PinPoint

[Big Bang!]

Good evening all.

Complete new kid here; words of one syllable please, I know nothing. 

I have had a 10 Micron 1000HPS mount for about 8 months and am having great fun with it.  Thus far I have done reasonably well building internal models by using my ATIK to visually centre stars.  As Per will confirm, these mounts are capable of 20/30 or more minutes unguided but I am struggling to get beyond 10 and as I get more adventurous this is holding me back. Clearly, I am operating at the limit of my models.   In layman's terms, how does one go about plate-solving? I have PinPoint on trial. As I understand it so far, all PinPoint does is solve the image and then annotate the FITS Header with relevant data. How does one get the data from the FITS into the mount?  Can that be done without further software? Are Maxim DL and PinPoint still required in order to use Per's Model Maker? That's a lot of dosh. Having spent a fortune on a really good mount it makes no sense to limit its performance so if I have to spend the money, so be it - I just want to be sure I am not spending unnecessarily and it is my lack of knowledge that needs addressing first.

So the question is, what do I need in order to get the best possible modelling solution?

Grateful for any help.

Gus

[ChrisLX200]

Well unfortunately at the moment ModelMaker does need MaximDL in order to run but hopefully it will be updated in the near future so that it doesn't. I wouldn't go buying MaximDL just for this purpose though - it's far too expensive to be a one-trick-pony. All MaximDL is doing in this case is capturing an image so that PinPoint can solve it - but ModelMaker cannot be configured to use anything else. You can still create models manually using a plate solver (like PinPoint) because all that ModelMaker is doing is automating the slewing to sync points in the sky, using Maxim to take an image, Pinpoint to solve, then sending the position data to the mount. So the first thing to do is to get PinPoint working.

Personally I use SGPro (Sequence Generator Pro) to capture images, and with that I can (if needed) take an image of the sky manually, plate-solve its position, and use SGPro to synchronise the position to the mount (either for creating the 3-point base model or to add more refine points). After each sync I would have to manually slew the mount to a new position and do it again 0 that's the bit that ModelMaker automates. It's quite an involved process running one of these mounts and I cannot possibly explain in detail everything I do but if you have any specific questions then fire away.

ChrisH

[Big Bang!]

Morning Chris,

Thank you!  That is exactly the clear explanation of functions that I needed to understand.  

I use the ATIK's own software to capture images and then PixInsight for post-capture calibration etc.  I have found this satisfactory for my needs, up to now at least.  Currently, therefore,  my only reason for buying Maxim would be to enable Per's Model Maker and as you say that is awfully expensive - especially as we remain hopeful of an upgrade to Model Maker that wouldn't require it.

 

I have PinPoint on trial; I have managed to get it to solve images and to annotate the FITS header with position data - seems to take about 5 seconds per image. The question is how do I get the position data into the mount if I don't use Maxim DL?  SGPro is much cheaper and your description of procedure seems straight forward enough.  

I would manually slew the telescope to each position; use SGPro to capture an image, PinPoint to solve it and then SGPro to send the data to the mount to add a sync point to the model?   Yes, there would be no automation but I would be far more likely to get a good model than the way I am doing it now (eyeball centering) and that has got to be worth the $149 + $99 for PinPoint and SGPro.  If I have understood you correctly this solution would certainly do until Per finally gets his finger out!  

 

Very grateful for your guidance.  It is so easy to make expensive purchase mistakes in this!!

 

Gus

 

 

[ChrisLX200]

You never need to save the file during this process so the FITs header is irrelevant. PinPoint is a great addition to your toolkit and I can recommend it - but it isn't essential either when using SGPro because it has its own plate solver. With SGPro you would slew to a point in the sky (for your first 3 base points which should be well separated and use both sides of the mount) and choose from the tools menu Solve and Sync (the Obs OC is not in front of me right now so I can't give an accurate menu choice...). You need to have the mount set to Syncs Refine=OFF for those first 3 points (use the virtual hand controler to check but I think that's the default unless you change it). Do that 3 times and those are your base points done and you now want to add refinement points so you first need to toggle the mount back to Syncs Refine=ON. You MUST remember to set that correctly before adding refinement points or you'll screw up the whole process! Then carry on as before, slewing to new random positions covering the whole sky, both sides of the mount, high and low DEC, and using SGP to solve and sync (it's a one-click operation that - SGP takes and image, plate-solves it, and sends the sync command to the mount). Doesn't take too long but it's boring, and of course you have to choose each of the points yourself - it's easier to use a planetarium program to do that (like StarryNight for e.g.,) just right-click on the display and choose 'Slew Here'. ModelMaker remembers the set of model points for your location which avoids trees and you neighbour's house - you'll just have to guess without it ;-)

ChrisH

[Big Bang!]

Thank you, excellent support.  I'm going to buy both now.  I can then be fairly sure that my model is as good as I can make it - there's just no point buying a mount of this quality otherwise.

 

Most grateful.

 

Gus

[Big Bang!]

Help, please!

I have purchased both SGPro and PinPoint and have spent 4 hours out in the cold this evening trying to understand them.  

I had initially thought PinPoint was solving frames taken using the ATIK sequencer because it was creating an additional file and putting some RA/DEC data into the FITS header; seems it wasn't, even though I had been careful to enter the plate parameters accurately. So I tried "All-Sky" (no RA/DEC hint or plate data required) and immediately it routinely solved in about 35 seconds, populating the FITS with RA/DEC data down to the 3rd decimal place! I was well impressed. 

 

So I opened up SGPro and set that up with Camera, Mount etc, being very careful to set the plate parameters correctly.  (My pixels are 3.69 microns and with a 2000mm fl that gives me 0.38"/pix at 1x1) I set PinPoint as the plate-solver and configured for 5s exposure at 2x2.  When I clicked on Solve & Sync it set off as if it was doing something (green bar apparently showing progress) but no end result - even after minutes.  Tried "Solve & Sync Blind" many times; still no joy.

 

Then, at the end of the evening I took a single frame using the SGPro sequencer (rather than Artemis' - the ATIK software).  Opened it up in PinPoint and it direct solved in 0.9 seconds, no "All-Sky" necessary.  This suggests a conflict?  Any known issues between SGPro and Artemis?  All evening I have had difficulties with software crashes and restarts including one "blue-screen" shut down. Never had any issues of this kind before with Artemis. 

 

In summary, PinPoint wouldn't solve frames taken by the Artemis sequencer but did solve a frame taken by the SGPro sequencer.  However, SGPro did not solve a single frame all evening - even with PinPoint set as its solver in the control panel, whereas stand alone PinPoint did.  I am flummoxed and I need guidance from somebody who knows what they are doing. I assume I am not doing/setting something correctly.

 

Gus

 

 

 

Solve 3.txt

[ChrisLX200]

Well the first rule is - Don't Panic!  These are complicated setups and the chances of getting it all working on the night are pretty slim.

Looking at your Solve3.txt file the plate scale is actually 0.44 not 0.38 - an error which shouldn't stop Pinpoint from solving (but it would stop Elbrus because it demands the exact plate scale to work with). Pinpoint should be solving any particular plate within a couple of seconds _provided_ it receives the correct 'hint' (i.e., it knows roughly where in the sky the scope is pointing). If it fails to solve then you would need to do a Blind Solve where it looks for a match in the whole sky - that takes much longer and can be 20sec or so.

I would put Artemis Capture aside for now, you don't need it and it only adds complication/confusion to an already complex operation. Check that all your kit is connected properly in SGPro (camera, mount, focuser, filterwheel), and for this operation I use the Red filter for better star contrast as it darkens the background sky.

The first thing to check is Pinpoint solving - take an image as you describe above (5sec @ 2x2) then right-click on the image you have two solving options - Solve and Blind Solve. Look at the data in the dialog box and check it makes sense - i.e., the plate scale and positional hint it will be sending (the RA/DEC coordinates), the binning is 2x2. The RA/DEC it will be getting from the mount (or where the mount thinks it is pointing at the time). Choose Solve and after a couple of seconds the result box should pop up with the data. If it doesn't then choose Blind Solve and you'll have to wait a bit longer for the results box. If successfull check the data in the box and see if you can spot where the Solve option was failing - in all probability the positional 'hint' it was sending was too far out (Pinpoint only looks within a narrow region for a 'Solve', if the pointing is further out than that then you have to 'Blind Solve' until pointing is more accurate).

If your mount was being setup for the first time (i.e., not an observatory mount) the you may have to Blind Solve the first three points you will be using as Base Points. After that pointing should be accurate enough to enable you to use Solve.

So lets assume you now have SGPro solving the images you are capturing. At this point you want to do the 3 Base Points, so set the Syncs Refine=OFF parameter on the handpad, point the scope somewhere in the sky for the first point and take your image. Check it solves OK by right-clicking (this is an extra check!). Then choose  the menu item for plate solving from SGPro's Tools Tab and click the the option Solve & Sync (or Blind Solve & Sync if required). It will take another image, download, solve (or Blind Solve), and send that sync data to the mount. This may take some time but you can check progress both with the green bar and messages on the bottom line. After a while it should pop up a message saying the operation was successful. You will need to repeat this 3 times for the 3 base points, slewing to different (well-separated) parts of the sky.

If that operation was successful then you change the parameter on the hand controler to Syn Refines=ON and continue taking as many extra refinement points you need to - I use the full 100 points but then I'm using ModelMaker which is automating the process, doing it manually I would aim for at least 20 points.

I will stop at this point and await feedback from you whether this is working...

ChrisH

[Big Bang!]

Hi Chris,

Thank you so much for your help here.

When I first set up the filters in SGPro a window came up that said something like, "In order to use the internal filter wheel the camera must have already been opened in another programme."  I  took that to mean that I had to have Artemis running in the background in order to allow SGPro to see the Camera. I was wrong.  I have just run the set up without Artemis and SGPro sees the camera OK.  I am now fairly sure this was the cause of the continuous conflicts and crashes - big issue solved, no crashed so far today.  This begs the question, does the same apply to PinPoint? Does SGPro activate it as required?

 

When it came to setting the positional hint I took data off the telescope's handset.  Obviously, being at the beginning of the modelling process there was no active model in the  mount and so there would inevitably have been some initial error there.  What is the best way of obtaining positional data for the hint? Just noticed that Artemis is presenting the images flipped compared with SGPro.  Could that be an issue?

 

As far as scale is concerned, both the SGPro calculator and my own maths say 0.38 @1x1 but I see you are correct, the scale is 0.44 in the solve report.  Is that big enough to trip up the solve?

 

Sorry, I am typing as results come in here... Just ran last night's image taken on SGPro; right click "solve" and it got it immediately and did the same image "blind" in 105 seconds showing scale 0.76 (which is correct @ 2x2).  However, it will not "solve" images off the Artemis Sequencer "possibly due to input plate scaling..." but it will "blind" solve that image (in 36s with %100 confidence).  

 

That is a good start, thank you.  Don't think I can do much more now until we get a clear night when I would hope to be able to get a 3 point basic refined by 20 or 30 into the mount and then have it stare at something for 30 mins or so to see how much better it is. 

Can't thank you enough,

 

Gus

[ChrisLX200]

51 minutes ago, Big Bang! said:

Hi Chris,

Thank you so much for your help here.

When I first set up the filters in SGPro a window came up that said something like, "In order to use the internal filter wheel the camera must have already been opened in another programme."  I  took that to mean that I had to have Artemis running in the background in order to allow SGPro to see the Camera. I was wrong.  I have just run the set up without Artemis and SGPro sees the camera OK.  I am now fairly sure this was the cause of the continuous conflicts and crashes - big issue solved, no crashed so far today.  This begs the question, does the same apply to PinPoint? Does SGPro activate it as required?

 

When it came to setting the positional hint I took data off the telescope's handset.  Obviously, being at the beginning of the modelling process there was no active model in the  mount and so there would inevitably have been some initial error there.  What is the best way of obtaining positional data for the hint? Just noticed that Artemis is presenting the images flipped compared with SGPro.  Could that be an issue?

Just don't even start Artemis, it's neither needed nor wanted! PinPoint has two modes working either as a stand-alone plate solver (Visual PinPoint) or as a scriptable plate solving 'engine' (ie., it works in the background and SGP calls it as a subroutine, sends it data, and Pinpoint returns the result to SGP). You never see PinPoint when working like this, it's being used as a plate solving engine. SGP always uses PinPoint as a service - an engine - so the PinPoint window never appears.

The first point for a new setup will be entirely wrong - or even showing as 00 RA and 00 DEC - so you will have to blind solve. If your PA is accurate then the second and third base points may just solve but more than likely you will have to blind solve for all three. SGP should 'know' which side of the pier the mount is on (East or West) because the mount tells it - but whether the image is flipped depends on your optics. There is an option (a check box) in SGP to change it to 'flipped' but if you have already managed to solve any plates then you don't need to bother as whatever you have in there is obviously correct.

51 minutes ago, Big Bang! said:

As far as scale is concerned, both the SGPro calculator and my own maths say 0.38 @1x1 but I see you are correct, the scale is 0.44 in the solve report.  Is that big enough to trip up the solve?

No, that should be fine for PinPoint - Elbrus (as I said) would fail with that error in plate scale though (one of the advantages of using PinPoint, it's much more relaxed about plate scale, you can have an error of 25% by default and it will still solve - and you can even increase that parameter if needed).

One extra point - you need to have Dual Axis Drive enabled on your mount (check your hand controller whether this is selected or not).

You also need to be 'reasonably' well Polar Aligned, and that adjustment is usually done first before moving on to adding refine points (at least I do). So you have your first 3 base points in the mount, if you then look at the Polar Alignment Info that the hand controler will show you (you need to memorise the menu structure of the controller so you can find these things ;-) ) then it will tell you roughly how well you are polar aligned. It will also make a suggestion as to how to adjust the ALT/AZ screws to get better alignment (like, AZ 0.5 turns to the Left for example). You can make this adjustment now and repeat the 3 base points. You should be able to get within an arc-minute or less of true PA quite quickly. This is an iterative process and although using Dual Axis Drive means an error in polar alignment will not affect either tracking or pointing (which may surprise you!) a polar alignment error WILL result in field rotation. So it's best to get it fairly accurate and for myself I aim for 30 arcsec from true aligment with the pole. Any more accurate than that is nit-picking and probably not worth the effort - unless you are going for 3hr subs or something.... Anyway, after you have good PA and 3 valid base points you can then move on to creating the refinement model.

51 minutes ago, Big Bang! said:

Sorry, I am typing as results come in here... Just ran last night's image taken on SGPro; right click "solve" and it got it immediately and did the same image "blind" in 105 seconds showing scale 0.76 (which is correct @ 2x2).  However, it will not "solve" images off the Artemis Sequencer "possibly due to input plate scaling..." but it will "blind" solve that image (in 36s with %100 confidence).  

That is a good start, thank you.  Don't think I can do much more now until we get a clear night when I would hope to be able to get a 3 point basic refined by 20 or 30 into the mount and then have it stare at something for 30 mins or so to see how much better it is. 

Can't thank you enough,

 

Gus

I think you must be using the same Atik camera as myself (490EX? or the Atik One equivalent with the built-in filter wheel...). You will need to using it at 2x2 binning at this focal length because you are way over-sampling at 0.3 arcsec/pxl! Even 0.7 is over-sampling but at least it's not quite as silly!

Update this thread with your progress and I'll keep an eye out for it. Prod me if I happen to miss it!

ChrisH

 

 

[Big Bang!]

Bingo!

Just managed to get one frame before the clouds rolled in. Deliberately chose a fairly obscure piece of sky with nothing much in it; right click, blind solved in 3 mins, 100% confidence. Showers due but I may well get some more clear sky tonight so I might get a chance to get a full model done. 

 

Procedure would be? Clear align, Refines - Off and then let SGPro  sync the mount from there? After three basics Refines - On and then keep going.

 

Gus

[Big Bang!]

I have an Atik One 9.  I discussed sampling with Ian King before I bought it.  He recommended the One 6 but because I also have a 75mm APO of 474mm FL I felt I needed to go for the 9 and then bin for the longer scope - otherwise I would have been sampling at 2.4"/pix in the small scope and I felt that was just too far under-sampled.  I have not found it a problem so far - not that I would necessarily recognise it if it were a problem.

 

Gus

[ChrisLX200]

14 minutes ago, Big Bang! said:

Bingo!

Just managed to get one frame before the clouds rolled in. Deliberately chose a fairly obscure piece of sky with nothing much in it; right click, blind solved in 3 mins, 100% confidence. Showers due but I may well get some more clear sky tonight so I might get a chance to get a full model done. 

 

Procedure would be? Clear align, Refines - Off and then let SGPro  sync the mount from there? After three basics Refines - On and then keep going.

 

Gus

3 Minutes?? wow, that's a long time... What is this running on may I ask? not that it should make that much difference really... It usually takes around 30secs on mine to blind-solve, and one or two seconds for a normal solve. You have the procedure correct, hope the speed increases when you have the 3 base points in there! Be aware the mount will remember the model data from the last time out, so if again setting up from scratch (if you put all the gear away and have to take it out to the garden again...) then you will need to clear the model in the mount before you start over. The existing model will be invalid.

[russp]

FWIW I use SGPpro to plate solve for automatic target aquisition and it works really really well and means I don't have to bother with star alignment or models at all, just get a half decent polar alignment and use guiding. I'm not sure why you'd need or want an accurate model of the whole sky if you're putting gear in and out and you're imaging. Set up scope and polar align - I just set up a target in SGP and let the plate solving do it's thing to get the scope pointed at the exact right bit of the sky, and then let the guide camera do it's thing for the imaging session. I don't use pinpoint so can't help with the setup of that but I do recall setting up plate solving in SGP took me a deal of head scratching - the good thing is that you can use the targeting module to platesolve images you get off the net to see if it works and don't have to waste imaging/observing time doing that  Excuse me if I've firmly grasped the wrong end of the stick.

 

[Big Bang!]

Its running on a good laptop - i7 processor 2x 240mhz, 8GB; absolutely man-enough for this.  I've done a couple more and they "Blind" solved in about 30 secs so I guess the initial pointing must have been way out. However, even after the first 3 or 4 they won't "solve" saying that scale may be too far out, which is a surprise because it has calculated the scale itself on the first 3.

 

After about 6 I went into the handset menu to look at the Align Info but it said no complex model.  So how do I know that SGPro has send align info back to the mount after each sync?

 

Gus

[russp]

I forget if SGP will work alongside a handset? How do you have SGP and the handset connected to your mount, I only ever use one or the other but never both. SGP connects through ASCOM on my set up - Theres only the one connection on my mount .. Either the align info is in the chips in the handset OR it's in the chips in the laptop (via EQMOD) - not sure how SGP is supposed to talk to your handset? 

 

[Big Bang!]

The 10 Micron mounts have their own internal computer, Linux based I believe so it can be operated PC-free and that is where the models are stored.  SGP talks to the mount through the PC (Ethernet or Wireless) and the Handset has its own direct input.  You can also use a virtual handset on the PC and what ever you do on that is replicated on the physical handset.

 

Gus

[Big Bang!]

I think I see my mistake.  If you right-click on a frame to bring up the option to "Solve" there is no "Sync" with it and one therefore assumes no data is sent to the mount.  I guess correct procedure would be to Plate Solve via the Control Panel as that provides a "Solve and Sync" option.

 

Expected to be clear all night down here so I am hopeful of good news tomorrow.

 

Gus

[Big Bang!]

Good progress tonight.

First frame solved in 35s, second and third in 20s.  Then did 40 refine points with Refines Sync ON followed by a brief examination of the database for rogue syncs.  Two attempts at a model have both ended up with an RMS  of 5.5".

 

I then decided to have the mount stare at M81 for 15 minutes; the completely raw frame is attached.  As would be expected, much noise etc but I would like an opinion about the stars and apparent tracking.  This is through a red filter.

 

Many thanks indeed for getting me this far into SGPro/PinPoint - SO much easier than the mark 1 eyeball and a digit.

 

Gus

 

 

[ChrisLX200]

Stars look pretty good to me, especially at this long focal length. What did you do about refraction data? It's not critical above 30deg DEC but if tracking lower that extra correction parameter becomes increasingly more important the lower you get.

ChrisH

[Physopto]

Well I have just read through this thread. Very interesting.

I bought my GM1000HPS last year. I do use Maxim DL as I have had it for several years. I also have Per's Model Maker and his StickStation. The last of which I have to be honest I have not been able to try out for any length of time as yet. It does seem to do what it says on the tin though. I control via The SkyX. 

I managed last year to get the mount fully operational in the solving department at Kielder. MM automated the alignment process with stunning results. As Chris says the first three points have to be repeated as it is an iterative process. But you must remember to clear the points in the model between each alignment after the manual repositioning of the mount by the adjuster screws and Dec settings. Otherwise the model will not work correctly, as it will then try to retain the wrong settings I believe. Once the initial polar alignment is set within a few arcseconds you can redo the initial alignment and just carry on with the fine model points. Each point taken over the sky will enhance the sky model and as it progresses the mount computer will recalculate the accuracy as it proceeds. I have never used any other programs but those mentioned to control the Micron so cannot help in that area. 

I have found the GM1000HPS to be almost unbelievably accurate in its pointing. Every point I send it to is spot on central in the image to within a few arcseconds. AND I do know I made some slight mistakes along the way initially! 

The plate solves at Kielder took less than a second each. At first I was convinced it was not possible and went looking for an error. But it was working! Almost unbelievably quick. If you can get the new MM that Per is working on that does not need Maxim DL you will be in for a treat.  

The only proviso is to watch your mount to get used to the automated slews. It can get uncomfortably close to bashing your CCD into the mount legs. The Micron can slew far too fast. I slowed mine down to 5 deg/sec during the initial testing. You need to view the MM screen and judge your points carefully so as not to let the mount crash your scope into the legs. 

Best of luck.

Derek

[Big Bang!]

Thanks chaps,

 

Refraction data: I have one of Per's StickStation's too but, erm, there I have been a bit of a numpty.  Without thinking I have been putting it in a USB port on the exhaust side of my laptop's heat ventilation and have spent ages pondering why it is showing quite a few degrees above that shown on the ATIK and Focuser!! The light came on this morning - no more to be said really.

 

The new level of modelling accuracy that I now have means it is now worth going back over the other set up characteristics: I have just re-leveled the bubble and although Polar Alignment was showing a respectable 1' 15" (the correction figures are only 0.01Lf and 0.01Dn) I am going to have a go at improving that tonight.

 

May I ask what RMS you guys are getting? I'm at 5.5" at the moment - how much better ought I be able to get it?  Fortunately, I am in a permanent set up with no nightly dismantling is required so I think it is worth spending some time minimising that.  What is the longest unguided you can go and still get round stars?

 

Gus

[Physopto]

Ahhhh Round stars!

As Olly has said, not an indication of much. You can get them round even with periodic error and poor seeing, just statistics. I have as yet not bothered with measuring RMS error. Too early for me to bother at present, so I can't help there. Your figures for polar alignment are OK. As the instructions say you really do not need to try to improve that. I had 0.01 (right or left unsure now) and 0.0Dn/Up at Kielder. It was good enough. If I had a permanent mounted setup, then I would  to try to improve it as you want to do. I made up a disc about 2.5" diameter and fixed it over the Dec adjustment knob. I graduated it in 10 divisions and a subset of ten. That allows me a better, more precise  and quicker adjustment. I also fixed a similar visible set of numbers and divisions to the Plastic Lock Knob Spanner ( for the want of a better name) for use on the azimuth adjusters. It just makes life a bit easier.

I think from memory you should get much lower RMS than 5.5" but it will depend upon you individual setup and skills.

I use the Stick Station on a short USB lead but out at the mount so it is sheltered and reads the correct temperatures. Definitely keep away from the computer, sorry, still having a chuckle at that one . I won't mention all the daft things I've done.......

Derek

[ChrisLX200]

The Polar Alignment is OK, but it isn't difficult getting it down to 30" (30 arcsec) if you feel like making the effort. I doubt you'll notice any difference though. The RMS is a bit of an oddball in that you may get a better functioning model sy using more refinement points yet which shows a larger RMS. That's because with the maximum number of samples (100 points) you are getting a more detailed picture of the flexure, whereas with fewer (say, 20 points) the errors are 'smoothed out' somewhat (so it actually performs worse). So don't worry about the RMS too much. For sure, 5" is plenty good enough to get you perfectly acceptable unguided tracking (in fact, anything less than 10" would be good but it does depend on the focal length at which you are imaging whether any trailing eventually becomes significant).

For myself I get 3.5" with the refractor (it being a very stiff setup with little or no flexure), my 8"RC gives me around 5", and I don't have enough data with the 12" ODK to say but the few occasions I've run a complete model it was around 6". All of those were using 100 points and all are plenty good enough. With the Refractor I've gone out to an hour subs just for a test but generally I use 30mins now (narrow band). Another important metric is the pointing accuracy - with the refractor it is always exactly on target to the pixel, and that is _both_ sides of the pier. I once did an all-night run imaging across the meridian so half were on the East side and half on the West with a flip in between. When it came to stacking this data all the frames stacked perfectly with no artifacts (overlaps) around the edge. If you think about that - the pointing accuracy needed to perform that feat - it is a remarkable achievement. Needless to say I can happily choose a very faint object to image confident in the expectation it will be framed perfectly.

ChrisH

[Big Bang!]

I 've managed to get an hour in the Obs this evening.  Fresh model; Polar Align now down to 30" - I had difficulty accurately moving the knobs to get any better than that but I would imagine field rotation should be undetectable at the sort of exposure lengths I envisage and therefore that is plenty good enough.  I managed to get 39 points into the model and the RMS is down to 4.8", compared to 5.5 yesterday. I understand the point about the RMS.  In view of that, can I just keep adding points to this model up to 100?

I am truly astonished to hear about your all night run - that seems like the stuff of professionals and very expensive kit.  I would really like to be able to image really faint galaxies and that has been my driver but I wonder if at 7.8 I am a bit too slow for that; perhaps a future upgrade. I cannot seem to get past 15 minutes without the stars starting to trail - how I would love to get to 30.  

Gus

[ChrisLX200]

2 hours ago, Big Bang! said:

I 've managed to get an hour in the Obs this evening.  Fresh model; Polar Align now down to 30" - I had difficulty accurately moving the knobs to get any better than that but I would imagine field rotation should be undetectable at the sort of exposure lengths I envisage and therefore that is plenty good enough.  I managed to get 39 points into the model and the RMS is down to 4.8", compared to 5.5 yesterday. I understand the point about the RMS.  In view of that, can I just keep adding points to this model up to 100?

I am truly astonished to hear about your all night run - that seems like the stuff of professionals and very expensive kit.  I would really like to be able to image really faint galaxies and that has been my driver but I wonder if at 7.8 I am a bit too slow for that; perhaps a future upgrade. I cannot seem to get past 15 minutes without the stars starting to trail - how I would love to get to 30.  

Gus

Yes it's difficult making these tiny tweaks to the adjustment knobs, just tightening up the opposite knob tends to shift it a little bit. You can keep adding refinement points to your model provided nothing has changed since the last time you added points (and I mean nothing - even ambient temperature can cause shrinkage/expansion in places that may affect it slightly). Using ModelMaker it's no effort to create a new model, which is just as well because in my case I have to unmount/remount the OTA each time (the mount itself stays on the pier) and that's enough to require a new model each time. I did try at one time keeping the model and just re-syncing it (a single point plate-solve will shift the entire model by whatever the offset was in pointing), but it wasn't very successful. The mount would appear to track OK on one side of the mount but after a flip pointing was then a good bit out (like 20-50 pixels). I suspect the cause was slightly different balance because the scope didn't go back in the dovetail in exactly the same position as last time. Anyway, it takes around 30-40 minutes to run a full model (depends on which camera I use) and it can be done in twilight by using a red filter so I don't lose any imaging time by doing this.

You should certainly be able to get beyond 15minutes and if it won't do that then perhaps there is some random (un-modelled) flexure in your system causing a problem (the focuser/extensions are the first place to look, then cables to see how well they are routed such they don't pull on the camera). Too early to say whether random flexure is a component because I doubt you have fully optimised your setup yet. My 12" ODK is 2040mm f/l and f/6.8 which I hope will be good for faint/small objects, so f/7.8 is not so much different.

ChrisH