Whirlwind

8 hours ago, astro mick said:

Nice dehumidifier,but a bit pricey for me.

These cameras must be built to withstand moisture,as they can become dripping wet on some nights,and there is no warning saying they have to taken back inside and dried out.

People who operate remote telescopes must have cameras attached all the time,but of course they might have a different set-up.

I,m leaning to-wards just leaving the camera attached.

Fingers crossed.

Mick.

A lot of remote observatories are located in dry locations (as high humidity also means cloud and rain) so it's less of an issue for these types of locations.  It is the external electronics you need to think about as well as connectors.  Most CCDs/CMOS are in purged chambers with some manner of reducing any moisture in the chamber (otherwise you get ice on the chip).  You never want any electronics sopping wet (that includes the mount) and you also have to be careful with the telescope as well because of mould or rusting cells etc.

Good dehumidifiers can be expensive but they are still cheaper than replacing the telescope equipment!  Think of it as insurance.  Dew heater bands are possible but is a bit of a fire risk as they aren't really designed to be always on so you might want to set it up that they only switch on in certain circumstances.

31 minutes ago, dan_adi said:

Well I have 90 mm at disposal between the flattner/corrector and the ccd sensor. I was looking into Optecs motorized Sagitta OAG but it eats up 31.7 mm. The G3 with filter wheel eats up additional 47 ish mm and I have a manual rotator with about 30 mm... So I am over the limit with Optecs Sagitta.

Would it be a bad ideea to place the OAG before the flattner? 

Never tried putting the OAG in front of the flattener but you will get more distorted stars in the OAG because of the lack of flattening.  It would be better to put the rotator ahead of the flattener as long as it can carry the weight.  Also I think the G3 with filter wheel is shorter than stated unless it is a third party filter wheel?

Are you automating the set up as if you are that usually means the camera is 'on' permanently and generating some heat so should keep some dew away.  I think a dehumdifier in an observatory is a must in the UK which would also help.

It depends on what you are going to use.  Both can be successfully used.  OAGs mean that you don't risk flexure between two telescopes but usually have a smaller field because of the limitations as to where the pickoff prism can be.  Also an OAG adds additional back focus consideration that needs to be accounted for (though most these days are relatively thin) but can be an issue if you trying to fit something else in like a rotator.

I tend to image with an OAG as I find it easier

Yeah, that would help as there is some compression issues happening somewhere?

20 minutes ago, souls33k3r said:

For the love of God I can't remember. I think they're the same but the only difference is that the first image was right out from PI and the second one (marked with red) was something I did on my phone. If that made difference to the file then I don't know

I've attached a comparison of the same stars between the image that shows that different in resolution.  The ringed one was giving a focal length of around 3500mm which seems high.  Also found out where the fixed noise was coming from. Apparently it was being added by the platesolve so that query is resolved.

Could you expand what you are plotting as all I see is some numbers and a plot?

On 23/11/2020 at 21:35, souls33k3r said:

I've marked the stars for you. You'll have to zoom in to see that. 

Are these the same two images or is the second one an uncompressed version of the first?  The plate solving of each one is completely different and there is obvious fixed pattern noise in the second compared to the first (which is much smoother and more random).  In the latter the ringing artefacts are much more apparent though.

Ah OK, I see them now, I used the bottom right one but couldn't see it (compression on my screen perhaps) but there are some better ones. I'll do some measuring tomorrow.

20 hours ago, souls33k3r said:

Apologies for the delay in replying, just had a busy few days. 

Here are the single and stacked Ha subs. Look at the bottom right hand corner star which like I said is bigger of the bunch. 

I'm a bit confused but I don't see any ringing on these images?

23 hours ago, souls33k3r said:

You will have to probably dumb it down for me. Run what numbers and how? 

So if we assume it is Newton's rings then there is a specific equation that can be run which is where the radius of the nth bright ring is given by multiplying the wavelength observed the radius of the spherical surface multiplied by (the ring number measuring - 0.5) all to the power of 0.5 (easier to read here https://en.wikipedia.org/wiki/Newton's_rings (first equation).

As such what we need is a distance measurement from the centre to the centre of a bright ring (and knowing which one) and as we have a specific wavelength we should then be able to determine the radius of the spherical surface. If this is run for each filter then we should be able to determine whether the radius of the spherical surface is the same.  If it is then that is a strong indicator that it is newtons rings and that it has one specific cause.  If it isn't the same then that might imply that Newton's rings isn't the cause or that it is a different source.  As the filters would be the only thing that was changing that would give a lot more empirical evidence that it is the filters at fault.

To start you could count the pixels from the centre to which ever ring you choose (and count which ring).  With the pixel scale that will give a distance.  You can run this for each filter (note you need one in the centre of the field of view so other optic distortions don't come into play).

If you have a solar telescope you could also use the camera on the sun.  It passes the same wavelength but you can remove the filters as potential cause if the same effect is seen.  On the other hand no effect would more strongly suggest a filter or other problem (or isn't N'rings).

9 hours ago, souls33k3r said:

I found someone on Instagram who luckily has the absolute exact same setup as mine, even to the second scope. I've been exchanging messages with him and he was kind enough to share some single subs, stacked masters and processed image and they don't exhibit the same issue. 

That's fortunate and interesting because that would suggest a different manufacturing tolerance level.  For example it could be that the chip is tilted slightly or even that the cover slip isn't lying against the micro-lenses.  Still think the best option would be to run the numbers and see whether you get the same radius causing the effect in each filter as that would point to a specific piece of equipment.

17 hours ago, Adam J said:

My sig is 2 years out of date so you can't work anything off my signature...

One thing I would say is that the place I have seen this effect more than any other is in narrow band images taken by the Hubble space telescope and I am sure there are no issues with its camera or filters. It will be a combination of  factors acting together as opposed to any single bit of kit being at fault. Hence I suspect op will just have to live with the effect. 

Yeah, I did note that it was based on your signature, so if it is out of date then less can be inferred.  But the principles still stand.  

As for space telescopes they don't tend to have cover slips that consumer versions do.  They are prepared in clean rooms so whereas consumer versions have to take some abuse from the environment.  They tend to introduce issues in the far red and infra-red which isn't great for photometry and spectroscopy of cool (or distant) objects.

1 hour ago, Adam J said:

I dont see this at all on my Astrodons with ASI1600MM Pro so its not to do with other filters masking the issue. It could be a result of the low F-ratio optics and large aperture resulting in newtons rings...The only time I have ever had this effect was with a coma corrector, I sent it back and it turned out that they had not coated two of the lens surfaces.

Adam

So just working on your signature, there are two things to note.  Firstly your Astrodons are 5nm, the extra width could be enough to 'bloat' the medium stars just enough to hide the effect.  Secondly you are imaging with a newtonian + coma corrector which may again bloat the stars slightly compared to the EdgeHD as I have no idea which produces smaller stars (also the airy disc size could be different/larger again masking the effect - airy disc is naturally smaller in a larger telescope/but have to factor in CO).  Another possibility is that the cover slip on your camera is tilted enough compared to the OP to remove the effect.  Without a side by side comparison and literally swap the cameras it is difficult to tell. I would still suggest running the numbers to see if we can find whether the sphere radius causing the issue is the same in all filters.

4 hours ago, souls33k3r said:

Unfortunately they are worthless to me if it doesn't work the way they are intended. I don't have to put up with software wizardry to fix something that shouldn't even be present in the first place. I really can't put up with this especially I paid nearly £500 per filter. If I was imaging a bright star then yes it makes sense for the microlensing effect to be very much visible and may even have the newtonian rings but on a small star? Certainly something is not right here. 

Newtons rings is a different from the more traditional microlensing effect you might get with and ASI1600.  In fact that you don't see it on the bright stars makes me more certain that Newton's rings are the cause.  The more traditional microlens effect is a result of pure reflection from surfaces that have less than ideal coatings (and hence are more obvious around brighter stars).   The newtons rings arise because there the you have a variable boundary between a planar and spherical surface resulting in constructive and destructive interference arising from reflection at the boundaries.  

The radius of any given ring is proportional to only the radius of the spherical surface and the wavelength of the light being observed.  This gives an explanation as to why very bright stars don't show the effect.  Regardless of the brightness of the star any given ring is at the same distance from the star.  However the flux of a much brighter star is 'scattered' over a wider radius.  This flux hence overwhelms the effect of the newtons rings - it's still there but is simply swamped by the actual signal.  For faint stars there is too little flux to detect above the background.  It's only the medium bright stars where there is enough flux but are still relatively small.  

I am assuming that you have moved to narrower filters than previously.  Broader filters are likely to have hidden the effect as well.  Newton's rings are wavelength dependent and as such each wavelength would constructively and destructively interfere at different radii from the star.  When you conflate the various wavelengths together they balance out across the same radius and it all merges (so the effect is masked).  

So where could this come from. If it is the filter then it would needed to have been made poorly as you would need part of the surface to be spherical and have pulled away from another planar element of the filter.  As it is an interference filter this is not impossible.  However, as it is present in all three filters then it would imply quality control at Chroma has gone down the pan and the lack of other similar issues reported elsewhere suggests it is unlikely or that you were very unlucky to get this with three filters.  Conversely we do know of an area of your system where this is a known planar and spherical touching surface.  This being the cover slip and the microlens. In probability terms given this is 'a known' for the sensor it is more likely it is arising at this location (it doesn't matter which filters you use it is always there but may just be washed out where broader filters are used).  To an extent each of these options can be tested.  The formula for Newton's rings is known.  As you know the radius of a number of rings for each wavelength, you could measure these and then calculate the radius of the spherical surface causing the effect for each filter band.  If this comes out as approximately the same then that would imply the same source (and more likely to be from the sensor).  If each were widely different then that would suggest another source (as the radius of the microlens would be the same regardless of wavelength and hence more likely to be the filters).  Note you would need to measure closest to the centre of the optics to limit optical distortion (i.e. near centre of the chip).

1 hour ago, souls33k3r said:

 

Thank you for the link and your input Ken. The CN thread is mostly tested on a super bright star which I won't ever be going anywhere near to but your data has somewhat smallish star in the elephant trunk image. Did you try flipping the filters afterwards to see if this effect goes away? 

This really worries me because I've paid a good chunk on these filters which are going to be worthless if they are not going to work for me. It really is heart breaking to say the least. 

They aren't worthless.  They are just of such a quality that you are showing up other effects in your system.  Newton rings occur when you have a spherical surface touching a planar surface.  For the ASI1600 this occurs because you have the microlens (spherical surface) touching a planar surface (the cover slip).  Sometimes you can tilt the system so that the light rays aren't parallel to the surface (so you don't get the effect) but if it occurs because of the camera then it's not likely to help - which I suspect in this case because of the small rings being visible.  Both CCDs and CMOS can show this effect (and depends on the design).

You can manage this from a software perspective by masking the stars and adding in a gaussian blur or similar (or you could replace with RGB stars).  It's also worth remembering that using 3nm filters is to get the best out of the nebulae which you can still do. 

Here are some other examples using different set ups.

https://www.cloudynights.com/topic/728648-reflection-artifacts-in-the-asi6200/

The effect is due to Newton's rings

https://en.wikipedia.org/wiki/Newton's_rings#:~:text=Newton's rings is a phenomenon in which an,the effect in his 1704 treatise Opticks

It's known on ASI1600s and believe it is an interaction between the cover slip and microlens of the sensor.  You are hence not likely to be able to get rid of the problem (because you can't tilt angle between these two glass elements.  It's probably occurring now because the filters are higher quality than previously (which probably masked the issue).

4 hours ago, tooth_dr said:

I have decided to buy a new CMOS camera to replace my KAF8300 CCD and was pretty much set on the QHY268M.

I now think the larger full frame might be a better option.

The Atik looks like a really good quality product. The QHY600 claims some sort of K sensor? 

Based on what QHY state there are both consumer and industrial grade of sensors; and that QHY offer cameras with the industrial version.  They state that these are more robust, more stable and tend not to degrade as quickly.  However, without any empirical evidence then it is difficult to determine just what sort of difference this makes for astronomy.  If the degradation occurs with less number of activations in the consumer version then this might be less of an issue with deep sky astronomy (with longer and less exposures).  A difference of 1million vs 2 million exposures is relatively insignificant (and even more so in the UK!).  On the other hand if you are taking a lot of short exposures (lucky imaging / planetary imaging) then it might be more of an issue.  CMOS hasn't really been around long enough for user experience to really come into play here.  On the other hand you might be taking a risk that the consumer version degrades quite quickly (especially at the cost and a small extra outlay for the industrial version could be worth the risk).

In terms of moving up to larger CMOS do take into account the extra grunt you need in your PC to process the images.  Each image of the full frame CMOS at the moment are about 120MB so if you are taking lots of short frames then that is going to take time to calibrate, combine and process.

On an aside if you aren't happy with the resolution of the KAF8300 you can always drizzle the images.  With the fast telescope you have that would also mitigate the additional noise this introduces.

21 hours ago, Adam J said:

Personally I would go with (in comparison to upcoming ASI2600mm or even the 294 for that matter):

1) Higher QE

2) 3x the surface area (larger sensor)

3) Higher dynamic range

4) Faster read times

5) larger full well capacity

6) Less fragile (one cosmic ray hit will not take out an entire column).

7) No dark frames required (IMX571 yes I have a friend who just dithers it) its actually cleaner than the CCD we are talking about here (provably so if you make me)

8 ) No amp glow.

 

So ill ask you for the IMX571 vs ICX694:

You say calibration is cleaner, can you prove it?

You say dark frames are more consistant, can you prove it?

You say CCDs cool better...can you prove it?

 

Diffraction Limited have a good introduction on the pros and cons of each choice (it also saves me waffling on for ages more than I do below)

https://diffractionlimited.com/ccd-versus-cmos-better/

In terms of the points noted

1) Yes some of them have better higher QE.  This isn't intrinsically linked to CMOS or CCD as you can get very high sensitive CCD as well.  Considering a balanced comparison of front and back illuminated sensors from Sony they are pretty equivalent.  The disadvantage is that back illuminated CCDs never entered the consumer market.
2) To an extent though the 6200 is similarly priced to the 16200's or 11000's anyway.
3) To an extent, yes but this is due to CMOS being user configurable.  CCDs can reach these levels but they are generally configured at the factory to balance noise/gain/well depth etc. It is relatively straight forward to get higher dynamic range but you have to sacrifice something (and the same goes for CMOS, it is just the users choice here)
4) Definitely faster read times.  Rather than one column/row that everything is read through CMOS can read each column simultaneously (although that brings noise issues - see below)
5) Again this largely based on what the consumer choice on the settings.  CCDs can have just as much well depth (if not more), but those without anitblooming gates could easily reach 100,000+
6) This is a bit of an urban myth.  Although it can happen it tends to be that very few cosmic ray hits will complete saturate and destroy a pixel (although I have seen it with some space telescopes).  For most users it damages the pixel so that the sensor always see it as having a residual charge which then feeds down the column or row so it is warm/hot but not completely useless.  In comparison CMOS seems to more generally degrade over time (especially the consumer versions) but there seems to be little evidence of how long this might take and/or based on number of exposures.  However they haven't been used widely for a while but will be interesting to see how these assertions pan out.
7) Yes, dithering averages out noise characteristics but doesn't remove the noise.  A lot is made of low noise characteristics, but really it is the calibrated frames that is important, but I'll come to this in a minute.
😎 No amp glow is the same for both cameras really so doesn't really matter (realistically it will calibrate out but with some cost in additional noise)

In terms of the questions, I can't prove it here because the CCDs are to be set up remotely and CMOS at home so can't really give you a direct comparison.  However, from a statistical perspective and the structure we can say some things.  Firstly the CMOS reads out from along all columns.  The columns each have their own electronics and it is impossible to get every column set exactly the same.  As such you do get column artefacts within CMOS (some worse than others).  Dithering and averaging out does help but it is residual noise that can't be fully calibrated out because it isn't truly random.  In comparison a CCD output (generally) discharges through one set of electronics so the noise artefacts nearly represent true random noise and hence can be subtracted to obtain the 'real' signal.  In comparison although you can make a proxy random noise character of the CMOS there is still a residual 'real' signal in the data which adds an element of noise to the calibration.  In addition there are structures within a CMOS that can be seen to some extent in the flats (for example see 10 minutes in this video):-

CCD and CMOS calibration files

Again these can be largely calibrated out.  At a very basic level though you have multiple non-random signals within the data.  Each of these signals has a small error within it and these add overall.  Again if you refer to the above video at about 4:40 you will see a similar effect in relation to subtracting amp glow (for the avoidance of doubt this is shown to prove the principle, not a discussion about amp glow as more modern CMOS are controlling this).  However from a principle perspective the noise where the amp glow was is still higher than other locations.  This is not an example of poor correction but a natural consequence of subtracting a non random signal from itself and leaving a higher overall level of error.  As such the more 'real' signal sources you have the more errors you compound.  Ultimately there are more 'real' signals in a CMOS camera that you have to remove.  Hence although by pixel noise levels can be lower, the calibrated signal can be worse and to an extent also more dynamic which doesn't favour being able to detect the faintest of objects.  You can never be quite sure whether it is real or not.  In comparison a CCD might have to expose longer but you can be more sure that the faintest objects (by flux) are real because the calibration is cleaner.

Finally as to whether CCDs can be cooled further, yes they can.  I've regularly cooled CCDs with liquid nitrogen with no issues. CMOS are more sensitive because the electronics are in the individual pixels and generate a lot of heat within the pixels (though I'm not an electronics engineer so the exact issue is not my forte.

From a personal perspective having used both, my preference at the current time is CCD over CMOS.  It isn't just just the stability of the noise but also most are manufactured in China and don't currently wish to support what I think is a regime with questionable environmental, social and democratic approaches (which isn't open for discussion).  I also worry that if lifespan of CMOS is relatively short then it encourages a greater throw away society and I don't support that approach for the future of my niece and nephews. Nevertheless in 5 years it may be a different question.

3 hours ago, Adam J said:

They are both Sony sensors. One is just vastly more capable than the other. The ASI6200mm pro is the same technology and it's talking images on astrobin that a 8 year old CCD can only dream of. 

It's not as simple as this.  There are pro's and cons to each type of camera.  Professional observatories are still sticking with CCDs, the specialist companies are still manufacturing them etc.  Yes, they are fading in the consumer market but that is because of factors outside the amateur astronomer 'region of interest'.

CCDs still have much more stable bias and dark frames because of the way they read out.  So for photometry and spectroscopy they are still the go to choice.  The calibration hence tends to be easier and cleaner overall.  CCDs also tend to cool better.  On the other hand you need to expose longer so need better mount and guiding, whereas you can get away with shorter exposures with CMOS and stack lots - although 6200MM has individual images of around 120MB so processing also takes a lot longer and you need a hefty computer.  Binning pixels on CCDs can be done at the hardware level (whereas CMOS is restricted to the software level so the noise reduction is a lot lower).  QHY also claim that consumer grade sensors degrade over time (but might be hype to sell the industrial version).  On the other hand damage to CCDs from cosmic ray hits can result in bad columns (but which can be averaged out).  

So there is no right or wrong choice.  Both have strengths and weaknesses but they can also both take great pictures.

56 minutes ago, gorann said:

That is what I have done - moved to a Bortle 2-3 sky by buying a small farm on the countryside. To compensate for cloudy nights in northern Europe I have over the last years built myself three obsies so I can optimise my efforts on clear nights (although so far I have limited myself to running two obsies in a single night). As many of us are going to work largely from home even after corona I think this is a way to go if your line of work allows it.

Unfortunately this isn't really an option in the UK as land is prohibitively expensive in most areas of England (barring random places in Scotland - and countryside houses are booming at the moment due to COVID).  

It would also be worthwhile seeing the breakdown by total entries and by individuals.  If multiple entries are allowed one particular prevalent imager continued to post many images then it is possible this distorts the figures somewhat.  There is always a risk that bias can creeps in as well from those considering the images (and it may well be sub-conscious).  This way you can ensure assessments are balanced (and it is always worth checking even if it turns out there is none). 

9 hours ago, Robny said:

Hi All

Does anyone use these: https://www.pixelskiesastro.com/remote-telescope-hosting-spain#

As far as I can tell, its a company that will host your own telescope (in spain) so you can use it remotely.  The fact that it doesn't state any type of guide price on their website, indicates to me that I cannot afford it 😂

But I'm just wondering how far out of reach it might be?  As its an interesting concept/thought given our perma-cloud situation 😬

Need to do a bit of research before bringing it to the attention of the finance director

Rob

They are more than helpful and very patient.  Hopefully I will have a system there shortly (they already have cameras etc) but am waiting for the mount to ship out to them.  Eventually it will be 12" corrected reflector and FLT98 but the world is not being kind at the moment to the plans.  

I have my own equipment at home as well, so that scratches the observing itch, but it's not a permanent set up (and probably won't be for a while) and it always seems to be clear when I have important work areas to attend which is more than frustrating.

It depends on what you like to do, whether it is the technical challenge, or you want the imaging.  I've sort of got over the technical challenge and actually want to get some decent images for a change (rather than a few hours here and there).

20 hours ago, kirkster501 said:

I find imaging with anything than refractors exceptionally frustrating.  There are enough variables at play with DSO imaging without adding collimation to the mix as well.  0.5arcsec/p is going to be very challenging mate on your guiding in NW Europe. I have tried it and gave up. YMMV though and maybe you might fair better than me.

 

You'd probably be better to bin the pixels to something more manageable.  That way you are still getting the benefit of the aperture with 'less' guiding issues.

A more expensive option (about twice the price after taking into account the flattener and reducer) would be the Vixen FL55 which would have a comparable focal length.   There is also the new Radian Raptor from Optcorp which I think is based on the same optics of the Sharpstar (although apparently they only take the premium lenses).  However the latter is new and untested and would likely need to be imported.