Ovoids with Mesu and Tak FSQ 85

[gnomus]

I am someone who is still trying to figure out when I should ‘soldier on’ rather than admit defeat.  Last night may have been a case in point.  The moon was pretty huge - mid-90% size.  Nevertheless, (and despite advice from people who know what they are doing) I decided that I would press on and try to get some Ha of the Pelican Nebula.  I went out at around 8pm, I rolled the roof off, and I fired everything up.  Then I took a peek over the observatory wall.  In the half hour or so since going out, it had completely fogged up - dense fog.  So, I shut everything down and rolled the roof shut again.  About an hour later, things seemed to have improved considerably, though ‘Clear Outside was still showing ‘Red’ for dew point and relative humidity.

 I went back out and tried again.   

Most of the time I got really good guiding – flat graphs with <0.2 pixel RMS error.  Every now and again the graph would throw a slight wobbly with Dec moving between +1 and -1 arcseconds for a few seconds, but overall it didn’t seem too bad – My Tak 85 and Atik 383L give me a resolution of 2.47” x 2.47”.  I was taking 20 minute subs and, on my warm room screen, SG Pro seemed to be showing me reasonably good images.  However …… Once I got indoors and looked at my subs on a decent screen I saw that I had some star trailing.  I attach some 100% crops taken from the middle and the four corners of my image - you may need to click on the image to make it the 100% size.:

It seems to me that all of these crops show stars with a top-left to bottom-right slope.

When I finished for the night (some 3 hours later) it was quite one of the wettest evenings I had encountered.  The equipment was all soaked and there was water streaming down the observatory walls.  The dehumidifier seemed to have dried everything up by morning.

Is this then simply a problem of 'seeing' or have the conditions pointed up a more fundamental problem with my set up?  I use an un-reduced Tak 85 (at least for the moment).  

[steppenwolf]

To my eye (but please bear in mind that I wear vari-focals), the bottom left and bottom right look pretty good to me - this could be a case of sensor tilt rather than tracking?

[gnomus]

Thanks Steve. If it is sensor tilt, what am I supposed to do about that?

[ollypenrice]

Firstly (call me sloppy) I'd be pretty happy with that and be inclinded to get on with some nice processing!

I can't see anything to complain about other than in the top two and, of those, the top left is slightly worse.

If part of the image is OK it can't be tracking, so that's ruled out.

That leaves tilt and polar alignment. My guess is tha latter.

The effects of polar misalignment will vary with the relationship between the position of the guide star and the centre of the chip. The only test of PA in which I believe implicitly is drift. 

Olly

[steppenwolf]

Thanks Steve. If it is sensor tilt, what am I supposed to do about that?

Sensor tilt can be a bit of a nightmare to resolve but it really boils down to tightening everything up within an inch of its life at the point that you are as positive as you can be that all the component matings surfaces are flush with one another.

Polar alignment? Could be but I suspect that you have that nailed?

[gnomus]

Firstly (call me sloppy)

.....

Olly

If you insist: You're sloppy.

Thanks for your response.  I went ahead and processed the 8 x 20 min subs I took last night.  I still think I'm getting ovoids in the centre (highlighted in red).  The other thing I have noticed on all the images I have taken with the Tak are small diffraction spikes (best one highlighted in yellow).  I must say I was not expecting to get these with a refractor.  Again, you'll probably need to click on the image to get the 100% crop.  Am I expecting too much from the equipment?

Sensor tilt can be a bit of a nightmare to resolve but it really boils down to tightening everything up within an inch of its life at the point that you are as positive as you can be that all the component matings surfaces are flush with one another.

Polar alignment? Could be but I suspect that you have that nailed?

I thought I had reasonably good polar alignment.  Certainly the PHD assistant tool reported excellent PA (but I don't know how reliable this is).  Following your comments, I went out and checked the tightness of everything.  I was able to tighten up the Tak rings and connections by a very small amount (a 'wee thochty' as they say back home).  I am not sure if this could have been responsible.  Something interesting did happen when I was doing this though - the whole scope rotated slightly in the clamshell.  It is possible that I hadn't tightened this up sufficiently (trying not to crush the tube).  Could this make a difference I wonder?

The final thing I have been thinking about are the 36mm unmounted filters that I have in my EFW2.  Fitting these filters into the wheel was a fiddly job to say the least.  Is it possible that one or more of these are not exactly square?  Would that make a difference?  If so, how do you get these things in square?????

I had hoped I'd be imaging by now - not still troubleshooting, but there you are. 

[ollypenrice]

I think you're pixel peeping, which has its virtures but should be kept in its place. If you look at Tak FSQ images you see a lot of excellent ones. You also see plenty of stellar artefacts. I get the odd PM or email asking if I've noticed this or that that is wrong with my stars and do I mind? I have noticed and I don't mind. Here's an example. Have a look at the bright young blue stars on the left. They have classic Tak 'inverse light house beams' coming from either side. Greg Parker tells me that they arise from pinching. He knows more about optics than I do so I'll settle for that. https://ollypenrice.smugmug.com/Other/Best-of-Les-Granges/i-Vtk6nfM/0/O/VDB152%20HaOIIILRGB27Hrs.jpg

But what interests me in this image is that the Tak has allowed me to catch more of the red SN remnant than I have ever seen before on the net. That doesn't mean nobody has gone deeper, it just means that I haven't seen it (and, yes, the processing is exaggerated but that was my intention - to drag it into view via colour.) So my advice - and my own policy - is not to let the details blind me to the big picture.

Anyway I'm still backing PA... but it could be tilt. The standard test for tilt is to rotate the camera.

Olly

[gnomus]

I think you're pixel peeping, which has its virtures but should be kept in its place. If you look at Tak FSQ images you see a lot of excellent ones. You also see plenty of stellar artefacts. I get the odd PM or email asking if I've noticed this or that that is wrong with my stars and do I mind? I have noticed and I don't mind. Here's an example. Have a look at the bright young blue stars on the left. They have classic Tak 'inverse light house beams' coming from either side. Greg Parker tells me that they arise from pinching. He knows more about optics than I do so I'll settle for that. https://ollypenrice.smugmug.com/Other/Best-of-Les-Granges/i-Vtk6nfM/0/O/VDB152%20HaOIIILRGB27Hrs.jpg

But what interests me in this image is that the Tak has allowed me to catch more of the red SN remnant than I have ever seen before on the net. That doesn't mean nobody has gone deeper, it just means that I haven't seen it (and, yes, the processing is exaggerated but that was my intention - to drag it into view via colour.) So my advice - and my own policy - is not to let the details blind me to the big picture.

Anyway I'm still backing PA... but it could be tilt. The standard test for tilt is to rotate the camera.

Olly

I am quite certain I am pixel peeping!  At 100% I am a little suspicious of those stars, but I really need to go to 200% (or 2:1 in PixInsight) to see the ovoids clearly.  

I will check polar alignment again if I get the chance (recent weather here has been beyond a joke).  Could this be cause by a little bit of cable drag on an insufficiently tightened scope (see my previous post)?  I can also rotate the camera - I presume that if I rotate it 90 degrees I would expect the direction of ovoid-osity to change by the same amount.

Edited October 29, 2015 by gnomus

[Zakalwe]

To my eye (but please bear in mind that I wear vari-focals), the bottom left and bottom right look pretty good to me - this could be a case of sensor tilt rather than tracking?

I would agree with that...the bottom corners look OK to me too. the top right looks marginally worse than the top left.

It could be sensor tilt, a bit of slop in the focuser, flex in the filterwheel front or rear plates. it could be a right old so-and-so to track down.

Steve, are you using a fully threaded connection from the focuser to the camera? Do you have CCD Inspector to run some subs through?

[Zakalwe]

I am quite certain I am pixel peeping!  At 100% I am a little suspicious of those stars, but I really need to go to 200% (or 2:1 in PixInsight) to see the ovoids clearly.  

I will check polar alignment again if I get the chance (recent weather here has been beyond a joke).  Could this be cause by a little bit of cable drag on an insufficiently tightened scope (see my previous post)?  I can also rotate the camera - I presume that if I rotate it 90 degrees I would expect the direction of ovoid-osity to change by the same amount.

Wouldn't the PA have to be a far bit out to get field rotation?

[gnomus]

I would agree with that...the bottom corners look OK to me too. the top right looks marginally worse than the top left.

It could be sensor tilt, a bit of slop in the focuser, flex in the filterwheel front or rear plates. it could be a right old so-and-so to track down.

Steve, are you using a fully threaded connection from the focuser to the camera? Do you have CCD Inspector to run some subs through?

I am using a fully threaded connection and all connectors are Tak parts.  I do not have CCD Inspector (and am not sure I'd know what to do even if I did).  I can post Dropbox links to a couple of subs if that would help. 

[sloz1664]

Hi Gnomus

Post drobox links and I'll check with ccd inspector.

Steve

[derrickf]

If you insist: You're sloppy.

  The other thing I have noticed on all the images I have taken with the Tak are small diffraction spikes (best one highlighted in yellow).  I must say I was not expecting to get these with a refractor. 

I don't believe these are diffraction spikes caused by obstructions (normally secondary supports etc.) but are probably artefacts produced, when imaging at fast focal ratio, with CCDs fabricated with micro-lenses over the pixels.

This was a  hot topic in the SBIG forums several years when the ST-10XE (non-micro-lensed) was superseded by the ST-10XME (micro-lensed). Many people who rushed to replace their ST-10XE with an  ST-10XME, because of its significantly better QE due to the micro-lenses, could not understand why they were suddenly seeing "diffraction spikes", often only vertically, in images made with fast refractors. It was eventually shown that the effect was only observed at fast f-ratios. The horizontal artefacts on the bright stars are probably caused by the pinching effect described by Ollie.

HTH

Derrick

[steppenwolf]

 I can also rotate the camera - I presume that if I rotate it 90 degrees I would expect the direction of ovoid-osity to change by the same amount.

Indeed, this will confirm whether or not there is sensor tilt - be aware though that there are other potential causes for the sensor to be tilted to the light cone including focuser slop, that awful adjustable connector that you get with the FSQ 85.

I would agree that the diffraction spikes on the bright stars are indeed the micro-lenses on the KAF 8300 sensor as this is a well known issue with this chip (note that the spikes are exactly N/S and E/W in the field of view).

Ollie makes a very valid point about the 'pixel peeping' though (and many of us including me are guilty of doing it) - your image is far removed from a disaster and with our limited imaging opportunities, you could just get on and enjoy yourself!

[gnomus]

Hi Gnomus

Post drobox links and I'll check with ccd inspector.

Steve

Thanks Steve (Is everyone on this site called Steve?)

Links:

https://www.dropbox.com/s/fbkvlc0rkdrykpe/Pelican_1200sec_1x1_HA_frame3.fit?dl=0

https://www.dropbox.com/s/6dl0o1d0mediw1e/Pelican_1200sec_1x1_HA_frame6.fit?dl=0

Frames 3 & 6 - the last ones taken before an autofocus run.  

I'm very grateful to you for this.

Steve

[gnomus]

Indeed, this will confirm whether or not there is sensor tilt - be aware though that there are other potential causes for the sensor to be tilted to the light cone including focuser slop, that awful adjustable connector that you get with the FSQ 85.

I would agree that the diffraction spikes on the bright stars are indeed the micro-lenses on the KAF 8300 sensor as this is a well known issue with this chip (note that the spikes are exactly N/S and E/W in the field of view).

Ollie makes a very valid point about the 'pixel peeping' though (and many of us including me are guilty of doing it) - your image is far removed from a disaster and with our limited imaging opportunities, you could just get on and enjoy yourself!

Thanks Steve

You are probably right.  I was quite pleased with the images last night as they downloaded from the camera.  I don't mind the mini-spikes, I just was not expecting them.  I understand that my image is not a disaster, but, at this price point, I was hoping for a significant step up.  I think I got rounder stars with my ED80 (and on a lesser mount).

[Steve 1962]

Hi Steve....another Steve here!

I agree with the other Steves (what sort of name is Olly for goodness sake?) - the spikes look like micro lensing. I get the same with my Atik 460.

The stars look bit odd to me, but given you're working at the native focal length of the scope and spacing isn't an issue I'd be inclined to live with it. I found that any egginess from my Tak 106 "stacked out" during processing.

I agree with Only about pixel peeking. Your camera (and mine) has small pixels and therefore has a high sampling rate. When you see perfectly flat images from a Tak they're gemerally paired with something like an 11000 chipped camera. These have 9um pixels which cover 4 x the area of our pixels. They therefore are more tolerant of tilt, curvature and all the others things we worry about. We've got higher resolution of the target..and field defects. So just process down the stars and enjoy the signal!

Edited October 29, 2015 by Steve 1962

[steppenwolf]

but, at this price point, I was hoping for a significant step up

I hate to say this but you know what? I had more fun and got more imaging done with my old blue tube ED80, EQ6 (EQMod) and SX M25C than I currently get with my 'inheritance' outfit!

Sadly, my experience has been that spending more money does not necessarily, a better image make - I even sent my own FSQ 85 back for a refund because I wasn't happy with the results! The problem with better gear is that one's expectations climb unrealistically which is why my current project (automation) is aimed purely at increasing my time on the sky with the gear that I already have.

What better gear should do is put the onus on the operator to get the best out of it and not blame the equipment if the results aren't great but the risk is pixel peeping - something I have done too much of in the past - now I just want to capture some photons during the gaps in the clouds ...........

I went through this with Hi Fi at one time and realised that I was listening for the 'pops' and 'flutter' and missing the music!

[sharkmelley]

I downloaded your two frames (frames 3 & 6).  In both frames, the stars have an overall elongation top-left to bottom-right slope, just as you said.  Now blink the two frames - you'll see all the stars move in that very same direction between frames 3 and 6.  Coincidence?

I strongly suspect your have some kind of flexure somewhere, so the image is slowly drifting relative to the guide camera's view.

Mark

Edited October 29, 2015 by sharkmelley

[gnomus]

I hate to say this but ...

What better gear should do is put the onus on the operator to get the best out of it and not blame the equipment if the results aren't great ...........

Ouch.  However, I have not found this to be the case with astronomical or indeed any other type of gear, and I too have suffered from the 'hifi sickness'.      

What I am trying to do with my question is ascertain whether or not I do have a significant issue and, if I do, seek advice as to how to fix it (thereby getting the best out of my equipment).  I may well be a nincompoop, but I was getting rounder stars with my ED80.  

[swag72]

.......but, at this price point, I was hoping for a significant step up.  I think I got rounder stars with my ED80 (and on a lesser mount).

I think that it's fair to say that once you start spending more money you begin to see less and less of a difference. I've often seen it quoted that you're now paying a few thousand ££'s more for a very small increase in quality.

The ED80 is a good scope and is excellent value for money. I wouldn't expect to see a huge difference in quality between that and a Tak.

..... Just my opinion

[Herra Kuulapaa]

Very important and good view. I personally have a tendency for speed blindness and that has had a cost effect for me.

Worst thing for me seems to be also that I don’t know what I want in the end..

[gnomus]

I downloaded your two frames (frames 3 & 6). In both frames, the stars have an overall elongation top-left to bottom-right slope, just as you said. Now blink the two frames - you'll see all the stars move in that very same direction between frames 3 and 6. Coincidence?

I strongly suspect your have some kind of flexure somewhere, so the image is slowly drifting relative to the guide camera's view.

Mark

Thanks for looking and confirmng my suspicions Mark. The elongation is all in the same direction - it is worse in some corners but it is present in the centre. I am dithering, so it is just possible that the 'blink' test is coincidental.

I will recheck cabling and balance. With the Mesu, one cannot 'declutch' and It is not always clear where the precise point of balance is.

Edited October 30, 2015 by gnomus

[gnomus]

I think that it's fair to say that once you start spending more money you begin to see less and less of a difference. I've often seen it quoted that you're now paying a few thousand ££'s more for a very small increase in quality.

The ED80 is a good scope and is excellent value for money. I wouldn't expect to see a huge difference in quality between that and a Tak.

..... Just my opinion

Thanks Sara. I quite understand the law of diminishing returns, it is the direction of travel (quite literally in this case) that concerns me.

[ollypenrice]

I downloaded your two frames (frames 3 & 6).  In both frames, the stars have an overall elongation top-left to bottom-right slope, just as you said.  Now blink the two frames - you'll see all the stars move in that very same direction between frames 3 and 6.  Coincidence?

I strongly suspect your have some kind of flexure somewhere, so the image is slowly drifting relative to the guide camera's view.

Mark

If the guide star were well off-axis relative to the Tak then wouldn't polar misalignment also produce this effect?

Indeed, this will confirm whether or not there is sensor tilt - be aware though that there are other potential causes for the sensor to be tilted to the light cone including focuser slop, that awful adjustable connector that you get with the FSQ 85.

Which connector is that, Steve? Since mine departed some years ago I can't remember it in every detail but the only connector to give me gyp was the tilt-adjuster (if that's what it was meant to be) on the reducer. Steve (Gnomus), are you using the reducer here? If you are then the 'three radial screw horror' is designed to test your tilt adjusting skills! I just loosened the screws, put the flattener on the table, pressed the top part down onto the bottom and tightened them. That worked in my case.

Steve. (Just going with the flow - or should that be FLO?  )

Edited October 30, 2015 by ollypenrice