Jump to content

Banner.jpg.b89429c566825f6ab32bcafbada449c9.jpg

Help with image creep.


Seanelly

Recommended Posts

Look forward to seeing the before and after images.

So after centering the third alignment star, it continues drifting across the dslr screen.

Is that in dec or ra, is it always in that direction, how long does it drift for ?

Are the stars in your images drifting in the same direction, or is it different in each image ?

Michael 

Link to comment
Share on other sites

20 hours ago, michael8554 said:

Look forward to seeing the before and after images.

So after centering the third alignment star, it continues drifting across the dslr screen.

Is that in dec or ra, is it always in that direction, how long does it drift for ?

Are the stars in your images drifting in the same direction, or is it different in each image ?

Michael 

 

So, if you've got time, I've got answers and I've got questions.

I can't provide images of the star drift during those four imaging sessions in question because while checking over each sub to potentially go into DSS I deleted all the bad ones. I didn't realize that I might need to reference them, and though I have plenty of room on my portable drives for storage, I didn't see any use in holding onto junk subs. I do have the good subs of M3 that I subsequently stacked, processed and posted in my album, 3+ hours out of about twelve imaged.

After centering the third alignment star, then exchanging the eyepiece/diagonal for the DSLR and positioning the star in the DSLR 10x box for focusing, the star continued to drift for a short period of time in the direction it came from after I took my finger off the handset directional control. This was under 10x, don't forget, so the drift was not much, but it was there nonetheless, and on the last disastrous night of the four sessions, after the star drifted and I re-centered it in the DSLR 10x and achieved focus and then slewed to M3 and framed the object, I turned my attention to setting up PHD2, and when I turned back to the DSLR  live-view screen a few minutes later, M3 had drifted halfway up the (unmagnified now, so a significant drift) screen of the DSLR. Seeing as how the image had already stopped moving and was still placed as such that my cropping would make it unnecessary to centre the image anyway, I left it that way and started imaging. I lost all subs on that last night.

The star trails were not long, but these subs were only two minutes duration, and I've gotten much better results just guiding off the mount alone. All trails are in the same direction, what I believe to be declination, as I was imaging toward the east and the trails were running to the top right of the subs (I hope that is not misleading).

Friday night I got my first chance to set up and test since that last disastrous session. I checked over all the gear, making sure everything was as it should be, and took great care in balance and locking everything down securely, etc. There was no drift of the third alignment star in the DSLR 10x live-view screen, so that was already an improvement, and after the mount slewed to M51, I framed it and then went through the Guiding Assistant routine.

I asked this question earlier in the thread but got no answer, and I've yet to get into the PHD2 analysis link provided: Is it recommended to use Guide Assist every time you set up, or perhaps only every time you want to image a new object, or just randomly on occasion? This is more relevant to me now because my guiding last night was better than I've ever seen it. There were occasional blips, but generally speaking it was exceptional.

I have another question that might be solved from the guide log provided below, but as stated, I have not yet looked at the analysis link so I do not know. While my guiding was very good according to the graph, I noticed that my polar alignment error in the Guide Assist routine ended up at around 8.0, whereas the first log I provided for the bad sessions earlier proved my PA alignment error at only 0.5. If I'm reading and understanding all this correctly and the discrepancy is truly an indication of bad PA, then it may be because while the handset provided Polaris position at 1:30, I just may have set the mount to 1:20 instead. It was just one of those things where after everything is done and you have started imaging, you run over the routine in your head to try and find any flaws, and this PA 1:20/1:30 stuck in my head all evening. There was no way I was going to start all over again, and I figured if PA was bad, it would still be possible to image while taking that into account, and so my question is, if all this be so, would guiding still be as good as it was last night?

And now for the results, good, bad, and just plain embarrassing. I imaged about ninety minutes each of M51, M3 and finally 2 hours M57 as a late-night addition, three different locations in the sky for a good overall test, and even got up before dawn to grab darks, flats and bias. But stupidly, in my routine to organize the folders before stacking, I somehow deleted the subs for M3 and M57 (I need to get more sleep, haha). Go ahead, hit me now and get it over with-tell you what, I'll do it myself right now. Ouch. What gets me in all that is I'm usually so careful anyway in my setup, and yet here I was at the most important of imaging sessions fretting over possibly clumsy PA and then dumping valuable images not even in the recycle bin where I could retrieve them and save the day. Anyway, I still had M51.

I lost about 10% of the subs to star trails or 'globbing', if I can use that word, but the types of losses were nothing like the consistent one-track trails of the bad M3 subs in question, and I'm hoping that this high loss can be attributed to the fact that while the mount RA was nearly parallel to the ground and so my balance there would have been spot on, the scope itself was nearly vertical for the roughly 90 minutes of imaging, and while my balance there was as good as I could make it, I'm wondering if this vertical attitude had anything to do with the high rate of loss.

Sorry for the long-winded explanations. I post here the 1h 4m M51 image, composed of 3m subs and about 50 each of the three other calibration frames in DSS and tweaked in PS, and also the guide log.

If the problem has cleared up, which won't really be known until the next session, fine and dandy, but it would still leave me wondering what caused it in the first place for all four of the imaging sessions of a new object.

 

PHD2_GuideLog_2019-05-17_215239.txt

 

1746467540_Autosave-1cropand16bit2ndtry2.thumb.jpg.c8a155d11cc7ad04c6be4a61ecc5104c.jpg

Link to comment
Share on other sites

A lot to digest there Sean, but here goes:

I don't have an explanation for the short period of drift after releasing any of the slew buttons - unless you have Backlash Compensation enabled on the mount, as mentioned before ?

Trailing in Dec on your images when PA is good could be due to Dec Backlash, which can be minimised with one-direction PHD2 Dec guiding.

Guiding Assistant gives a confirmation of your PA, and gives recommendations for PHD2 settings based on the Seeing for that session. In your Friday run it shows Dec heading off south due to your 8.1arcmin PA, that your RA PE has peaks of about 20arcsecs. Dec Backlash is 5,839 milliseconds, so the recommendation is Dec guiding in one direction. Also suggests a 1 second exposure, you used 3 seconds.

As you gain confidence you will use GA less frequently.

Guiding in one direction entails:
Setting PA to be no better than about 4 arcmins, so that the mount gently drifts in one direction.
Balance the scope to be slightly heavy in the opposite direction, which takes up the backlash.
PHD2 will correct the gentle PA drift, and if the scope balance is right, it should not go out of balance in the opposite direction, which could take 5.839 seconds to correct !

Friday night, no drift after releasing slew buttons - let's hope you are now good to go for future sessions 😆

You deleted the M3 and M57 images and emptied the Recycle Bin ? I've used my Huawei phone to hack a North Korean nuke missile test, it's heading your way.......

Your M51 image, composed of your good subs, is as good as I get in terms of star shape, but a jpeg of one of your 10% of poorer 3 minute subs, stretched if necessary, would have been more informative. 

Now to your guiding.
There are many Dec excusions of up to +/- 5 arcsecs, which will give you Dec drift on some of your 3 minute subs. Try Dec guiding in one direction only ?
I notice you are ST4 guiding. PHD2 can compensate for the different guide corrections needed for high and low Dec targets, if you have cabled to give RA and Dec positions to PHD2. 
You haven't, so in your case PHD2 recommends you do a Calibration for every different target, so that PHD2 can give correctly sized guide corrections for each target Dec.
I can only see one Calibration, despite targeting three different objects at 47 Dec, 28 Dec, and 33 Dec ?

Now a word about your guidescope.
Your imaging scale is about 1 arcsec/pixel I would guess, and your guide image scale is 6.62 arcsecs/pixel, according to your PHD2 settings.
That means if your guidecam shifted only 1 pixel (5.2um) due to a less than optimum mount, your images would move nearly 2 pixels on your T6i.                                                                                                                                                                       Is your guidescope mounted rigidly enough not to move small fraction of the thickness of a hair under gravity as the mount tracks to different positions ? When you have experienced unwanted moves in you imaging, it's worth removing a possible
source.  Hence the concerns of some of us !

Michael

Link to comment
Share on other sites

16 hours ago, michael8554 said:

A lot to digest there Sean, but here goes:

I don't have an explanation for the short period of drift after releasing any of the slew buttons - unless you have Backlash Compensation enabled on the mount, as mentioned before ?

Trailing in Dec on your images when PA is good could be due to Dec Backlash, which can be minimised with one-direction PHD2 Dec guiding.

Guiding Assistant gives a confirmation of your PA, and gives recommendations for PHD2 settings based on the Seeing for that session. In your Friday run it shows Dec heading off south due to your 8.1arcmin PA, that your RA PE has peaks of about 20arcsecs. Dec Backlash is 5,839 milliseconds, so the recommendation is Dec guiding in one direction. Also suggests a 1 second exposure, you used 3 seconds.

As you gain confidence you will use GA less frequently.

Guiding in one direction entails:
Setting PA to be no better than about 4 arcmins, so that the mount gently drifts in one direction.
Balance the scope to be slightly heavy in the opposite direction, which takes up the backlash.
PHD2 will correct the gentle PA drift, and if the scope balance is right, it should not go out of balance in the opposite direction, which could take 5.839 seconds to correct !

Friday night, no drift after releasing slew buttons - let's hope you are now good to go for future sessions 😆

You deleted the M3 and M57 images and emptied the Recycle Bin ? I've used my Huawei phone to hack a North Korean nuke missile test, it's heading your way.......

Your M51 image, composed of your good subs, is as good as I get in terms of star shape, but a jpeg of one of your 10% of poorer 3 minute subs, stretched if necessary, would have been more informative. 

Now to your guiding.
There are many Dec excusions of up to +/- 5 arcsecs, which will give you Dec drift on some of your 3 minute subs. Try Dec guiding in one direction only ?
I notice you are ST4 guiding. PHD2 can compensate for the different guide corrections needed for high and low Dec targets, if you have cabled to give RA and Dec positions to PHD2. 
You haven't, so in your case PHD2 recommends you do a Calibration for every different target, so that PHD2 can give correctly sized guide corrections for each target Dec.
I can only see one Calibration, despite targeting three different objects at 47 Dec, 28 Dec, and 33 Dec ?

Now a word about your guidescope.
Your imaging scale is about 1 arcsec/pixel I would guess, and your guide image scale is 6.62 arcsecs/pixel, according to your PHD2 settings.
That means if your guidecam shifted only 1 pixel (5.2um) due to a less than optimum mount, your images would move nearly 2 pixels on your T6i.                                                                                                                                                                       Is your guidescope mounted rigidly enough not to move small fraction of the thickness of a hair under gravity as the mount tracks to different positions ? When you have experienced unwanted moves in you imaging, it's worth removing a possible
source.  Hence the concerns of some of us !

Michael

The recent Friday session I finally managed to fit in that showed no drift after releasing the directional buttons I think tells me that previously either something was not battened down or my balance was off, but the fact it happened four times in a row while getting progressively worse and now it seems fine, even with the extra caution Friday given to the setup, just has me baffled. If I never see it again I'll be happy, but it will bug me for a long time to come.

I will have to check on the settings for backlash hopefully Tuesday night with a bit of lucky clear weather. All these settings (backlash,etc.) are new to me, as well as the Guide Assistant recommendations, which I never noticed after it finished it's routine. I just assumed it would take care of those by itself (pretty sure you knew who you were dealing with here from the start but this is more proof how new I am at to this).

As to that silly business of deleting M3 and M57 subs, I never emptied the recycle bin, did not even send the subs there, I just looked there for them but they were not to be found, nor anywhere else. I can't even recall how I lost them. I thought I had separated everything into folders but after checking each sub of M51 I moved on to check M3 and M57 but the folders had vanished. Maybe I should have been a magician!

As to that nuke crack, why are you using Chinese Huawei technology after all the [removed word] they've stirred up, both the country and the company?  And you can't fool me, any N. Korean missile fired in this direction, if it ever got off the launch pad, would no doubt land closer to Japan than Canada.

Re guiding: My experience with noticeably improved guiding Friday night after using Guide Assist was an eye opener. I did not know how necessary it was in my case, and I will use it for each new target in future.

Guidescope: You will have noticed the SW scope I am using for imaging. I don't know if you've seen it, but the metal wedge mount for the guidescope is dovetailed to fit the metal Orion guidescope base and secured by a single, strong screw. I went over it thoroughly on Friday during setup and it is very solid. I can't vouch for materials or design used in these two areas in the past which may possibly be the source of the questionable reputation, but I work as an industrial millwright, and while the technical side is not my strong suit (you will have noticed, haha), I really do know a strong, mechanical connection when I see one, so while I don't rule it out entirely as as contributing to my problem, it is the least of my worries. If my drifting had occurred on just one night, or on a couple of separate nights, I might perhaps suspect that the dew heater I use on the guidescope had caught up on something and pulled on the guidscope, but four sessions in a row blows that up.

Tuesday night (hopefully) I will follow all the Guide Assist recommendations and briefly test-image several more widely separated objects, and pass on any bad sub examples and the log. I hate to say it for your sake but I hope all my troubles over those four nights were an anomaly that I'll never see again and I've just wasted your time. If that is the case, hopefully the compensation on your part will be that I know a fair bit more about what I'm doing than I did before all this happened, which will no doubt improve my images, and you've shown yourself to be very selfless with your time.

But first things first.

Link to comment
Share on other sites

Hi Sean

My Hauwei phone ? No way Jose, of course I haven't got one, for obvious reasons 🙂. English humour at its worst.

Unfortunately the North Korean missiles seem quite reliable. 

GA is a good starting point each evening, unticking the Check Backlash box once it's started will save a lot of time since you already know it's bad on your mount - down the road you might want to try adjusting your Dec worm ?

You can set your mount's Backlash setting to zero during the day, look in your instruction manual first.

Your missing files? Try the Windows search routine. Search for files created on the date in question, and for file types enter a * before .CR2 (eg  *.CR2 ) which will find all CR2 files created that day, if CR2 is the format you saved them in.

Not so much how your guidescope is attached to the imaging ota, more the way the guidescope is held in its mount with two screws, which if nylon or soft tipped, are going to move those microns I mentioned.

My own finderscope is mounted the same way as your guidescope, and needs realigning every session, it's that imprecise.

But it was just one item to tick off the list, if you find down the road that you're getting good guiding across the sky, then no problem.

Michael 

 

Edited by michael8554
Link to comment
Share on other sites

I will be using the GA for each new object at least for the immediate future, and will see to the backlash setting.

I just broke away a few minutes to check the handset backlash settings for the mount and both Dec and Alt are set at 0 value, which is the disable mode (manual).

What is your opinion of belt drive conversion for the HEQ5? I've read here only positive results, that it can reduce backlash, etc., and though I'm in no hurry to explain another astro-related purchase to my wife, it surely can't be that expensive, and I'm confident I could do it myself, as I deal with relatively similar stuff-albeit usually on a larger scale-on a daily basis. If it can help improve my images, I would seriously consider it.

Sorry to leave you hanging again for the last ten minutes (haha), I just tried to recover the lost files as you suggested and also went deeper for online strategy but they are nowhere to be found. I might be persuaded to believe in a mislabeling issue and they are hanging around somewhere if I were not sure that I did not because of their relevance to my situation, also because mislabeling two files at the same time is a boner I just can't believe I'd make, besides, the date search has revealed nothing. Quite frankly I now feel that if they are so intent on hiding from me they can bloody well stay there, I will know soon enough how my imaging is faring without them. (That said, my preliminary curious look at a couple of the subs of M57 (no star trails!) before the folder mishap were fascinating (great colour!), and I can't wait to get in a solid night imaging this ring nebula.)

As for the guidescope issue: I see that the dovetail (I called it a wedge earlier) foot of the guidescope bracket can be viewed as quite inflexible in the main scope base, especially as in my case there is no extra shoe or adapter added that might increase the possibility of unwanted movement, but the guidescope tube in its bracket is less so, as the three alignment screws (you mentioned two) are not steel as I said in error earlier, but nylon. If I were having these drifting issues constantly I might be inclined to see this setup as more of a culprit, but at any rate I will keep an eye on things in this area.

I hope you can bear with me a little further come Wednesday (weather permitting, which looks good for Tuesday night) and hopefully celebrate a little on helping a fellow skywatcher get back on track-actually get onto a better track.

Cheers, Sean.

 

 

Link to comment
Share on other sites

Hello, Michael. If you could indulge me once more (a figure of speech, perhaps I mean once or twice more), I have the results from Tuesday night's imaging session, and results are mixed.

I was very conscious of my setup routine, tension, balance, cable placement, etc., but while centering the third star (Vega) of alignment in the DSLR 10x live-view box for B-mask focusing, I noticed it was drifting briefly uncommanded, not out of the box, so the drift was not much, but some drift nonetheless, and I had to recenter a couple of times before it settled down. Then after focusing and re-placing the star in the unmagnified center of the DSLR where it was originally after 3-star align, I selected M51 and after framing this galaxy with a couple of 30s exposures I turned my attention to activating GA, and when I returned to the DSLR (2 minutes?) to check positioning of M51 with another 30s exposure, M51 had drifted up the DSLR live-view screen exactly as it had done before! Not all the way out, but a significant drift considering that the DSLR live-view was unmagnified, and far enough that I had no choice but to re-center it, which I did with a couple of short exposures, and then moved it around a few times to see if it would drift again but it seemed stable this time, so I went back to PHD2 to reset.

Here is the last short (30s) exposure taken to frame M51 in the DSLR live-view. Magnified, you can see that the image shows star trails (significantly, in the same direction as all the bad subs in the earlier sessions that started all this trouble) over the 30 second exposure, which I only discovered later when studying the exposures:

IMG_5248.CR2

 

And here is the 30s exposure after activating GA where I discovered that the image had drifted in the DSLR live-view along the lines of the star trails in the image above:

IMG_5251.CR2

 

All that aside, once GA gave me the go-ahead and I set the recommendations it provided, I started 3 minute exposures. I could not find calibration setting in GA and only discovered last night that it lies in the 'brain' section of PHD2. This I will rectify next session, but it remained as-is for the two objects I imaged Tuesday night.

Results from 26x3m M51subs were varied. 9 were very good, 10 showed very slight star trails under heavy cropping but adequate for wide-field look, and 7 showed significant trails, not as bad as they were in the previous sessions where all the trouble started, but pretty much useless nonetheless. Following are an example of each.

Good 3m M51 sub:

IMG_5269.CR2

 

Mediocre sub:

IMG_5275.CR2

 

Bad sub:

IMG_5289.CR2

 

Keep in mind that even the bad subs show star trails significantly shorter than the trails that first brought all this to mine and your attention a couple of weeks ago, which were perhaps 3 times longer, and I'm wondering if it's possible that the 7/26 bad M51 subs from Tuesday might be attributed to the scope orientation, which was nearly vertical during the 90 or so minutes of exposing?

I wonder this because after M51 I switched to M57 and my percentage of bad subs dropped from 7/26 for M51, to 14/69 for M57, though a loss that is still far more than I was getting before all this trouble started.

I imaged 69 3m exposures of M57. 42 were good, 13 mediocre, and the 14 bad ones just mentioned. Following are examples of each.

Good 3m M57 sub:

IMG_5297.CR2

 

Mediocre sub:

IMG_5305.CR2

 

Bad sub:

IMG_5309.CR2

 

Again keep in mind that these short star trails are nowhere near as bad as the ones that brought all this to our attention in the first place, so something has changed for the better. That said, I remind you that my sub losses before all this were minimal, perhaps 2-3% at most. I thought that maybe all this scrutiny of my subs was making me too critical of their appearance, but I checked a bunch of my subs from six sessions a couple of months ago comprising 17 hours total of 5 minute exposures on M81-M82, and they are near perfect, an example of which follows.

Typical 5m sub of M81-M82:

IMG_2775.CR2

 

I'm at a loss as to why even the best subs from Tuesday night can't match the typical M81-M82 sub. They are okay, but looking at my M81-M82's thoroughly shows me they are crisper without a doubt. This is driving me crazy.

Finally, I attach the guide log from Tuesday night.

PHD2_GuideLog_2019-05-21_215054.txt

 

What on earth, or should I say Heaven, is going on? (A rhetorical question for the most part, but at this stage I'll try any suggestions you might have!)

Yours in agony, Sean

Link to comment
Share on other sites

Before I reply fully, some questions for you, but I would say that you are at the stage of fine tuning your PHD2 settings.

1)  I notice the 4 failed Calibrations before your final good one. Are you allowing PHD2 to autoselect a suitable star, or choosing your own?

2)  The GA suggested 0.8 seconds exposure, which suggests very choppy Seeing, but you used 2.5 seconds - is your camera not sensitive enough to use faster exposures ?

The GA is saying that to keep around the 0.10 Min Move, you needed fast corrections, and 2.5 seconds was too slow to catch the RA moves.

3)  However I'd say 0.8 seconds is a bit excessive, but what are your guidestar exposures like at say 1.5 seconds ?

Michael

Link to comment
Share on other sites

1) Four calibrations: For some reason lately the first few stars I choose for calibration do not move and eventually I get a fail message 'star did not move enough'. I'm choosing good unsaturated stars with 8>SNR but it is like I am choosing bad pixels or something even though I've tried stars with 100> SNR with the same results. So now I can tell within ten seconds or so that the star is not moving and I try another and after a few times I get one that starts to move right away. I have no idea why this is, but it never happened before this recent trouble began except for a couple of times when I was just starting out with PHD2 and I chose bad pixels. Is it possible to select bad pixels with high SNR? (BTW, you mentioned in your previous response, quote: "GA once only at the start of each night, Calibrate near each new object." Is the calibration we are speaking of above what you are talking about? (it had me somewhat confused as I thought it might mean a different calibration elsewhere, but I believe now that it is one and the same, meaning I use GA to start the night, and when I switch to a new object I simply go through the calibration routine again.)

2) GA suggestions: You mentioned taking care to follow these recommendations and that is what I did, and I swear that I saw GA recommending exposures 'between 1 and 3 seconds', and since I was exposing at 2.5 already, I left it that way. The weather may be clear enough tonight to set up at least for a while and I will scour GA recommendations to see where I missed this absolute (0.8 or whatever it says tonight) number. My guide camera will take faster exposures, I believe down to 0.1 seconds or maybe it is 0.01 seconds, but at any rate I've tried exposures occasionally at .5 second to see the how guiding reacted, so I know it will do this at least.

3) Unfortunately I can't answer that because I left the exposures at 2.5 seconds because of what I saw (what I swear I saw!) GA recommend, between 1-3 seconds. If fortune favours me with a few hours of clear weather tonight I will find the correct GA exposure recommendation you refer to and perhaps experiment with it around that recommendation to see if I can tweak it better.

From what you've said and asked I feel much better about nailing this down, and believe that you are right that a little fine tuning will see this overwith for good. Fingers crossed. Sean

Link to comment
Share on other sites

1)  To avoid locking on a hot pixel try setting the HFD to 1.5 in the Brain button/Guiding tab, and untick Star Mass Detection.

You say you are choosing good unsaturated stars, but you should allow PHD2 to select a star, it knows what's good for it ! Tools tab/auto select star

2)  In the left-hand pane of the GA results is Drift-Limiting Exposure, which was 0.8 seconds. But then it also always says to use 1 to 3 second exposures too in the right-hand of the display !

3) All I'm suggesting is to try a faster exposure, say 1.5 or 2 seconds, for 5 minutes without changing anything else, and see if it gives faster reactions to the excursions, and a smoother guiding graph as a result.

If the graph starts to yoyo then reduce the Aggression until it settles.

As I said, time to experiment a bit, but allow enough time for each new setting to stabilise.

 

Did you try balancing to take up the Dec backlash ?

 

Easing into my thoughts on your results, I didn't think your "poor" exposures were drifting that much, I'd be interested to see if others agree. 

What is your criteria? Is it the Score for each exposure in DSS ?

Try stacking all your subs and comparing with your best stack, if you're unhappy then ignore me !

Michael

 

Edited by michael8554
Link to comment
Share on other sites

23 minutes ago, michael8554 said:

1)  To avoid locking on a hot pixel try setting the HFD to 1.5 in the Brain button/Guiding tab, and untick Star Mass Detection.

You say you are choosing good unsaturated stars, but you should allow PHD2 to select a star, it knows what's good for it ! Tools tab/auto select star

2)  In the left-hand pane of the GA results is Drift-Limiting Exposure, which was 0.8 seconds. But then it also always says to use 1 to 3 second exposures too in the right-hand of the display !

3) All I'm suggesting is to try a faster exposure, say 1.5 or 2 seconds, for 5 minutes without changing anything else, and see if it gives faster reactions to the excursions, and a smoother guiding graph as a result.

If the graph starts to yoyo then reduce the Aggression until it settles.

As I said, time to experiment a bit, but allow enough time for each new setting to stabilise.

 

Did you try balancing to take up the Dec backlash ?

 

Easing into my thoughts on your results, I didn't think your "poor" exposures were drifting that much, I'd be interested to see if others agree. 

What is your criteria? Is it the Score for each exposure in DSS ?

Try stacking all your subs and comparing with your best stack, if you're unhappy then ignore me !

Michael

 

Just on my way out the door as the weather is clear for the time being, but caught your response in time to write everything down. I will see to your suggestions and get back to you.

Link to comment
Share on other sites

Here's what the PHD2 developers said about your Guide Log:

"Guiding appears to be limited by the 13.6s period RA error component.  I think that the RA guiding could be improved by using shorter guide exposures, like 1s, or 1.5 s at most.

I'd also suggest upgrading to the latest version of PHD2 (or better, the latest dev version), and running the Guiding Assistant with 1s exposures to get recommendations for the RA and Dec min-move settings.

The ~765mm focal length of the imaging scope should be relatively forgiving and the above-mentioned RA guide error may not even be visible in the main camera subs."

 

So set exposure to 1 second or thereabouts BEFORE running GA, not after as an experiment as I suggested,  my bad.

Latest dev version at https://openphdguiding.org/phd2-2.6.6dev1-installer.exe

Just download and run it, don't uninstall the older one.

Without seeing your subs, thay too don't think there will be appreciable drift.

The largest component of your PE has a 13.6 second duration,  this can be seen in the PHD2 Log Viewer if you select a smoother section of the log graph and click on Statistics, that's why a faster exposure is suggested.

Michael 

Edited by michael8554
Link to comment
Share on other sites

8 hours ago, michael8554 said:

Here's what the PHD2 developers said about your Guide Log:

"Guiding appears to be limited by the 13.6s period RA error component.  I think that the RA guiding could be improved by using shorter guide exposures, like 1s, or 1.5 s at most.

I'd also suggest upgrading to the latest version of PHD2 (or better, the latest dev version), and running the Guiding Assistant with 1s exposures to get recommendations for the RA and Dec min-move settings.

The ~765mm focal length of the imaging scope should be relatively forgiving and the above-mentioned RA guide error may not even be visible in the main camera subs."

 

So set exposure to 1 second or thereabouts BEFORE running GA, not after as an experiment as I suggested,  my bad.

Latest dev version at https://openphdguiding.org/phd2-2.6.6dev1-installer.exe

Just download and run it, don't uninstall the older one.

Without seeing your subs, thay too don't think there will be appreciable drift.

The largest component of your PE has a 13.6 second duration,  this can be seen in the PHD2 Log Viewer if you select a smoother section of the log graph and click on Statistics, that's why a faster exposure is suggested.

Michael 

More very helpful suggestions, Michael, and I hope in future to be able to take over analysis of my guiding/imaging, etc., with my newfound knowledge of the technical details we are delving into. You've been very supportive and I appreciate it immensely.

A quick summary of last night's tests: I  set up as usual, with one exception, allowing the scope to settle down for a few minutes after the slew to the third alignment star (again Vega), and no star drifting ensued or any trouble on that front, so I will continue to do this in future, though the times in past that the image drifted a good deal across the DSLR live-view while I was busy with PHD2 setup can't be attributed to the scope 'settling', in my opinion because the drift was so far and continuing so long after the scope was stationary. Hopefully, however, with the more patient approach I'll be taking it will be the last I will see (or you will read!) of this strange phenomena.

I set the parameters you suggested in PHD2 Brain (HFD: 1.5; untick Star-mass Detect), and used auto-star detect, but after several failed attempts where I got messages: HFD too low, I experimented with those two parameters and though I was able to eliminate the error messages by going back to the old parameters, I still could not get auto star detect to find a star that moved, so I selected a few of my own, but again, kept getting error 'star did not move enough'. So I abandoned M51 and sent the scope to M57, which by that time (app. 11 P.M.) was just above the tree line in the east and ready to image, though a little close to the horizon for my to attempt testing of sub quality.

I reset the Brain parameters (HFD; Sta-mass) you suggested, activated GA, then selected a star with what was probably too high an SNR at around 275, but at least PHD2 started calibration immediately, much to my relief. (Question: GA says select an unsaturated star with SNR >=8; Is 275 too high? I did't think the star I selected was saturated, but I'm not experienced enough to know.)

After GA routine I set the parameters it suggested, including the Drift-Limiting Exposure (which said 1.4), and started 3 minute exposures.

Seeing/imaging conditions were bad, with definite haze and considerable, what I believe is called ECMWF cloud, which is not so much cloud as the haze just mentioned.

Results were better than I expected, considering the challenging conditions, which probably dictated that I not attempt to compare or use the subs to change parameters in future, but I'm anxious to get this as right as possible before I wear out my welcome with you and didn't want to waste even these crappy conditions.

You asked previously after looking at the subs I passed on for your inspection what criteria I was using to grade them. I just magnify each sub and if I see 'egg' shapes then I consider them either iffy, if just noticeable; or bad, if there is a definite lengthening that would mar or blur the final image if too many were stacked with the 'good' subs. You said you didn't think my poor subs were all that bad, and that drove me again to inspect a good portion of each of the six nights of my 5m M81-82 subs taken a couple of months ago, and if I was pressed, I would reconsider somewhat my opinion that they are superior to the good subs I am capturing now. Good subs from last night and from earlier experiments through these troubles tells me that they are roughly the same as M81-82, though M81-82 are 5m exposures. The big difference is in the number of what I would consider poor subs these days, significantly higher than M81-82. Conditions may contribute to poor subs, but even some of the truly 'overcast-like' subs from last night are sharp with no hint of egginess, while on the other hand some of the poor subs came during the odd improvements in conditions. I am looking to the new PHD2 parameters and the PHD2 developers you contacted to hopefully correct these losses with yours and their analysis, and perhaps with better imaging conditions and the latest recommendations this will be solved.

 

Anyway, for what they're worth considering the lousy conditions, and after my struggle to begin imaging, I got 49 3m subs (800i), and found 18 in the poor class, as defined by me; 9 in the mediocre class that according to my criteria I would probably scrap; 9 in the so-so class that I would probably use, though with reservation; and 13 with no discernible faults whatsoever.

Ideally my goal of course is to see all my subs like these, but practically, I realize this can never be, considering all the variables that go into capturing subs. So my reasonable goal, like else's, is to limit my total losses and reduce the iffy class.

That's about it, unless you want examples from last night, though I doubt that will accomplish much considering the latest exchange between us and the terrible conditions, except to say that the GA recommendation to guide at 1.4s exposure was changed by me to 2s after checking the guide graph (not as good as a couple of nights ago, but as stated, conditions were very poor) some time into the imaging session, but with little difference that I could see, and so left it at 2s. From what you and the PHD2 techs have said, I suppose I couldn't really ask for better until I implement the changes suggested and give it another go, which looks good for Sunday night with much better conditions forecast.

I will download the link you've supplied (much thanks) to the latest version of PHD2 today and set the latest recommended parameters, including exposure setting to 1s before implementing GA.

In spite of all these troubles and my claim to going crazy over them, I find all this extremely interesting, and I'm looking at this as a very good lesson in guiding/imaging, and can see that much has been rectified up to this point, and the potential there to get fully back on track.

Big thanks, Sean

Edited by Seanelly
Link to comment
Share on other sites

HFD 1.5 might be inappropriate for your setup, I still think it's the best way to avoid autoselecting hot pixels:

"Minimum star HFD' - specifies the minimum half-flux-diameter (roughtly the 'size') of a suitable guide star.  This is probably the best way to prevent PHD2 from mis-identifying clumps of hot pixels as usable guide stars.  You can determine a suitable value for your system by manually selecting some small stars that you know are not just hot pixels, then use the star-profile tool to see the HFD values of those stars.  You'll want to specify a minimum HFD value that allows selection of legitimate faint stars but not hot pixels."

Star Profile is in the Tools menu. A saturated star will have a flat top to the profile, and you'll get a red SAT message on the bottom of the PHD2 window.

 (You can refine your guidescope focus by getting the lowest HFD figure in the Star Profile window as you focus.)

I don't believe you can have too high a snr, but you can have too low, the SNR message at the bottom of the PHD2 window should be green, it goes brown if getting too low.

I noticed the other night you focused on Vega, then slewed over to M51. Remember my concerns over micron moves in your guidecam mount ?

Well  the same applies to your dslr, I always set focus on or near the target because the heavy dslr will sag differently after a slew to a different position. 

Worth doing if you're picky about star shape 🙂

You should continue with your sub rejection process because you are the best judge of what's suitable.

But you should aim to improve your guiding to lower that reject rate.

Michael 

Link to comment
Share on other sites

2 hours ago, michael8554 said:

HFD 1.5 might be inappropriate for your setup, I still think it's the best way to avoid autoselecting hot pixels:

"Minimum star HFD' - specifies the minimum half-flux-diameter (roughtly the 'size') of a suitable guide star.  This is probably the best way to prevent PHD2 from mis-identifying clumps of hot pixels as usable guide stars.  You can determine a suitable value for your system by manually selecting some small stars that you know are not just hot pixels, then use the star-profile tool to see the HFD values of those stars.  You'll want to specify a minimum HFD value that allows selection of legitimate faint stars but not hot pixels."

Star Profile is in the Tools menu. A saturated star will have a flat top to the profile, and you'll get a red SAT message on the bottom of the PHD2 window.

 (You can refine your guidescope focus by getting the lowest HFD figure in the Star Profile window as you focus.)

I don't believe you can have too high a snr, but you can have too low, the SNR message at the bottom of the PHD2 window should be green, it goes brown if getting too low.

I noticed the other night you focused on Vega, then slewed over to M51. Remember my concerns over micron moves in your guidecam mount ?

Well  the same applies to your dslr, I always set focus on or near the target because the heavy dslr will sag differently after a slew to a different position. 

Worth doing if you're picky about star shape 🙂

You should continue with your sub rejection process because you are the best judge of what's suitable.

But you should aim to improve your guiding to lower that reject rate.

Michael 

Okay, Michael, I will familiarize myself with HFD settings and guidescope focus to find a value that works for me.

As to focusing on Vega before slewing to M51, I have no choice at present but to focus using the Bhatinov mask on a bright star because even in the 10x DSLR box the cross-hairs and floating focus bar are still quite dim and focus tricky. I've tried the B-mask on dimmer stars around the objects I'll be imaging but they are too dim unless a luckily positioned bright star is nearby. I realize that using APT or similar app (I assume this is how you get focus with your DSLR) to control my sessions is optimal and focusing is easier but I had no problem without it, really, using a digital remote shutter release as I do. I avoided using APT (I already have it on the laptop) right off when I started into imaging this past winter because I wanted to concentrate on the basics and getting things up and running before jumping into another set of parameters. Once I've got things under control here I'll definitely look into setting it up.

Yes, it seems I am at the stage of refining the tools at my disposal to limit bad subs and reduce unwanted egging. Perhaps the new suggestions to be implemented tomorrow night will do the trick!

Edited by Seanelly
Link to comment
Share on other sites

I've looked at the screenshots of BYE and APT and found them to be too complex for my needs, filling the display with unwanted tools and info.

I use Canon Utilities which takes up little space on the screen.

Good luck with your next session

Michael 

Link to comment
Share on other sites

This my Focusing setup:

Untitled.jpg.99ca6f8bd0493175f916532c38c38eee.jpg

The Canon Utilities Zoom View is showing the magnified LiveView image

The red circle in Bahtinov Grabber hovers over the centre of the cross when correctly focused, with focus error text to the left.

The small Canon Utilities remote has most of the camera settings available, the red circled icon opens a small long exposure sequencer that fits nicely in the space under the remote.

As each sub downloads it opens ZoomBrowser full screen and  displays a large image and a Histogram.

Nice and simple and relatively uncluttered.

Michael

  • Like 1
Link to comment
Share on other sites

Canon Utilities, I will look into this. Are you able to get focus with dim stars around your object or must you locate a fairly bright star nearby?

I have seen the tutorial in the link you've supplied, and now that I know more than I did about PHD2 I'll keep it handy to reference, but having initially looked through it a few times it can be intimidating to the uninitiated. I have better aptitude for hands on teaching rather than being supplied off the written page, sort of like being shown rather than being told, if that makes sense, and retain the lessons better this way, which is why I've leaned so heavily on your advice. To which I am stoked to tell you has paid wonderful dividends!

4 hour imaging session last night: Having to rely on a bright star (for now) for focus, and keeping in mind your warning about possible flexing after slewing the heavy DSLR to a distant object afterwards, I went straight from focus (Vega) to M92, a less unpredictable slew than M51, and having applied all the GA suggestions (1s Drift-limiting exp suggested, which was already in use from the pre-GA setting you mentioned to change), I fired off twenty 2m subs of M92, then having M57 clearing the treetops around 11:30P., slewed there and got calibration again, then let the scope really do it's works with 72 3m subs.

Results: Fantastic! Conditions were much better, though seeing was still rated rather poor, and after minor fiddling on a hunch (finally taking some initiative in this), in which I reduced the aggression on changes to Dec and RA to control initial sporadic spiking, guiding was so good it brought tears to my eyes. It only occasionally climbed beyond 1" either way all night. What a difference! I kept creeping out to the laptop as time dragged on (I couldn't help myself and sacrificed some sleep over it) to check and see if the other shoe was going to drop but it remained excellent all night.

Subs came out great, most better than I've ever captured, and I didn't lose a single one of the 20 M92's, all stars pinpoint with no hint of stretching. Out of 72 M57 subs I lost 7, which were only 'bad' when compared to the good ones, which were so crisp and clean the central star showed nicely in each 3m sub, with the occasional hint of the second. Another 8 of these 'good' subs showed minor variations, but again only compared to the excellent ones, and most I will try at least once in a stack to see if the results are better or worse than just the best ones.

 

PHD2_GuideLog_2019-05-26_220049.txt

 

I supply the log for you verify the results, and perhaps it may be seen that I've exaggerated somewhat, but I believe not much, as my subs do not lie. I may have seen the improvements from M92 and M57, but I'm over the moon about it all!

Link to comment
Share on other sites

47 minutes ago, Seanelly said:

Are you able to get focus with dim stars around your object or must you locate a fairly bright star nearby?

You can up the ISO to silly figures, but a bright star nearby is better.

 

48 minutes ago, Seanelly said:

I have better aptitude for hands on teaching

Same here, which is why I can't get my head around PemPro, SGP, PixInsight etc, Still a worthwhile read.

Successful Calibrations first try - what did you do ?

Scond guide session - outstanding !

Now you can offer advice to the next  problem poster !

Canon Utilities comes on the CD with your camera, or you can download a CD image from Canon USA

Michael

Link to comment
Share on other sites

Calibration: I set up the parameters again that you mentioned, HFD, minus Star mass detect, and then on a whim rotated the guide camera slightly back and forth in it's mount, and it turns out I cannot always get ideal focus on the stars unless I pull the camera out a few millimeters, which may have been part of the guiding problem, I don't know, though I don't recall in the past selecting stars or having them selected that looked out of focus. Anyway, once I had ideal focus on what I knew were actual stars, Auto select grabbed the first one and away it went no trouble. I'll just have to keep an eye on this procedure in future, maybe make a spacer for the guide camera.

Ya, again, guiding just blew me away. Once I got into that second session and tinkered a bit and saw the results, I sat back and held my breath a while but other than the odd spike or short meander, it kept on beautifully. I captured a couple of screen shots but silly me I can only find the text documents and not the photos, which are probably lurking around with those subs I was looking for earlier, haha.

I'll certainly do my best to return the favour of help to anyone I can.

I'll check out the Canon CD (I didn't know anybody actually looked at the CD's that come with most electronic devices these days). When I am absolutely certain my issues are behind me I will look into the Utilities or try APT or some such to consolidate my sessions, if only for more precise focus, and download the latest PHD2 version, which I hesitated to to jump into until things had settled down. I still hesitate, now that it seems my guiding is where I want it. No rush, at any rate, I'm going to enjoy a few nights of solid imaging without headaches, or without those specific headaches.

I'd hug you without reservation if I could, Michael, but you'll have to settle (whew) for a big thank you.

 

 

Link to comment
Share on other sites

It's gratifying when one's long range detective work turns out to have been helpful. 

A final thought. 

I reported on Saturday that your RA had a 13.6 sec component.

It has been suggested on the PHD2 forum that down the road you rotate the current mesh position of the  intermediate gear in the RA gearbox (5).

I would look before and after at the Analysis section of the Log Viewer to see the effect, finding the sweet position will be hit or miss. 

motor gear play.png

Michael 

Link to comment
Share on other sites

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 account

Sign in

Already have an account? Sign in here.

Sign In Now
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.