I'm new at the whole PMA/PMX idiom and TheSkyX so I am sure I am just overlooking some basic step. But I can't get TheSkyX to plate solve (image link) targets west of the meridian in the automated mode. It happily solves away east of the M, but not west. I can manually slew to individual west targets and they will solve on a 1-at-a-time basis, but not in the automated mode.
Any ideas anyone?
Peter
Peter, have you tried increasing the Bad Pointing Sample Criterion in TheSkyX? What maybe occurring is that post a meridian flip moving to the west, it the tolerance falls outside this value and is considered invalid so does not update the model.
Go to Preferences | Advanced. You'll see it in there. Try changing it to 0.75 or so as a starting point.
You could also add a few points manually from the west and not finalise the model to see if that would work.
Peter, have you tried increasing the Bad Pointing Sample Criterion in TheSkyX? What maybe occurring is that post a meridian flip moving to the west, it the tolerance falls outside this value and is considered invalid so does not update the model.
Go to Preferences | Advanced. You'll see it in there. Try changing it to 0.75 or so as a starting point.
You could also add a few points manually from the west and not finalise the model to see if that would work.
Thanks Jase. I'll flag that to try when I get back. I'm away for the next 3 weeks - this is when I start to wish I had the obs on full remote control - one day perhaps.
Peter
Index out of range usually just means that your raw pointing is so far off that you're nowhere near where you think you are. Way off. Not even close. Plate solves won't work when you're that far off. This is usually only the case prior to your first calibration run right after setting the mount up.
If you're off, but not way the heck off, you can make them work by going into Tools->Preferences->Advanced and increasing the "Bad pointing sample criterion (degrees)" value. This is the setting that controls this "sanity check." Increase it within reason, it's there for a reason, to warn you that you're way the heck off.
Also, any time you set up your mount for an observing session at a new location, after homing the mount and syncing on a star and prior to your first automated calibration run, manually map three or more stars on each side of the meridian. Make sure to map at least one star at a high declination and one at a low declination on each side of the meridian.
Here is my usual polar alignment and calibration routine. This assumes that the finder and scope have already been aligned. I'll gloss over steps like attaching the camera and focusing it and stuff like locking down your mirror if you have an SCT (always lock it down before doing an automated calibration).
1. Consult GPS receiver and make sure computer time is correct and that I have entered and set my correct location. Connect to and home mount.
2. Use TSX to slew to a star near the celestial equator at least 5-degrees away from the meridian.
3. Manually move the whole mount and tripod to center star.
4. Go into Telescope->Tools->Bisque TCS->Utilities and clear the sync history, Go into Telescope->Tpoint Add On->Setup->Tpoint Add On Settings->New and start new Tpoint model, Go into Telescope->Startup and sync on the star, go into Telescope and Add Pointing Sample.
5. Manually slew around to a half dozen stars on each side of the meridian, center them and add some pointing samples.
Now we should have good enough pointing so that plate solves (image links) will succeed.
6. Go into Telescope->Tpoint Add On->Calibration Run and start an automated calibration run of 25 or so points. Let it run. Finish it and click on supermodel and accept the model. Make suggested polar alignment changes. (Note that when I fire up a calibration run, I always have it first determine the image scale and position angle. I also enter the actual humidity and atmospheric pressure. I actually get a little better results this way.)
Now we should also be able to sync on a linked image instead of a star. This is better.
7. Clear sync history, start new Tpoint model, take a picture, image link it, sync on linked image.
8. Fire up an automated calibration run of 25 or so points. Let it run. Finish it and click on Supermodel and accept the model. Make suggested polar alignment changes.
9. Repeat 7 & 8 until I'm happy (usually two more times gets me to sub arc minute polar alignment).
Now we're polar aligned. Let's improve our pointing and tracking by running a really big calibration run.
10. Clear sync history, start new Tpoint model, take picture, image link it, sync on linked image.
11 Fire up an automated calibration run of 180 or more points. Let it run. Go look through other folks scopes for awhile. Finish the calibration run and click on Supermodel and accept the model. Turn on ProTrack.
I hope I got that right. If I've forgotten anything else, I'm sure someone around here will point that out...
thanks for that step by step guide. You should write the manuals. I have not done mine yet but this will make it much easier when the clouds clear.
Only question I have is item 2.
2. Use TSX to slew to a star on the equator about 5-degrees away from the meridian.
Where about is this position roughly ? Meridian is right overhead. Celestial equator ? Dec 0 right ? Do you have an example star you would use at say roughly 8:30 pm ?
I hope I got that right. If I've forgotten anything else, I'm sure someone around here will point that out...
In the nicest possible way, of course.
Thanks Ernie. That's a comprehensive workflow and I will definitely run that through the next opportunity I get.
Peter
Where about is this position roughly ? Meridian is right overhead. Celestial equator ? Do you have an example star you would use at say roughly 8:30 pm ?
The Zenith is the point straight up directly overhead.
The Meridian is a line that runs from due North, up and over your head through the Zenith and then down through the pole to due South.
If the celestial equator runs through your Zenith, then you must live somewhere on the Earth's equator.
Image link and sync on linked image is better than syncing on a star. But you usually have to do that very first sync on a star because when you first set up, oops, raw pointing isn't usually good enough yet for image links to work.
In general, avoid synchronizing on a star that is located too near the meridian. Do not synchronize on stars above 60 degrees declination either. Where "too near the meridian" is a relative term. I stay at least five degrees or so away from it when syncing. But then I have an ME. I'm not at all sure if this is 100% applicable to the MX with it's larger zones around the meridian.
Syncing invalidates your Tpoint model. If you're used to other mounts where you sync anytime you want to improve local pointing, now is the time to get out of that habit. If you have third party software that is set to do (real physical) syncs, turn that option off. Differential pointing with a plate solve and a jog is okay, but not syncs.
Also, in my experience ProTrack works very well only after you've mapped a lot of points and have a good model. If you've only mapped a small number of points, then my experience is that it can make things worse. In my experience it can take over 80 mapped points all over the sky before ProTrack goes from making things worse to making them better and it can take upwards of 150 points or so before it really starts to do wonders for tracking. YMMV. Fortunately automated calibration makes it very easy to quickly acquire a lot of mapped points. (You need to collect them all in the same session and same night. Can't just add more the next night or anything like that.)
I highly recommend reading the Paramount MX manual several times, then several times more. It's the best mount manual Software Bisque has ever produced. It has many years of experience distilled into it.
Also Ernie how long does this routine usually take you?
About 45 minutes to an hour for the set up and polar alignment tweaking part. Another 45 minutes to an hour for the long mapping run. If that seems long, sorry, I'm being brutally honest here. And that long mapping run is worth it to have the center of every object fall dead center on your CCD (or within only a handful of arc seconds thereof). A real time saver later. And ProTrack! What a wonderful feature that is! The smile ProTrack puts on my face is priceless. But it needs that long calibration run with all those mapped points to work so beautifully. Because of this set up overhead, when I set up, I like to stay on station for at least three nights. One of the reasons it takes me so long is that my camera is slow and TSX does one thing at a time. It does not overlap slewing with downloads and plate solves. It would be really good if more people than just me would request SB to enhance the automated calibration run process to have more overlap. After all, the program is multithreaded and could do the image acquisition in one thread, overlap the download with the following slew, and fire off a separate thread for each plate solve. If they did this, calibration runs would go much quicker. No big deal for those of us who keep our mounts in observatories, but the MX is a portable mount and so MX owners would be expected to be going through this routine painfully often. Making it faster is in their best interest...
By the way, you can and should bin the maximum amount during automated calibration runs and it will NOT hurt your calibration. I bin 4x4 and set my exposure between 10 and 15 seconds. That speeds things up and doesn't hurt anything.
About 45 minutes to an hour for the set up and polar alignment tweaking part. Another 45 minutes to an hour for the long mapping run. If that seems long, sorry, I'm being brutally honest here. And that long mapping run is worth it to have the center of every object fall dead center on your CCD (or within only a handful of arc seconds thereof). A real time saver later. And ProTrack! What a wonderful feature that is! The smile ProTrack puts on my face is priceless. But it needs that long calibration run with all those mapped points to work so beautifully. Because of this set up overhead, when I set up, I like to stay on station for at least three nights. One of the reasons it takes me so long is that my camera is slow and TSX does one thing at a time. It does not overlap slewing with downloads and plate solves. It would be really good if more people than just me would request SB to enhance the automated calibration run process to have more overlap. After all, the program is multithreaded and could do the image acquisition in one thread, overlap the download with the following slew, and fire off a separate thread for each plate solve. If they did this, calibration runs would go much quicker. No big deal for those of us who keep our mounts in observatories, but the MX is a portable mount and so MX owners would be expected to be going through this routine painfully often. Making it faster is in their best interest...
By the way, you can and should bin the maximum amount during automated calibration runs and it will NOT hurt your calibration. I bin 4x4 and set my exposure between 10 and 15 seconds. That speeds things up and doesn't hurt anything.
Hi Ernie,
How are you using Protrack? Do you image unguided?
How are you using Protrack? Do you image unguided? Or can you use it if you are still autoguiding?
You can use it while autoguiding. ProTrack does not interfere with autoguiding. It extends the amount of time you can image unguided to the point where you don't need to guide at all for short subs. Note that if you're using CCDsoft, you can use DirectGuide mode to eliminate the guide cable.
You can use it while autoguiding. ProTrack does not interfere with autoguiding. It extends the amount of time you can image unguided to the point where you don't need to guide at all for short subs. Note that if you're using CCDsoft, you can use DirectGuide mode to eliminate the guide cable.
That's good to know. I'll try that out. I have used it without guiding and it did seem to help.
One other question. After initial pointing run to get polar alignment the T point model makes reccomendations for the PMX like
Losen West , tighten East Azimuth knowb by 12 arc minutes etc...
Or Raise polar axis by 20 minutes.
When we do this from southern hemishphere I am asuming we just follow the instructions and dont have to do anything funky like tightening east if it says west. Or lowering axis if it says to raise it ?
Can someone just 100% confirm this so I dont drive myself crazy.
Did I mention I love the Joystick on the MX. makes centering stars so smooth and easy.
One other question. After initial pointing run to get polar alignment the T point model makes reccomendations for the PMX like
Losen West , tighten East Azimuth knowb by 12 arc minutes etc...
Or Raise polar axis by 20 minutes.
When we do this from southern hemishphere I am asuming we just follow the instructions and dont have to do anything funky like tightening east if it says west. Or lowering axis if it says to raise it ?
If your latitude and local time is set correctly and accurately, then all else being equal,
the advice it provides will be consistent with you being in the Southern Hemisphere.
TPOINT was originally developed in Australia for the 3.8m Anglo Australian
Telescope.
Yes that is what I have found the instructions on what to do are what you actually do and you don't correct for being in the Southern Hemisphere.
But it also usually states how many ticks of the mount to move Each knob on the adjustment metal knob is 1 tick. It makes it very precise.
Move east 3 ticks is turning the knob 3 of those raised knuckles on the adjustment knob. An awesome feature.
Got T-Point add on in SKy X now working and doing fully automated T-Point runs.
Works great. Took a little fiddling with Plate solve parameters but now I am binning 2x2 with 5 sec exposures and solving everytime. Takes 12 seconds to move solve and move between targets. Very cool. Have large run happening as I type this. Whats a decent pointing accuracy to aim for ?
It pains me to write this, but I've seen people achieve better than four (4) arc-seconds RMS all sky on a Paramount ME. That's with a large refractor as their OTA. Alas, I can only dream of being able to accomplish that.
The very best I've ever been able to do so far with my EdgeHD 14 on a Paramount ME is between 8 and 9 arc seconds RMS all sky. Despite having the mirror locked down, mirror movement is no doubt still my limiting factor.
Once you've finished your run and had Supermodel compute and accept your model, please go into Tpoint, select the Model tab, select "Multiple Graphs" from the Graphs pulldown on the lower left corner, drag the edges of the window to make the graphs as large as they'll get, make a screen snapshot using the Windows snipping tool accessory (if you're running Windows) and post it here. Thanks.
Oh, if you haven't tried going into Telescope->Tools->Bisque TCS->Parameters and increasing the mount's acceleration and slew speed to their maximum values, give it a whirl. Very cool.
Ok here it is. This is after a 50 point automated run. So not much data. I also noticed after this my polar alignmenet was showing I needed to adjust 3min W and lift 2 min. So once this is done I will do a 100 or so point model and see if it gets any better.
Still pretty happy with 11.1 arc sec. I probably need to also double check all my dovetails, focus locks, etc.. and make sure everything is nice and tight.
I have no idea what all these graphs mean other than scatter plot. Your step by step guide below was great by the way so thanks.
I am very pleased how quick and easy it was to get plate solve/image link working. Did not need to download any additional catalogues etc... it just worked using the Sky X, Camera Add In and T-Point Add in which all come with PMX
