si6pack Posted March 30, 2014 Report Share Posted March 30, 2014 Hi Robert, I was too curious - I gave it a try already yesterday! I didn't change anything but made my test sequence file MLPT positive... To cut a long story short, first image stars were lines, second image stars were still lines but shorter, third image little improvement, all following images stayed the same - short line shaped stars Switching back to a NON-MLPT-based sequence file, round stars, no issue! Any ideas what is going on there? Regards, Christian Link to comment Share on other sites More sharing options...
robertp Posted March 30, 2014 Report Share Posted March 30, 2014 Hi Christian, hm, interesting. So you also see worse tracking for the first few images. Philipp thinks that it might be that the scope needs more time to settle after the MLPT-calculation with the various slews finished. One thing that I want to try is to slew to the imaging target, slew e.g. 2 degrees further east of the target in RA, do the MLPT calculation there, send it to autoslew, slew "forward" to the imaging target and start to image. If the tracking is then perfect right from the beginning, it might indicate a problem with the first portion of the fitting curve. I still can't understand why you do not get decent tracking at least after a few minutes. I'm not sure whether you tried the following: * clear the configuration, so that no pointing model is in use * sync with Sequence * do the three star polar alignment procedure with Sequence and check whether polar misalignment is smaller than 1 arcminute. * slew to the imaging target (might be a little off center as you have no pointing model in use), synch with Sequence, slew to the image target again (it should be dead center now) * do an MLPT with e.g. 6 points and 1 hour imaging time * start to image e.g. 5 minute exposures So basically you do an MLPT without any pointing model with an appropriate polar alignment. This is under the assumption that you have the fixed encoders and do not have to do an encoder calibration. One thing that is vital is that you have a working GPS connection. To be on the safe side even thought it should not matter make sure that before you start with the MLPT the computer time is as accurate as possible (sync it via the internet). Maybe you can do some screenshots on the way, e.g. showing your polar alignment results and the fitting curve in Sequence. Good luck, Robert Link to comment Share on other sites More sharing options...
si6pack Posted March 30, 2014 Report Share Posted March 30, 2014 Hi Robert, Thank you very much for your detailed advice - I will give it a try tonight! Concerning GPS: I am using WinTec GPS devices - WBT202 or WBT301 coupled via bluetooth, everything looks fine there... I suppose, that the MLPT function in a sequence series uses the same algorithm as the dedicted one in SEQ (MLPT button)!? To make this cristal clear: So far I always used "embeded" MLPT during a series run in SEQ, never the dedicated function... I will do some screenshots - my wife will kill me, if I do spend my time once again outside... ;-) I keep you posted! Thank you, Christian Link to comment Share on other sites More sharing options...
robertp Posted March 30, 2014 Report Share Posted March 30, 2014 Hi Christian, Maybe you could try the dedicated MLPT-function to rule out some weird issue coming from the additional functionality of the automation processes. I would expect that the algorithms are identical but for a quick test the dedicated function is quicker anyway as you don't have to select a target etc. Please tell your wife that your tests are of great value for the community and that we, the community, really appreciate her patience ;-) Good luck, Robert Link to comment Share on other sites More sharing options...
si6pack Posted March 30, 2014 Report Share Posted March 30, 2014 8-) will do! Cs, Christian Link to comment Share on other sites More sharing options...
si6pack Posted March 31, 2014 Report Share Posted March 31, 2014 Hi Robert, here is my readout from yesterday's session... I strictly followed your instructions. Before I started, I carefully revisited my levelling and corrected it minimally... homefind, cleared old config sync within SEQ 3 star polar alignment on E - RMS RA before 0,77 arc min, RMS DE error before 1,05 arc min resulted in 6,6 arc min Alt up and 6,6 arc min Azi cw redone polar alignment to get below 1 arc min - for any case, I deleted old config (which should have been deleted anyway) 2nd 3 star polar alignment on E - RMS RA before 0,05 arc min, RMS DE error before 0,40 arc min resulted in 0,5 arc min Alti down and 1,6 arc min Azi cw redone polar alignment to check final polar alignment error - RMS RA before 0,02 arc min, RMS DE error before 0,09 arc min (would have resulted in 0,1 arc min Alti down and 0,3 arc min Azi cw) slew to imaging target (M106) - SEQ sync MLPT 100 min 6 points MLPT calculation: RA flatline, max RA error 6,0 arc sec MLPT calculation DEC constantly decreasing, max DE error 44,2 arc sec LPT forwarded to AS To cut a long story short, first image stars were lines, second image stars still elongated, third image egg-like shape, forth image dito, fifth image no improvement... GPS readings ok Then I used SEQ with a sequence file with MLPT enabled for the same object - round stars (300 sec exposures)!?!?!? To be honest, I have no explanation for this... During the sequence file, a pier flip took place. Without pointing file, the object was approx. 12 arc min out of center on the other side of the meridian, but perfectly round stars...then SEQ stopped due to focus failure, overcast... Hmm, right now, I have no clue what is going wrong Best, Christian Link to comment Share on other sites More sharing options...
robertp Posted March 31, 2014 Report Share Posted March 31, 2014 Hi Christian, that's very strange. First of all, I'm surprised that doing the MLPT you got a max RA error of only 6 arcsec and a max DE error of only 44,2 arcsec. My max-errors with similar polar misalignment are always way higher (I just checked the last session, max RA error war 9 arcmin, max DE error 2,5 arcmin). For me that would indicate that you have a very stiff setup. Did you measure flex and hysteresis with your setup? I would expect that with such perfect max errors, you should get perfect tracking. I have no clue why using a sequence file at the same region, you suddenly got good tracking. Could you please do the following: When you open Sequence an go to the MLPT-screen, there's a button labeled "Load and Draw LPT File". When you click that, you can choose a directory starting with LPTxxxx and Sequence will redisplay the fitting curve. Could you do that for the two sessions from yesterday and compare the fitting-curves of the MLPT run and the run with the sequence file? I guess you might have stoppped the first run after a while and started the run with the sequence file so that portions of both fitting curves might overlap? Do they look similar? Have you tried to contact Philipp Keller with your problem? Best regards, Robert Link to comment Share on other sites More sharing options...
robertp Posted March 31, 2014 Report Share Posted March 31, 2014 Hi Christian, one addition: you can do what I described above indoors by starting Sequence, clicking Connect and then tell autoslew to start in simulator mode. Robert Link to comment Share on other sites More sharing options...
si6pack Posted March 31, 2014 Report Share Posted March 31, 2014 Thank you Robert for giving feedback - I sent you an email. I attached a Word document where you will find the MLPT screenshots... For my impression, both files would look similar if one would extrapolate the "by hand MLPT" into the future: initially there is nearly no error in the pointings, but the last point starts taking off...this is in a way reflected in the sequence driven MLPT RA error - initially a small error, then it takes of... in both cases, the DEC error comes up to approx 44 arc sec... Maybe a lot of good will in this interpretation, but in sum I would assume, that both MLPT files should have delivered... Christian Link to comment Share on other sites More sharing options...
tbaehnck Posted April 1, 2014 Report Share Posted April 1, 2014 Hi, Then I used SEQ with a sequence file with MLPT enabled for the same object What is the meaning of this? Wich sequence file? Do you record a pointing file with MLPT turned on? BR Thomas Link to comment Share on other sites More sharing options...
si6pack Posted April 1, 2014 Report Share Posted April 1, 2014 In the beta version of "SEQUENCE" (SEQ) you can set up sequence files - automated imaging. You enter your targets and save them as "target files" (.trg). Then you have to define your exposure file, where you have - besides the basic imaging settings (time, filter binning) - the possibility to define whether you want to dither, use guiding, or make use of MLPT. If I remember right you save those as .exp files. Now you have two possibilites: 1) You do the imaging by hand i.e. select your object in your preferred imaging program and do an explicit MLPT file in SEQ (you define the duration, number of shots, etc.); after having finished your MLPT file, you start imaging driven by your preferred software... 2) Using the .trg and .exp files in SEQ simplifies everything - you just load your target file and load the exposure file, press "run now" and off we go...depending on your settings in the exposure file, SEQ will run the sequence file with (w/) or without (w/o) MLPT... Just for clarification: In my first attempt I used possibility 1, good polar alignment, and a "by hand" MLPT (NO pointing file) -> star sripes, shorter stripes, culminating in egg-like shape... In my second attempt, I used possibility 2, same polar alignment but MLPT performed automatically due to thepresets in SEQ (NO pointing file) -> perfect round stars (600sec exposure, before and after meridian flip) I hope this answers your questions... By the way, if I use - after a good polar alignment ( Christian Link to comment Share on other sites More sharing options...
si6pack Posted April 7, 2014 Report Share Posted April 7, 2014 Latest news on MLPT: Yesterday I set up the mount and did a fast 3 star polar alignment on E (no config in use) - after polar alignment, RA error 0.3 arcmin, DEC error 0.5 arcmin Interestingly, the normal message of clearing the pointing file after polar alignment did not pop up. Instead AS showed me that a non-saved config is running (which was the 3 star polar alignment). Therefore I deleted it by hand (clear config)... Then I slewed to M51, centered the object after solving with pinpoint, synched via SEQ and started a MLPT run (100 min, 5 points). In RA I observed an increasing error from 12 arcsec to 2.4 arcmin, in DEC it started with 4 arcsec and ended at 1.5 arcmin. Then I started imaging 2x600sec and 2x900sec - perfect round stars! After that, I set up a sequence in SEQ (2x600 sec and 2x900sec) for M51 including MLPT. After starting this file in SEQ, MLPT run took place (similar read-out compared to above) and imaging started: perfect round stars. However, after meridian flip, M51 was not longer on the chip...but the stars from the neighbouring area kept perfectly round! If possible, I will give it another try today (no change in polar alignment), but now generate a pointing file (12 stars E & 12 stars W) and use it while repeating the MLPT runs from yesterday... Christian Link to comment Share on other sites More sharing options...
tbaehnck Posted April 7, 2014 Report Share Posted April 7, 2014 (edited) Hi Christian, thanks for keeping us informed. Very interessting. I have recognized very similiar also. But one question, when you normally do a PA with three stars, you will be asked if you want to delete the pointing (the window which not appear this time) afterwards. If you say yes, do you have ever looked at "show corrections"? When I have done so, there are still a few numbers in there. When I used "clear and reset config" its all set to zero. Do you know what is the difference in the two ways? BR Thomas Edited April 7, 2014 by tbaehnck Link to comment Share on other sites More sharing options...
si6pack Posted April 8, 2014 Report Share Posted April 8, 2014 Hi Thomas, To answer your questions - no I never looked at "show corrections"...will do next time! And I do not have a clue what the differnece is...sorry for that. However, in AS there is this message around, that AS is using a "non saved config" even if the window pops up and I discard the pointing (3 star polar alignment) file... There is still some testing necessary, in order to fix the issue leading to elongated star using MLPT...I keep you posted. Christian Link to comment Share on other sites More sharing options...
si6pack Posted April 12, 2014 Report Share Posted April 12, 2014 Yesterday I could continue with my examinations on MLPT... Using my wheely bar, I brought the complete setup in front of my garage, performed, after homefind a 3 star polar sequence run on E. WITHOUT doing any further alignment (the mount was 4arcmin RA and 6.2 arcmin DEC off, respectively) I immediately proceeded to MLPT and investigated the behavior within a .seq file and without performing a .seq file on my pet M51. Worthwhile to mention: Before doing a new MLPT, I cleared the orange LPT label in Autoslew by hand and my slew settling time is 10 sec. To cut a long story short: MLPT (RA error from 42arcsec to 3.2arcmin, DEC error from 12arcsec to 56arcsec) within a .seq file - round stars during a 900sec exposure followed by a MPLT without running a .seq file - round stars during a 900sec exposure (similar RA/DEC errors to above were detected by Sequence) During the first run, observatory control in MDL was coupled to "Astrooptik Server Telescope"; I repeated both runs, but now MDL observatory control was linked to "TheSky-controlled telescope" - no issues, perfect round stars, error range as above... Conclusion: MLPT with a roughly polar aligned mount seems to be no issue, no matter how the telescope is controlled in MDL... I even did a MLPT run without having coupled MDL observatory control to any telescope - perfect round stars... However, I was also suffering from elongated stars from time to time - I have no clue why this happens sometimes and sometimes not. Maybe I unconsciously make mistakes during setting up my equipment leading to unforeseen tracking errors...any ideas wellcome! Regards, Christian Link to comment Share on other sites More sharing options...
rharter Posted April 22, 2014 Report Share Posted April 22, 2014 Hi Christian, reading your last sentence this sounds quite familiar to me. I do have a DDM85 on a permanent location. My observations with Sequenz (Ver0.9) and MLPT are almost the same, see my post "Error difference between a manual pointing file and Sequence and MLPT". Somtimes it works, sometimes not and I also do not have a clue what´s going on. So hints for solving this are very appreciated! ;-) kind regards Ralph Link to comment Share on other sites More sharing options...
si6pack Posted April 24, 2014 Report Share Posted April 24, 2014 Thank you Ralph, It is always good to get confirmation from another independent user...from time to time, I think I am the only one having problems Tonight, I was using Sequence for revisiting M106 and friends...I did a rough polar alignment on E, and a 12E &12W pointing file. My sequence file was using MLPT and dithering by 4 pix...then I came across a few problems. I selected an "auto" focus star in the .trg file ( http://forum.astrosysteme.com/index.php?/gallery/image/68-m106-focus-star/ ) - focusing did not take place, supposingly due to the weired coordinates selected by the "Auto Select" button (and confirmed by the error messages of Sequence, see below:) 00:34:01.292: Running now Nr. 1 of this Sequence00:34:04.084: Error when selecting Plate.Catalogstars00:34:04.209: Could not select suitable focus star For a better composition of the image, I did not center on M106, but somewhere inbetween M106 and the neighbouring galaxy, maybe this screewed the focus star selection!? Sequence ignored this issue and started the MLPT run and the imaging session... Everything was fine - http://forum.astrosysteme.com/index.php?/gallery/image/64-1-m106-revisited-2014-04-23-g-0011-scaled/ But with the next image the trouble started: http://forum.astrosysteme.com/index.php?/gallery/image/65-2-m106-revisited-2014-04-23-g-0012-scaled/ Continued: http://forum.astrosysteme.com/index.php?/gallery/image/66-3-m106-revisited-2014-04-23-r-0001-scaled/ http://forum.astrosysteme.com/index.php?/gallery/image/67-4-m106-revisited-2014-04-23-r-0002-scaled/ and finally ended in an error message of my system - AstrooptikServerTelescope is not longer responding... Autoslew was not longer running and disappeared from the tasklist; however Sequence was still running... These are the last messages of Sequence: 04:03:39.109: Time to autofocus now04:03:40.199: Selecting Filter 104:03:45.329: Need new focus star because cannot reach focus star on West side04:03:45.439: Error when selecting Plate.Catalogstars04:03:45.439: Autoselected new Focus Position at 00h00m00,0s +00d00m00s04:03:45.529: This Autofocus Position could not be reached on West side, so I am skipping Autofocus for now04:03:51.529: Slewing to 12h15m27,0 +47°30m19 Pierside West Hmm, what might have happend? I crosschecked with my weatherstation, no high gusts or anything...very strange and very frustrating that - at least I - can NOT 100% rely on automated runs...however, one should not neglect, that Sequece 2.0.3 is a beta version, there are bugs, for sure! BUT: I do not see any attempt to move this soft any further or feed back from the dealers side, despite all efforts of the community... Frustrated regards, Christian Link to comment Share on other sites More sharing options...
tommy_h Posted April 24, 2014 Report Share Posted April 24, 2014 Hi Christian It seems sequence has not found a suitable focus star in the closer area of your target (depending on the mag-limits in the autofocus settings tab). In this case try to find a focus star near your target manually and enter the coordinates in your target file. Otherwise sequence uses the focus star from the target-file for synching at the beginning of a sequence session (if this is turned on in the settings tab). Best Regards Thomas Link to comment Share on other sites More sharing options...
si6pack Posted April 24, 2014 Report Share Posted April 24, 2014 Hi Thomas, Thank you very much fir your feed back and the info - acually, my focus star limits were between mag 4 and mag 8...however, when I crosschecked the settings, the values were set between mag 1 and mag 3!!! No idea how this happend...anyway, I did as you recommended and selected a focus star by hand and corrected the values in the settings... Best Regards, Christian Link to comment Share on other sites More sharing options...
P.Keller Posted May 9, 2014 Report Share Posted May 9, 2014 Hi Christian, although this may seem a lame excuse, the main problem with trailed stars is in 90% of the cases a timeflex problem. MLPT cannot correct a problem where the pier is sinking into the ground while the exposures are made. But you can measure this with timeflex procedure in Sequence. The first reflex after getting elongated stars is to think its the pointing file or some software bug or some MLPT problem but the truth is that in most cases its just bad luck with some time flex problem. We even have this problem on fixed installations, where a complete building is bending by 10 arc seconds due to temperature changes in the night and causing tracking problems. We have not ordered a tip/tilt sensor with A lot of thought in mobile telescope setups should be put in the way the pier/tripod is fixed to the earth. To check wether MLPT is working or not, one way is to make a bad polar alignment and clear the configuration. Link to comment Share on other sites More sharing options...
si6pack Posted May 12, 2014 Report Share Posted May 12, 2014 Hi Philipp, Thank you for your thoughts and suggestions. I tested the timeflex with my setup and it was around 6 to 8 arc sec... However, when I did the testing, the mount was on a tripod without wheelybar...maybe this wheelybar adds to the undesired star elongation, I have to test this hypothesis and will come back... Nevertheless, if you could have a look into this forum's gallery, you will find a few shots where the setup did wired things all of a sudden - something like a "zig-zag" tracking: http://forum.astrosysteme.com/index.php?/gallery/album/10-sequence-autoslew-issues/ Could you please also comment on these? The mount as such is performing very well, Martin from ASA optimized all settings during an online session... Thank you for you support in advance, Christian Link to comment Share on other sites More sharing options...
Recommended Posts
Create an account or sign in to comment
You need to be a member in order to leave a comment
Create an account
Sign up for a new account in our community. It's easy!
Register a new accountSign in
Already have an account? Sign in here.
Sign In Now