Lodestar/MaxIm Guide Fade Fade Reprise 3

[Ronald Crouch]

Hello All,

Posting this on Maxim too. The problem: Super bright guide stars fail to be detected by Maxim.

OK. Hardware.. Lodestar 1 shot binned 2x2 as guider. Full calibration fresh darks.. In ACP a guider S/N of 12 was requested. I moved that up from 6 after guiding simply failed totally. A typical guide star box image has noise of 30 and the star at 23000 counts. SNR in Maxim aperture says > 90/1. Problem.. guide star fade over and over. If Maxim registers a "fade" during actual acquisition not a big problem cause tracking is superb.. My Mach 1 is < 1 arc-sec PE with the 120mm APO.. If fade prevents guiding from ever starting it can be a small problem for 180 sec subs but a big problem for 900 sec subs. The attached zip file is a directory which contains the ACP log, MaximLog, Tracking Log and another folder with all the guide star images over the same period which was about 3:56 til 4:56 UT.

My inspection of guide star images suggests they are undersampled being binned 2x2 which is about 2.1 arc-sec/pixel. Main OTA is at 1.5 or 1.75 arc-sec/pixel depending on reducer. It is a crazy sharp 120mm APO with the lodestar on the guide port of a QSI583wsg8. Guide chip/path look fairly free of dust and smudges.

Basically this long standing Maxim guide star fade problem means I cannot reliably use Maxim 6.13 which means I cannot use ACP or Focusmax 4. I can go back to Maxim 4.61 / Focusmax 3 and CCDA which does not have the problem. TheSkX/CCDA is also an option. Both TSX and PHD will guide forever and never fail to see the guide star. It seems a pity to have invested so much money in 3 software packages which have great promise but can be non-productive because a bright guide star cannot be be seen by one of the packages. I LOVE ACP and hate to have to give it up . Really great clear nights are precious and I simply cannot lose subs with a great setup for such a reason.

Any suggestions appreciated. Does anyone see something to try other than not binning the guider but I know others have tried that to no avail already. I sure wish ACP for just image production without an observatory could escape Maxim 6. Just sayin..

Best,
Ron

[Ronald Crouch]

I have from the beginning shot the lodestar binned 2x2 because it is cleaner that way and with, say the SW120 F/7 the binned resolution is 2.1 arc-sec per pixel. The 120 is so crazy sharp that a typical guide star might only be 4 pixels which is I suspect why Maxim fails to recognize it. Likely trying to do a good thing and avoid hot-pixels.

So today I decided to move to unbinned lodestar and in Maxim selected camera 2 and shot a sequence of 31 bias and 31 10sec darks. i took one of the darks and made a rejection map with hot pixels threshold at 1700. In calibration setup i associated that map with the darks and made master bias/dark.

The attachment shows unbinned lodestar 3 sec exposures with or without bias/dark calibration. Some of those hot pixels are 65000counts! and the calibration with hot-pixel map really cleans it up great. It's a 5 min sequence to repeat this library. Hopefully temperature is not crazy critical. Hopefully thesesteps will solve the guide star faded unless the sky goes bad. We will see. AM I thinking clearly?

[Steve Brady]

Hi Ron--

A quick note to let you know that FocusMax does not relie on using MaxIm or TSX for star detection so there is no need to fall back to FocusMax V3....

Clear skies,
Steve Brady

[Ronald Crouch]

Ah yes I know and that is good.. My reference to fall back to 3.4 was in the case I moved back to Maxim 4.61 which guides fine with my hardware but does not work with FM4.

Best,
Ron

[Bob Denny]

I can't do anything in ACP for this. It seems to be limited to some models or some subset of Lodestars. The killer is that it will guide OK in PHD etc. I am interested to hear how your experiment works out. I have moved this to the "other stuff" section because it isn't being (directly) caused by ACP.

[Ronald Crouch]

Thanks Bob! Just to be complete I post here what i said in reply to a user with a STT-camera with the same problem on the Maxim forum!! It is terribly sad because Maxim's failure to be interested in supporting their customers will force me to abandon ACP. It is becoming the point where how can I in good faith do business with Maxim if they do not even respond to pleas for help? Maxim 5 (early) and Maxim 4.61 do not have the issue. So we pay to support our vendors by upgrading to 6 and it effectively kills the entire process. It is the silence and lack of interest which hurts. It is interesting that both CCDAutopilot and SGP can use PHD2. Apparently it has all the hooks and more to support ACP and is very well structured. My selfish side suggests that ACP could offer that as well with a script.

Anyway I appreciate everything you do!! Oh well.. here is my reply in Maxim just to keep things complete..

---

Yes I am using PEM. Without it my Mach one is a very smooth ~+- 3 arc-sec. A couple of weeks ago on a night of great seeing I redid it with PEMPRO and the results are a solid <1 arc-sec. One course my pixel scale was ~1.5 so it's hard to pin the PE down exactly other than to say it is very very good.
The mod of connection to the mount does not matter.. movement is precise with a cable or ASCOM. Return from dither is fast as well. The additional of a hot pixel map to my guider darks (full calibration from 31 bias and 31 10sec darks) yields a very very clean unbinned guide exposure. I guide at 0.5X and that puts the Maxim guide rates around 3.2. When the guide does not fade the corrections are very tiny or nothing at all. Now with ~900 sec subs if the guiding goes away I am hosed. The bottom line is that is if my new approach of unbinned guider + super clean processing does not solve the problem and if Maxim does not engage to help I have no recourse but to leave the wonderful ACP because it is tied to Maxim. CCDAutopilot now supports PHD2 which does not have this guiding bug. Also Sequence Generator Pro looks interesting, connects to all my gear and might do the job as well.


The point is that Maxim 6 has a problem. Guide frames were requested along with Logs and I sent them. Hopefully someone will look and answer. I can go back to Maxim 4.61 which does not have the problem but once again by doing so I lose both ACP and Focusmax 4. It is terribly terribly sad that such a basic problem continues. I see you use a STT camera so the fact the the software is provided by the same company and you also see the problem is quite interesting.

---

[Ronald Crouch]

Hi Bob!
Regarding guide star S/N parameter.. How can we better understand this one :-). i found in recording guide frames (a zip on Saturday on this thread i think) that in ACP i set at 12... yielded a guide exposure of ~5.6. When I look at those frames in Maxim aperture reports ~90/1! Er those "90/1" stars are said to have "faded". The track log also in the zip said the level was about 40,000!!!! I would never go so strong.. About 1/2 to 1/3rd maybe.. Sheesh. Too faint to see gee wiz. So far no response to the info and data in that zip which is another issue. Not only do PHD2 and Maxim 4.6 work fine, TSX Camera add-on does too!! How might we commission an ACP PHD2 guide script LOL.. PHD2 is an awesome thing now compared to what I used 10 years ago. It just works.
Best,
Ron

[Bob Denny]

Ron, I am doing what I can to urge Diffraction to solve the mysterious Lodestar guiding problems. It's not ACP's actions to set the exposure length. For virtually all guide cameras the 3.0 SNR is fine. The behavior of the Lodestar makes ACPs actions seem crazy and you're looking for answers in the wrong place. I usually try to refrain from casting a bad light on other software, but in this case the guiding sensing in MaxIm is somehow choking on the images coming from the guide camera, or at least it seems so...

I'm trying to get this worked out with Diffraction.

PS the STT problem I think you're referring to appears to be a broken shutter. If that's not it then I apologize. I can't keep up with all of the traffic on the MaxIm official support forum...

[Ronald Crouch]

I agree totally! The problem is most definitely not with ACP. Here tonight looks pretty good so I'll be able to get another data point with the guide images being calibrated with scaled darks and a pixel rejection map. Wow it is clean. The other part is not binning so the guider will be pretty much at typical sky resolution..overkill.. 2.1 arcsec.. Not under sampled. The thing is that the lodestar has absolutely no banding and a fft is clean too. My suspicion is the super sharp optics are messing with Maxims algorithm. The frustrating part is no information back from them so we cannot do thing to help cooperatively to try and solve the issue.

Thanks for all you do!
Ron


Sent from my iPad using Tapatalk

[Ronald Crouch]

Hi John,

When I look at what PHD2 did to allow automation hooks in CCDAutopilot it is very simple. PHD2 and PHD2Broker are running and they know all about one another. There are simple BAT files in PHD2Broker which by the name is the go-between PHD2 and the rest of the world. Those BAT files are simple two line switches really to begin_slew, end_slew, dither, guiding_on, guiding_off, pause, resume, etc.. i look at acquiresupport and the AutoGuide() routine and that section could use those functions leaving all else alone. To avoid never-ending code entropy (I love your code!! clearly commented) acquireimages I would try and not touch. Were I successful LOL I would have an acquire support which would only work with PHD2 of course. Might be fun to try actually because in the end PHD2 is well-honed software built to simply guide and help analyze guiding behavior nothing more really. I do not expect or see it as your task to look into this solution of course. I might ask for advice though!

You might guess that I am not expecting success with my lodestar tonight even with utterly perfect unbinned guide frames.. Others have been there before me :-). We will see.. If guiding fails I just will have to go to a session solution which does not use Maxim for guiding until some action fixes the problems. Tonight i supposed to be clear which is a good thing!

Best,
Ron