper II).

The pin point program has run out and the AAG was free and well maxim came to me from some other place (er I mean I ahh bought it).

So maybe I will go and get CCDsoft (must remember to load the drivers for the QSI if I do so it will work) and automapper II and see how this works. Do I still need to keep the GSC (star catalogue) or does automapper II use the sky6 catalogue?

Aaah, yes. You'll need to augment TheSky6's star database to improve plate solving. Check out which ones it supports yourself but I downloaded USNO-A2.0 which TheSky6 supports natively. It was free to download but I don't think it's available any more - it's been superceded. If you want it, PM me your address and I'll send you a DVD.

[Doomsayer]

If you are using the Sky6 you are limited to either using Pinpoint or CCDsoft for platesolving with either Automapper2 or AAG Tpoint Mapper as I understand. If you are using TheSky6 it really is worth purchasing the full pinpoint - DC3 can supply a disc with the GSC loaded - I find the GSC is sufficient for all plate solving with my 12"RC. Pinpoint or equivalent plate solving capacity also makes a lot of powerful automation possble with software such as CCDautopilot, such as precision slews (based on plate solves), dithering, acquiring stars for focusing or G2V calculations etc.

I don't have theSkyX, so I will not comment on Marcus's Tpoint mapping insights there as it seems to be far more self contained.

If you are not using Tpoint fully with the PME you are really under utilising capability and possibly being counter productive to its performance.

Before I moved to automated plate solving for mapping, I used to just map 25 or 30 points manually, by centering the star on a digital cross hair in CCDsoft using the main imaging camera bin2x2 and main OTA - it takes about half an hour or less to do this, using the n-s-w-e motion controls in TheSky. After 2-3 mapping runs I have the mount well aligned and then get typically 15arcsec pointing accuracy. The latest PME manual now covers all of this procedure well in my opinion. There are several Tpoint tutorials buried on the SoftwareBisque site, applicable to TheSky6 suite. It sounds like TheSkyX has made this stuff a bit easier.

[Paul Haese]

Thanks for the tips and help guys. I will go off and work this out. I am thinking that I will go with CCDsoft and automapper II. I will PM you Marcus for the catalogue.

[frolinmod]

Originally Posted by strongmanmike

I dunno...the PME just needs a good Polar Scope...

You're making me laugh Mike.

Honestly, the focal length you are working at with that gorgeous fast refractor is so low that polar alignment is probably good enough just by pointing the scope somewhere near North. Drool! I want one! Try working at FL=3911mm. Or better yet, don't. It requires very good polar alignment. In any case, Tpoint gets me there.

[rogerg (Roger)]

For the sake of completeness and to help others who search for this in the future, I'd like to finish off my discussion earlier on my problems of night-to-night sync location/pointing.

I've found that my problem was my scripting. It turns out that during the night my scripts were causing the Paramount to flip sides of the meridian (unexpectedly) and when doing so perform a plate solve and synchronization (unexpectedly). Once this sync was done, the TPoint model was polluted and pointing off for consecutive nights (at best) or the plate solve was incorrect and pointing was off even more than just pollution of the model (at worst).

So, now that I have ensured there's no errant plate solves and synchronizing due to my scripst, I'm finding my pointing between nights is repeatedly within about 10 pixels (about 8 arc seconds) of the previous night. Very happy to finally sort this out so I can do more productive work with the mount.

My scripting is done using CCDCommander but it's not it's fault, it's simply my choice of settings in it, having come from a LX200 which didn't concern its self with meridian flips.

[gregbradley]

How do you program in meridian flips?

Greg.

[rogerg (Roger)]

Originally Posted by gregbradley

How do you program in meridian flips?

Greg.

I'm not sure on all the details as I typically avoid going past the meridian, my eastern hemisphere is where I try to tay.

When taking images CCD Commander has an option to plate solve after meridian flip. During taking the image if meridian flip is occurs the plate solve will be done (if option enabled) and mount re-slewed to desired co-ordinates. I think.

Move-to commands will obviously flip the mount according to slew limits and setup options, so if object within x of meridian it will flip (and then optionally perform late solve and reslew also).

Roger.

[gregbradley]

Thanks Roger.

Greg.