Help - Fixing that last tiny bit of polar mis-alignment when using an SCT

[g__day]

Folk I'm interested in fixing that last bit of polar misalignment I am seeing on a permanently mounted rig (Atlux on a pier in an astrolab). At my disposal are the latest versions of MaxPont, PEMPro and Tpoint for the Sky 6 - and plenty of time.

I guess my polar alignment is within 1-3 arc minutes of the SCP (true versus refracted SCP - I don't know).

My main scope is a C9.25 with a Canon 400D on it. I use the central focus spot on the Canon viewfinder to centre my stars when aligning (shining a very small red led into it so I can see this spot when its otherwise pitch black! Finally I assume my C9.25 has some very slight mirror shift that changes based on elevation or other random factors. Being Carbon fibre my focus rarely changes night by night. I have a starlight fine focuser on the SCT (greatly helps remove any mirror shift) - but I focus using a Meade motorfocuser into a JMI PC controller and a Bhatinov mask. So Once I'm set up and focused - I never physically touch the mount or scope during the night's run (so rule out human errors from knocking gear). The OTA is firmly attached using Losmandy bars - so I assume there is minimal differential flex happening via the mounting rails.

Last time I checked with multiple 30 minute runs on PEMPro on both Alt and Az - it said I was aligned with 0.2 arc minutes of the SCP.

MaxPoint with a 80 star model says I'm 25 arc seconds too low and around 3 arc minutes West of the SCP.

Tpoint with a 70 point star model says I need to rotate the mount 4.7 minute to the East (Sigma 93.987)and elevate it -2.1 minutes (Sigma 95.899) (which I guess means drop it 2.1 arc minutes - which would be disagreeing with MaxPoint)?

So lovely - three modern pieces of software - all purpose built to aid polar alignment - all disagree!

Oh the Hand controller itself - a SkySensor2000-PC Final version EPROM - set to polar aligned mode. Interesting behaviour when I do an initial 3 star alignment from mount level and facing East. I generally align on a triangle of stars Achnerar, Dipha and finally Rigel Kentarus - always in that order - always twice - one pass through alignment followed by the next. Once I have a two star alignment and slew to Rigel Kentarus the first time (before 3 star alignment is reached) the mount always stops pointing about 1-2 arc minutes too far West and 1-2 below my target. I am pondering if this closing error on my 3rd star is a very good indication of my polar mis-alignment error. If my mount overshoots the third star to the West by say 2 arc minutes - sounds to me my starting position was too far West to begin so I have to rotate my mount 2 arc minutes to the East - agreeing with Tpoint. Similarly if my mounts end up too low on the thrid star does that mean it started too low by this inverse amount?

If I was using a large refractor - I guess I could rule out any mirror shift during drift alignment - by eye or using PEMPro. As this is not the case I am pondering is this why I see three modern pieces of software giving conflicting advice.

I am tempted to rotate the base of the mount 2 arc minutes to the East then doing a 3 star alignment on the hand controller and noting if homing to the 3rd alignment star becomes a lot more precise. Then I could repeat the exercise in Altitude.

Can I ask for advice - why what you do in my place to nail the final polar alignment tuning? Do I trust Tpoint over the rest of the software and or the SkySensor2000-PC hand controller itself?

Thanks,

Matthew

[gary]

Hi Matthew,

To determine whether to raise or lower the RA axis, what was the actual value of the
term ME including its sign, as distinct from the value in the verbose report TPOINT
provided that you transcribed above?

[Barrykgerdes]

Being within 3 or 4 arc mins of polar alignment is probably the best you will practically do with the simpler GEM's. There are many factors that will influence getting any closer than that. The most significant in your case is probably flexure in the system depending on the attitude of the "load". A "perfect" polar alignment done by drift alignment on a star moving a few degrees either side of the meridian will not necessarily be the same as one done totally one side or the other even if you have perfectly balanced the system.

A mount like the G11 has the polar axis supported between two bearings that are only about 9" apart. Even with the bearings as tight as possible there will be some give with a heavy load.

Other factors like orthogonality may also be having a bearing when you get down the the 2 or 3 arc minute errors.

The error readings given by various software programs are usually based on positions of stars after alignment and depending on where the stars are will affect flexing and refraction so you could easily be "chasing your tail"

Barry

[g__day]

Gary,

The details are:

Tpoint – 70 Stars 4 Oct 2011-10-04

Southern Hemisphere

************* AZIMUTH ************
MA: +232 seconds (+3.9 minutes).
Rotate axis East (counterclockwise).
For latitude -33.88°, the azimuth adjustment is 4.7 minutes.
Sigma=93.987

************* ALTITUDE ************
ME: -128 seconds (-2.1 minutes).
The polar axis should be raised -128 seconds (-2.1 minutes).
Sigma=95.899
coeff change value sigma

1 IH -86.192 -86.19 92.467
2 ID +286.867 +286.87 151.600
3 NP +22.545 +22.54 131.151
4 CH +87.512 +87.51 93.321
5 ME -128.002 -128.00 95.899
6 MA +233.006 +233.01 93.987
7 TF -42.010 -42.01 18.051
8 FO +59.133 +59.13 201.502

Sky RMS = 52.72
Popn SD = 56.02

* Correlations between terms:

ID +0.0429
NP -0.4144 +0.8097
CH -0.9910 -0.0658 +0.4032
ME -0.2123 +0.6394 +0.8707 +0.1991
MA +0.0483 +0.9797 +0.8269 -0.0690 +0.6584
TF +0.1176 -0.2965 -0.2115 -0.0348 -0.0109 -0.2705
FO +0.0734 -0.9306 -0.9106 -0.0538 -0.8695 -0.9136 +0.1766

IH ID NP CH ME MA TF

* fit

coeff change value sigma

1 IH -0.572 -86.76 92.540
2 ID -1.039 +285.83 151.450
3 NP -0.001 +22.54 130.689
4 CH +0.668 +88.18 93.320
5 ME +0.431 -127.57 95.333
6 MA -0.604 +232.40 93.818
7 TF +0.168 -41.84 18.084
8 FO +0.415 +59.55 200.598

Sky RMS = 52.72
Popn SD = 56.02

* Correlations between terms:

ID +0.0368
NP -0.4212 +0.8087
CH -0.9910 -0.0609 +0.4090
ME -0.2185 +0.6360 +0.8692 +0.2043
MA +0.0424 +0.9796 +0.8259 -0.0643 +0.6550
TF +0.1189 -0.3053 -0.2198 -0.0357 -0.0179 -0.2793
FO +0.0799 -0.9303 -0.9100 -0.0591 -0.8676 -0.9131 +0.1859

IH ID NP CH ME MA TF

IH : = -86.76
ID : = +285.83
NP : = +22.54
CH : = +88.18
ME : = -127.57
MA : = +232.40
TF : = -41.84
FO : = +59.55
* GSCAT

*****************************

So not nearly as good pointing under Tpoint RMS 52.7 arc second - versus +/- 16 arc seconds for MaxPoint the other day...

MaxPoint tells me (on 80 stars - well distributed) Telescope Errors are:

Hour Angle Bias: -00 04’ 11” Polar Azimuth 00 03’ 04”
Declination Bias: 00 02’ 17” Polar Altitude 00 00’ 25”
Collimation: -00 02’ 32” Tube Droop -00 01’ 38”
Axes Misalignment 00 04’ 13” Mount Flexure -00 09’ 01”

With the RMS being a circle 31.6 arc seconds in diameter

[gary]

Hi Matthew,

A negative value of ME for the Southern Hemisphere means the RA axis must be
raised, in this case by 127 arc seconds for the true pole position. You can ignore the
minus sign in the verbose report. The refracted pole in Sydney is about another
86 arc seconds above the true pole. Typically you will want to be aiming somewhere
between the true and refracted pole. The specific point depends where you are
imaging in the sky. (Come to Lostock at the end of this month, attend my presentation
and I will be discussing this topic more :hi:).

Notice, however, that the sigma value for ME is 95 arc seconds. Sigma is
its standard deviation. A value of -127 +/-95 is saying that the value of -127
is in the noise.

Your MA term looks "real" enough. I would adjust it in Az and then repeat a star pointing run.



A pointing performance of 56 arc minutes RMS is a pretty good result for a mount of
this type.

Firstly, drop the FO (Fork Flexure) term. Based on the above, it is not
apparent you have any. Notice the value is +59 but the standard deviation is
200? That is telling you that it is more noise than signal and that it will
not be making a meaningful contribution to the model and quite possibly "stealing"
a little from your other terms, in particular a little from MA and NP.
FO models fork flexure on fork mounts. On your GEM, DAF which is the
Declination Axis Flexure, is the one you should investigate.

Keep the TF (Tube Flexure) term for now.

However, the signal to noise ratios for NP and CH aren't good. The two terms have
very similar signatures and when you sample stars, you need to get a good spread
of stars at differing altitudes in order to help distinguish them. Having said that,
the report shows the two aren't highly correlated but I recommend you see what
happens if you keep one and drop the other. Experiment by first seeing what happens
when you drop the NP term. Untick the NP term or you can type at the command line
reset
lose np
fit
See if the signal to noise ratio of CH then improves. If not and in any case,
also experiment by dropping CH but including NP. Untick the CH term and tick the NP term
or you can type at the command line -
reset
lose ch
use np
fit

It might well be you only need one or the other. CH is the more likely of the two
and especially so for a GEM.

After experimenting with CH and NP, try the suggest term dialog and add one or more additional recommended
terms if they have good signal to noise ratios and the RMS/PSD values drop as well.



Given the standard deviations in some of the above terms, you can expect some
to and fro in the RMS values.

Based on the above, your FO is not worth keeping and even more so because
you have a GEM. Check what happens when DAF is added. The RMS
is likely to drop some more when you add some additional terms. DCEC, DCES,
HCEC and HCES are always worth checking as these often account for various
runout errors. Looking at the above, I would try DCEC first. Depending upon gearing
ratios, higher order harmonic terms may also assist.

[gregbradley]

Great thread, some valuable advice there.

Another view on this. As a contractor I always prefer to measure to the acutal rather than the theoretical. Working on a few building sites will teach you that lesson very quickly! What something should be versus what they are can be 2 significantly different measurements.

Keep in mind these models are exactly that - theoretical not actual. They are a computer generated of model of the mounts performance, not the actual mounts performance.

So in the case of yours where there is disagreement I would check the actual performance of the mount. If you autoguide as-is and check the guide errors in a few spots. Then take one of the model's suggested improvements and the autoguide again. Now its a simple case of are the guide errors on average (not just one guide star) better or worse?
Now you are checking against the actual and not in theory land.

As the lowest possible autoguiding errors I assume is your target, then this should be the measure of what you judge against.

With T-point you also have to make sure you get rid of outliers in the model before you do the polar alignment assessment otherwise you are adding in noise to the model and it will skew the result.

I found T-point able to get me very close and once I was getting round stars routinely I stopped at that point even though a further tiny tweak no doubt was possible.

Greg.

[g__day]

What amazingly detailed responses - I am blown away!

Guiding has always been excellent on this mount - but I can certainly understand the maxim of the closer to the pole the better your guiding should be. With the SkySensor2000-PC its a bit more complicated than that, given a three star aignment the SS2K will point and track very well even if your 20 degrees off the celestial pole. It compensates for refraction and huge polar mis-alignment in how it runs the motors. So much so that when I run PEMPro I always only do a one star alignment; I didn't think to take this path with Tpoint and MaxPoint modelling the sky - I did full, fresh 3 star alignments before I put them to work.

Given MaxPoint and Tpoint (and if I'm interpreting the 3 star alignment of the SS2K itself correctly) all seem to say move the mount about 2 arc minutes to the East - I think will try that first and see what happens to pointing and tracking.

Next I'm going to try and spend a fair bit of time going through Gary's advice and see what of it I can apply (and understand)!

Wow again folks!

[frolinmod]

If you guys would upgrade to TheSkyX and the Tpoint Add On, the Supermodel feature would automatically get rid of the outliers for you. It might also come up with a better model than you can do manually.

[g__day]

I always try to think my way out of challenges before I buy my way out! I seriously haven't scratched the surface of what the Sky 6 PE and Tpoint can do.

Over time I am likely to jump to current versions of SB products (even though I hate annual licence software), but I'd rather learn the basics well enough - and gain skills - rather than just have software gift me the answer. Maybe I'm old fashioned thinking that way!

[gregbradley]

I will have a copy of that with my PMX due soon. Its good to hear some positive feedback about the Sky X as Software Bisque are not the best marketers in the world. So info on what the Sky X actually does and how its different from the Sky 6 is vague at best.

They need a slick little video.

Greg.

[g__day]

I've duplicated this question on the PEMPro forum of CCDWare, then noted a query between PEMPro vs Tpoint dating back to April 2011. It's of real interest because the author of TPoint - Patrick Wallace - comes into the thread on page 2 and he and Ray Gralak do a real masters course on the issue!

http://ccdware.infopop.cc/eve/forums/a/tpc/f/401101098/m/1377011106

[g__day]

Gary,

Adding in just the DAF term and removing the Fork Flexure term brings the pointing error down by half and modifies the correction advice!

Southern Hemisphere

************* AZIMUTH ************
MA: +155 seconds (+2.6 minutes).
Rotate axis East (counterclockwise).
For latitude -33.79°, the azimuth adjustment is 3.1 minutes.
Sigma=18.554

************* ALTITUDE ************
ME: +31 seconds (+0.5 minutes).
The polar axis should be lowered +31 seconds (+0.5 minutes).
Sigma=23.135

* fit

coeff change value sigma

1 IH -175.012 -175.01 42.430
2 ID +172.928 +172.93 27.058
3 NP +203.402 +203.40 26.316
4 CH -206.566 -206.57 46.235
5 ME +30.698 +30.70 23.135
6 MA +155.245 +155.24 18.554
7 TF -79.727 -79.73 8.419
8 DAF -550.472 -550.47 33.764

Sky RMS = 24.06
Popn SD = 25.57

* Correlations between terms:

ID +0.3269
NP -0.8292 -0.3387
CH -0.8553 -0.1352 +0.6089
ME -0.3242 -0.9464 +0.4607 +0.1261
MA +0.3078 +0.8857 -0.1415 -0.1185 -0.7154
TF +0.1358 -0.2360 -0.2045 +0.0825 +0.1687 -0.1558
DAF +0.1257 +0.3517 -0.3402 +0.3920 -0.3554 +0.3399 +0.2689

IH ID NP CH ME MA TF

* fit

coeff change value sigma

1 IH +1.430 -173.58 42.355
2 ID -0.912 +172.02 27.085
3 NP -0.965 +202.44 26.225
4 CH +0.232 -206.33 46.234
5 ME +0.718 +31.42 23.114
6 MA -0.446 +154.80 18.596
7 TF +0.724 -79.00 8.412
8 DAF +1.488 -548.98 33.656

Sky RMS = 24.06
Popn SD = 25.57

* Correlations between terms:

ID +0.3274
NP -0.8283 -0.3378
CH -0.8564 -0.1361 +0.6094
ME -0.3234 -0.9462 +0.4595 +0.1259
MA +0.3077 +0.8867 -0.1405 -0.1188 -0.7163
TF +0.1277 -0.2390 -0.1979 +0.0886 +0.1717 -0.1589
DAF +0.1211 +0.3515 -0.3364 +0.3942 -0.3548 +0.3396 +0.2659

IH ID NP CH ME MA TF

IH : = -173.58
ID : = +172.02
NP : = +202.44
CH : = -206.33
ME : = +31.42
MA : = +154.80
TF : = -79.00
DAF : = -548.98
* GSCAT


PS

Ray Gralack's suggestions:

Well, as I like to say, "best-fit" is not always the "right-fit"! TPoint and MaxPoint are just doing a best fit of the data given to them. Try unchecking a term and you'll see that polar alignment changes without having changed anything physically, so who's to say that the model is accurate enough to be trusted precisely?

In particular, look at the sigma value for the Alt/Az terms. If you believe the model and sigma value, which I am not sure I do, then there is about a 68% likelihood that the value is within 1-sigma of the reported values. That means there is almost a 1 in three chance it's outside the range, again *if* the sigma value is accurate.

So, I would trust drift alignment over the best-fit models. It could be that one or other of the point modeling software apps is not adjusting J2000 coordinates to JNow.

-Ray

[g__day]

Sorry to drone on folk,

I just followed the rest of Gary's advice and added the Harmonic terms(DCEC, DCES, HCEC and HCES), which brings RMS down to 20 arc seconds (so +/- 10 arc seconds = wow!) - now you're talking!

This changes the Tpoint alignment SCP advice yet again to:

Southern Hemisphere

************* AZIMUTH ************
MA: +166 seconds (+2.8 minutes).
Rotate axis East (counterclockwise).
For latitude -33.79°, the azimuth adjustment is 3.3 minutes.
Sigma=15.716

************* ALTITUDE ************
ME: +82 seconds (+1.4 minutes).
The polar axis should be lowered +82 seconds (+1.4 minutes).
Sigma=21.243


* fit

coeff change value sigma

1 IH +70.234 +70.23 35.008
2 ID +93.833 +93.83 45.150
3 NP -27.242 -27.24 15.434
4 CH -170.121 -170.12 36.278
5 ME +81.423 +81.42 21.243
6 MA +165.643 +165.64 15.716
7 TF +91.513 +91.51 38.162
8 HCES +40.758 +40.76 18.575
9 HCEC +124.462 +124.46 26.770
10 DCES -142.175 -142.18 31.814
11 DCEC -56.605 -56.61 45.030
12 DAF -118.588 -118.59 22.289

Sky RMS = 20.27
Popn SD = 22.27

* Correlations between terms:

ID -0.2593
NP -0.5549 -0.1389
CH +0.0004 -0.0070 +0.4676
ME +0.5718 -0.5949 -0.1076 +0.0692
MA -0.0024 +0.1444 +0.2958 -0.0012 -0.5019
TF +0.9704 -0.1218 -0.5491 +0.1521 +0.4671 +0.0285
HCES -0.3847 -0.2571 +0.6540 +0.6933 +0.2381 -0.3068 -0.3131
HCEC -0.7244 +0.1203 +0.1882 -0.6552 -0.3797 -0.0065 -0.8145 -0.0815
DCES -0.8065 +0.6383 +0.2446 -0.1047 -0.4589 -0.2263 -0.7542 +0.1637 +0.6304
DCEC -0.4243 -0.6793 +0.5010 -0.0913 -0.0937 +0.2144 -0.5275 +0.2585 +0.4139 -0.1285
DAF +0.5095 -0.1736 +0.1972 +0.8341 +0.3376 +0.1204 +0.6016 +0.3332 -0.9257 -0.5351 -0.2203

IH ID NP CH ME MA TF HCES HCEC DCES DCEC

Action was taken to correct ill-conditioning. The
current FITTOL tolerance is 0.01, which exceeds
values from 7.38e-006 to 0.00994 encountered
during the fit.

* fit

coeff change value sigma

1 IH +0.062 +70.30 35.025
2 ID -0.002 +93.83 45.206
3 NP +0.301 -26.94 15.434
4 CH -0.072 -170.19 36.258
5 ME +0.210 +81.63 21.241
6 MA -0.071 +165.57 15.717
7 TF -0.156 +91.36 38.170
8 HCES -0.062 +40.70 18.571
9 HCEC +0.225 +124.69 26.775
10 DCES +0.272 -141.90 31.837
11 DCEC +0.031 -56.57 45.012
12 DAF -0.019 -118.61 22.258

Sky RMS = 20.27
Popn SD = 22.27

* Correlations between terms:

ID -0.2597
NP -0.5559 -0.1377
CH +0.0003 -0.0066 +0.4667
ME +0.5717 -0.5966 -0.1090 +0.0680
MA -0.0031 +0.1453 +0.2968 -0.0013 -0.5020
TF +0.9706 -0.1233 -0.5500 +0.1517 +0.4669 +0.0282
HCES -0.3843 -0.2544 +0.6529 +0.6946 +0.2373 -0.3071 -0.3122
HCEC -0.7244 +0.1205 +0.1895 -0.6552 -0.3788 -0.0059 -0.8144 -0.0829
DCES -0.8069 +0.6385 +0.2458 -0.1047 -0.4603 -0.2247 -0.7554 +0.1645 +0.6308
DCEC -0.4238 -0.6796 +0.5013 -0.0911 -0.0912 +0.2137 -0.5260 +0.2561 +0.4134 -0.1284
DAF +0.5100 -0.1737 +0.1955 +0.8337 +0.3372 +0.1198 +0.6020 +0.3340 -0.9259 -0.5357 -0.2200

IH ID NP CH ME MA TF HCES HCEC DCES DCEC

Action was taken to correct ill-conditioning. The
current FITTOL tolerance is 0.01, which exceeds
values from 0.000102 to 0.00996 encountered
during the fit.

IH : = +70.30
ID : = +93.83
NP : = -26.94
CH : = -170.19
ME : = +81.63
MA : = +165.57
TF : = +91.36
HCES : = +40.70
HCEC : = +124.69
DCES : = -141.90
DCEC : = -56.57
DAF : = -118.61

[frolinmod]

Very good. Can you please post an attachment containing a zipped copy of either your .tpt file or an exported (from within Tpoint) .dat file? I'd like to import it into TheSkyX, run a supermodel on it, then post back the results for comparison purposes (i.e., for better or worse).

[g__day]

Great idea, thank you - both files are attached - am very interested to hear what Supermodel produces!

Thinking over the changes last night - it struck me as unusual to the lay person how whilst adding coefficients that decrease sigmas and hone RMS (by adding about 8 harmonic terms RMS went down around 18 - so +/- 9 arc seconds pointing - which seems superb - the polar alignment corrections swung by quite a bit - far more indeed than the improvement in RMS pointing - is that usual?

As Ray Gralack says - doesn't inpsire confidence when a slight tune of a model significantly alters the polar corrections it offers!

Matthew

[frolinmod]

Here are the raw results from supermodel. I didn't attempt to fine tune these at all. All I did was click supermodel and accept. Note that supermodel excluded your point #64. You may or may not have a fork mount. That term still applies though even if you don't. It's not an error. I talked to Patrick Wallace about this and he said to keep it. The term itself will be renamed to something more appropriate for non-fork mounts in a future release of Tpoint. Note that the polar alignment report is to the refracted pole, not the true pole. TheSky6 reports polar alignment for the true pole. TheSkyX reports polar alignment for the refracted pole. My location setting doesn't appear to matter. It's probably data dependent (i.e., is for wherever the data was collected).

[gary]

Hi Matthew,

I only spotted it on your graphs after I had posted the other night but it appears that
you only sampled stars exclusively on the east side of the mount and therefore
missed sampling any with positive hour angles.

I cannot emphasize enough the merit in sampling stars across the whole sky,
which in the case of a GEM, includes performing a meridian flip.

It may be that your western skies are obscured or that for some reason you
can't perform a meridian flip, but if you can, I highly recommend on your next
sampling run you do so.

Some mount errors reverse their direction when the mount is flipped and this
signature can further help betray them. When you sample stars
on one side of the mount and then do the flip and begin to sample stars on the
other, don't be surprised to see your raw RMS pointing error residual suddenly jump up.



What your results so far have demonstrated is that polar alignment is only
one source of error within your system. When you consider how all the
other fabrication errors are entangled with polar misalignment in such a complex
and knotty way and when you consider the relative magnitude of the other
errors compared to the polar misalignment errors, then it will come as no
surprise that as you add or remove terms that are correlated with the polar
misalignment terms, that the polar misalignment terms will swing quite a lot.

Let us put it another way. If one were to try and determine the polar misalignment
without taking into account these other various sources of errors that have relatively
appreciable magnitudes, would one expect to arrive at a reasonable value for the
polar misalignment terms? The answer is, of course not.

Consider a simple exercise in statistics -
One wishes to determine the average age of all students attending a high school which
has a population of 1000 students.

One interviews the first two students and determines their average age to
be 14 and 14.5 and therefore calculates their average age to be 14.25 and the
sample standard deviation to be 0.35. Does this result likely provide a reliable
estimate of the average age of the entire population?

The more data you have, which means in this case the more of the sky you sample,
if the mount is free of random errors, then the more confident one grows of the
result when pointing to some arbitrary position of the sky.

There is no one single magic point in the sky to which one can align the polar axis
which then results over an extended period in zero field rotation for all RA/Dec points
in the sky.

In other words, if you do a drift test and tweak the mount to achieve minimal field
rotation at some point in the sky, don't expect that tweak to be as good across
the whole sky. In that sense, a drift test is only providing a small sample. If that
is the only point of the sky you plan on imaging and your imaging times
are sufficiently short that refraction and the other errors don't start to create
significant field rotation, then that may be all you need.

But if you want to achieve the best whole sky pointing performance and with
it the best chance of nailing where the pole is as a reference point, then the optimal
prescription is eliminating random errors with the mount/OTA, sampling a
large number of points across the whole sky, performing a statistical analysis of
the type TPOINT performs, optionally tweaking the polar elevation to help counter
for the effects of refraction in the area of the sky you are going to image and most
importantly, have a pointing kernel which is applying the same pointing model and
therefore the correct associated dynamic tracking rates.

Again I will be touching on some of this at Lostock.

By the way, an RMS of 18 arc seconds refers to a radius, so if the errors are
normally distributed, roughly 68% will fall in a radius of 18 arc seconds, which is
of course a diameter of 36 arc seconds.

Ernie mentioned using FO which as he points out is legitimate to do on a GEM
in a "if the glove fits" kind of way. However, because the TPOINT methodology
is to try and provide terms that mimic some expected mechanical phenomena,
then I would recommend that if it does not significantly improve the model
compared to those terms that have a ready and straight forward mechanical
explanation, such as HCES, then I would drop FO in favour of these more
less surprising terms. In other words, apply Occam's Razor. Then only use higher
order harmonics or polynomials or terms such as FO to 'mop up' anything that can't
be readily accounted for. :thumbsup:

[g__day]

Gary,

Again - thanks for your insight. Yes my sampling was only in one quadrant South East to North East, reading Tpoints documentation that was the way I thought you were supposed to start off!

MaxPoint was equally distributed over half the sky on my last run (see below).

Next steps - when the clouds finally clear:

1. Ray Gralack suggested I re-run PEMPro - just to see if the pier has settled since my last run - happy to do that.

2. I understand the need for a realistic sample size - I think I'll shoot for something with several hundred data points. I have the full pintpoint and AutoMapper and MaximDL (but not CCDSoft v5) - so I might as well put that what I have to automated used and try an automated sky map with plate solves to build a more realistic model.

MaxPoint states that the circles it places around stars are at 1 and 3 sigmas and the numbers are the diameters of these circles - not the radius. That's why I read MaxPoint saying I had a radius of +/- 16 arc seconds for 1 sigma. So given Tpoint's circle said 20.3 arc seconds - I presumed that was the diameter of the circle - not the radius - I didn't image it would be twice as bad as MaxPoint's model.

3. Once I have MaxPoint auto calibration done I'll post results, then I'll try AAG_Tpoint mapper (which works with SKy 6 PE and MaximDL) and see what result it gives me with a similar sized data set.

Challenging to nail down that final bit of polar alignment - when you want to go do very low arc minutes of even a few tens of arc seconds. Hope it doesn't sent me crazy!

Thanks,

Matthew

[PTW]

For Matthew's data, I get a slightly different Super Model:

coeff change value sigma

1 IH +0.000 +735.72 202.978
2 ID -0.000 -520.59 153.884
3 & HHSH4 +0.000 -600.58 175.044
4 & HHSH5 -0.000 +281.81 81.956
5 & HHCH5 -0.000 +211.24 60.775
6 & HDSD -0.000 -281.12 44.287
7 & HDSD4 +0.000 +16.88 4.663
8 & HDSH2 +0.000 -885.09 270.273
9 & HDSH3 -0.000 +370.05 112.750
10 & HDCH3 -0.000 +228.06 71.055
11 NP -0.000 -364.09 196.648
12 CH +0.000 -82.02 39.926
13 & HXCZ -0.000 -1009.76 367.233
14 ME +0.000 +298.88 68.782
15 MA +0.000 +358.46 77.630
16 TF +0.000 +220.96 48.128

Sky RMS = 16.58
Popn SD = 18.88

I also get these recommendations for polar axis alignment:

Polar axis settings:

ME MA

+299 +358 current setting
+88 +0 refracted pole
+48 +0 minimize field rotation
+40 +790 minimize unguided declination drift
+175 +754 minimize unguided total drift

The polar axis is currently 299 arcsec above and 358 arcsec
to the right of the true pole.

To reach the refracted pole, lower the polar axis by
211 arcsec and rotate the mount anticlockwise by 431 arcsec.

To minimize field rotation, lower the polar axis by
251 arcsec and rotate the mount anticlockwise by 431 arcsec.

To minimize unguided declination drift, lower the polar axis by
259 arcsec and rotate the mount clockwise by 520 arcsec.

To minimize unguided total drift, lower the polar axis by
124 arcsec and rotate the mount clockwise by 476 arcsec.

[g__day]

I think this might be the year I shift to The SkyX + Tpoint! I have done very, very little astronomy lately - works too busy and I am getting fit with Aikido, training for my senior brown belt and parts of the black belt grading...

Don't know what it is like for others but I find it surprising how one day you wake up and just want to prioritise some activities much higher then others, then some random time later - everythings re-orders the priority list again!

Reckon I spent almost more time trying to tune my gear than use it - but that is a common trait in almost all I do!


Thank you for the continuing advice - I will be sure to use it! :)