Jump to content

  •  

CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.

Photo

OAG Guide Camera Filter

  • Please log in to reply
20 replies to this topic

#1 hmaron

hmaron

    Vostok 1

  • -----
  • topic starter
  • Posts: 128
  • Joined: 01 Dec 2020

Posted 22 July 2021 - 12:34 PM

This has been touched on before. I am just curious to hear from those who have experimented with using a red, IR or any filter in front of their OAG camera to improve guiding. Have there been any successes or a sense of any benefits in guiding stability? Or, equally important, have you noted negative outcomes. 

 

 



#2 jerahian

jerahian

    Apollo

  • *****
  • Moderators
  • Posts: 1,468
  • Joined: 02 Aug 2018
  • Loc: Maine

Posted 22 July 2021 - 01:24 PM

This has been touched on before. I am just curious to hear from those who have experimented with using a red, IR or any filter in front of their OAG camera to improve guiding. Have there been any successes or a sense of any benefits in guiding stability? Or, equally important, have you noted negative outcomes. 

I've tried using the ZWO IR Pass filter on my ASI290MM Mini.  Aside from making it harder to find guide stars in my OAG, I saw no perceivable difference.  Caveat:  I did not do any controlled tests or any tabulated comparisons and such, just my gut feel as to how my system guided.



#3 hmaron

hmaron

    Vostok 1

  • -----
  • topic starter
  • Posts: 128
  • Joined: 01 Dec 2020

Posted 22 July 2021 - 02:31 PM

I've tried using the ZWO IR Pass filter on my ASI290MM Mini.  Aside from making it harder to find guide stars in my OAG, I saw no perceivable difference.  Caveat:  I did not do any controlled tests or any tabulated comparisons and such, just my gut feel as to how my system guided.

Thanks. That tradeoff is the issue and your experience is exactly what I would suspect. Yet the question intrigues me.  When I am in Seattle (the summers) the seeing is particularly bad so often and I'm looking for some magic. The best solution to guiding is a great mount that requires little. I fortunately have that for the most part.  But......I can never leave well enough alone.



#4 jerahian

jerahian

    Apollo

  • *****
  • Moderators
  • Posts: 1,468
  • Joined: 02 Aug 2018
  • Loc: Maine

Posted 22 July 2021 - 03:15 PM

Thanks. That tradeoff is the issue and your experience is exactly what I would suspect. Yet the question intrigues me.  When I am in Seattle (the summers) the seeing is particularly bad so often and I'm looking for some magic. The best solution to guiding is a great mount that requires little. I fortunately have that for the most part.  But......I can never leave well enough alone.

Yep, I feel your pain.  I'm on the equivalent but opposite side of the country from you, with equally bad seeing.  My best nights are listed on Astrospheric as Average, and are generally, Below Average :/

 

CS, Ara



#5 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 10,628
  • Joined: 12 Oct 2007
  • Loc: Melbourne, Australia

Posted 22 July 2021 - 06:52 PM

I'm not aware of any demonstrated benefit from filtering a guidestar - and there are good reasons not to expect any benefit with IR-only, because the slight improvement in centroid stability is countered by a corresponding increase in the star size due to the longer wavelength.  This is known as "The Infrared Myth" : http://articles.adsa...000587.000.html

 

At the same time, I think everyone knows there are benefits in guiding on a star with strong signal and high SNR - so that sensor noise doesn't contribute error in the centroid - causing the guider to "chase the sensor noise."

 

And I don't know why people feel a need to make corrections less often.  Mounts can respond very quickly to short guide pulses, and there is no reason to let time pass between corrections and allow the error to grow.

 

In my case I get fwhm's in the low 1 arc-sec range with 0.27" pixels and corrections every second - using a completely unfiltered guidestars with a cgx-l.  A recent result with minimal processing (no sharpening or smoothing - just global stretch) is here:

 

get.jpg?insecure

 

No need for high end mount, no need to minimize the number of corrections - and no filter so the guidestar is as bright as possible.

 

Frank



#6 ks__observer

ks__observer

    Surveyor 1

  • *****
  • Posts: 1,665
  • Joined: 28 Sep 2016
  • Loc: Long Island, New York

Posted 23 July 2021 - 07:14 AM

I'm not aware of any demonstrated benefit from filtering a guidestar - and there are good reasons not to expect any benefit with IR-only, because the slight improvement in centroid stability is countered by a corresponding increase in the star size due to the longer wavelength.  This is known as "The Infrared Myth" : http://articles.adsa...000587.000.html

 

At the same time, I think everyone knows there are benefits in guiding on a star with strong signal and high SNR - so that sensor noise doesn't contribute error in the centroid - causing the guider to "chase the sensor noise."

 

And I don't know why people feel a need to make corrections less often.  Mounts can respond very quickly to short guide pulses, and there is no reason to let time pass between corrections and allow the error to grow.

If you don't use any filter than presumably your star will be bloated by infrared in any event -- so you might as well get the centroid benefit as well with the IR filter.

Guiding is supposed to counter mechanical imperfections -- so, obviously, you don't want to chase noise and you don't want to chase seeing.

I guess depending on how many guide-stars you have in your guide-camera FOV, if you have plenty of stars with strong signal, than noise would not be an issue -- find a star just below the saturation point.

I have not studied the article in detail re the "infrared myth."  Yes -- the sky is distorting the signal even if we don't detect it in infrared.  But guiding is supposed to detect mechanical imperfections -- so reducing the seeing factor with more infrared should help.  I would also think a larger star might help with centroid calculations.

 

Re guide-camera exposure time: 

I think the alternative theory, which I think the ONAG developer presented a while back, was that guiding too short will compound short term errors.  Just to present a counter viewpoint.



#7 Benschop

Benschop

    Ranger 4

  • *****
  • Posts: 320
  • Joined: 12 Jan 2015
  • Loc: Eastern Ontario

Posted 23 July 2021 - 09:15 AM

Defocusing the guide camera would be an alternative to filtering in "less than ideal" seeing. The guiding software does centroid calculations, so the focus is not extremely critical. The guide star is "larger" or occupies more pixels on the guidecam sensor when defocused.



#8 rgsalinger

rgsalinger

    Cosmos

  • *****
  • Moderators
  • Posts: 8,834
  • Joined: 19 Feb 2007
  • Loc: Carlsbad Ca

Posted 23 July 2021 - 10:35 AM

All that defocusing or filtering is going to do is to make it harder to get a good SNR on your guide star. It means that in places with sparse stars, you may even have trouble finding any. That's particularly true if you are using a long focal  length scope and/or an off axis guiding system. Centroid calculations (my experience) in practice are simply not as good as the various papers make them out to be. However, you can certainly calculate it to within a fifth of a pixel. If you want more guide stars upgrading the camera or the scope is the way to go. 

 

As far as guiding cadence is concerned, my experience is that the better the mount the less the need for fast corrections. My original iEQ45 preferred 1 - 2 second corrections to give the best guiding. My current CEM120EC2 will oscillate (some kind of encoder issue) if you correct it more than 1 a second. The other expensive mounts I've used also seem to prefer slow corrections. My two Paramounts work best in the 3-8 second range. I think that you need to work out the right cadence for your mount.

 

Note also that you don't have to use the exposure length as your cadence. All three of my systems have a built in 1 second delay. That way I can take short exposures without "overguiding". Without that, using a big scope I get tons (technical term) of saturated guide stars. YMMV.

 

I'm intrigued by multi star guiding but the one time I tried it the numbers (results) seemed no different from what I got using a single star. So, that was that. I did not trust the guiding numbers - I trusted the FWHM and eccentricity measurements of th actual stars in the images to tell the tale.

 

 

Rgrds-Ross 



#9 t-ara-fan

t-ara-fan

    Surveyor 1

  • -----
  • Posts: 1,946
  • Joined: 20 Sep 2017
  • Loc: 50° 13' N

Posted 23 July 2021 - 04:30 PM

An IR filter is supposed to give a more stable image, because the IR light is refracted less by a turbulent atmosphere. So in theory there is less seeing jitter in the location of the guide star. 

 

Using multi-star guiding in PHD2 will also reduce the random jitter due to seeing, because a number (10?) of stars have the jitter averaged.  The latter seems easier to do.



#10 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 10,628
  • Joined: 12 Oct 2007
  • Loc: Melbourne, Australia

Posted 23 July 2021 - 06:40 PM

If you don't use any filter than presumably your star will be bloated by infrared in any event -- so you might as well get the centroid benefit as well with the IR filter.

Guiding is supposed to counter mechanical imperfections -- so, obviously, you don't want to chase noise and you don't want to chase seeing.

I guess depending on how many guide-stars you have in your guide-camera FOV, if you have plenty of stars with strong signal, than noise would not be an issue -- find a star just below the saturation point.

I have not studied the article in detail re the "infrared myth."  Yes -- the sky is distorting the signal even if we don't detect it in infrared.  But guiding is supposed to detect mechanical imperfections -- so reducing the seeing factor with more infrared should help.  I would also think a larger star might help with centroid calculations.

 

Re guide-camera exposure time: 

I think the alternative theory, which I think the ONAG developer presented a while back, was that guiding too short will compound short term errors.  Just to present a counter viewpoint.

All wavelengths are perfectly fine because the scale of the seeing effects is down at the diffraction level and much smaller than a typical guidestar spot.  If you actually want to image the Airy pattern I recommend a red or IR filter because the pattern will be larger and more steady - but that assumes you are doing a form of lucky imaging with aligned and stacked short exposures - and the result is diffraction limited.  For guidestar spots with 1s or more exposure the spots may be 2" fwhm or so - and there will be more impact from chromatic aberration than seeing or anything else.

 

It's just unfortunate that the seeing theory was largely developed in the 60's and it led people to wrong conclusions about infrared behavior, but those misconceptions were clarified in the 70's.  Yet in CN there is a belief that infrared guiding will have some inherent benefit - when much of the reasoning behind it is a known myth, and it doesn't apply to guidestars in the first place.  And I haven't seen any stand-out results showing benefit from such guiding, while my own results with full spectrum and a bright guidestar are quite good - at just over 1" fwhm with a mid-range mount.  And I am using short guide exposures and correcting rapidly - since anything else makes absolutely no sense to me - and my results show that it works well.

 

There are so many conflicting attitudes about what makes a good guidestar.  Some people bemoan oag's because it is so hard to find a guidestar, while others want to use dark filters and throw away all those precious photons.  Others say the size and shape of the guidestar doesn't matter at all since the centroid will be perfectly fine either way (and I disagree!), while yet others expect a big benefit from multi-star guiding - which implies single star centroiding maybe isn't so great after all.

 

I don't expect to see any stand-out results from either IR guiding or multi-star.  For multi-star I see people happy about better *looking* guide plots - but in all this stuff I'm looking for stand-out results that indicate a given theoretical benefit actually shows as improved images with smaller fwhm.  And I haven't seen any.

 

Frank


Edited by freestar8n, 23 July 2021 - 06:46 PM.


#11 hmaron

hmaron

    Vostok 1

  • -----
  • topic starter
  • Posts: 128
  • Joined: 01 Dec 2020

Posted 24 July 2021 - 12:54 PM

This has all been very helpful. It speaks speaks that that virtually no one uses a filter for guiding. In the end, if there were discernible benefits we'd see it being utilized more often.   

 

Similarly, there is the question of using a reducer on the guide camera!?!? That too, while a compelling concept, is uncommonly implemented or lauded.



#12 ks__observer

ks__observer

    Surveyor 1

  • *****
  • Posts: 1,665
  • Joined: 28 Sep 2016
  • Loc: Long Island, New York

Posted 24 July 2021 - 01:30 PM

Similarly, there is the question of using a reducer on the guide camera!?!? That too, while a compelling concept, is uncommonly implemented or lauded.

For guidescopes, as opposed to an oag, the guidescope aperture is small, so the Airy Disk size is a function of f-ratio.

The smaller the f-ratio the less pixels accross the FWHM.

I would say a slower guidescope is better so you get bigger star for better centroid calc.

So i think an extender might be better.



#13 rgsalinger

rgsalinger

    Cosmos

  • *****
  • Moderators
  • Posts: 8,834
  • Joined: 19 Feb 2007
  • Loc: Carlsbad Ca

Posted 24 July 2021 - 05:11 PM

I've used a reducer on an OAG, not on a guide camera because the scope had a really long focal length and the guide camera we were using had a tiny chip. Worked great. Eventually we bought a better guide camera and we no longer use the reducer. I think it worked because the scope had a really big imaging circle. It might now work on other scopes, so YMMV.

Rgrds-Ross



#14 Oort Cloud

Oort Cloud

    Ranger 4

  • -----
  • Posts: 373
  • Joined: 19 Nov 2020
  • Loc: New Jersey, USA

Posted 24 July 2021 - 06:24 PM

I tried using an IR pass on my guidescope since it's cheap, and I figured would have a lot of CA. My reasoning was that with the 462mc having 100% QE in IR, and IR being less affected by seeing, that it would give more stable guidestars. I honestly can't say if it helped, as I did it early on durinng my initial testing with the guidescope, but it definitely worked. I was getting sub-arcsecond RMS values, and my scale for the imaging scope was 2.25"/pixel. On a camera less sensitive to IR though, I could easily see it being a hinderance. I'm probably going to leave it off moving forward though since I removed it for use with my SCT and OAG.

#15 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 10,628
  • Joined: 12 Oct 2007
  • Loc: Melbourne, Australia

Posted 24 July 2021 - 06:33 PM

I think a reducer won't hurt in most situations but there are several things going on, and it isn't simply turning a 10" f/8 guidestar into a 10" f/4 guidestar.

 

OAG optics are simple in principle but there are subtle things involved that can have a big impact on guidestar quality.  So a while ago I made a tool to visualize OAG geometry and the light loss involved.  Here is an example:

 

sample_oag_layout.png

A key thing to note here is that the cone of light usually doesn't fit into the prism - and all that light is lost.  The focal length is the same, but you have effectively lost aperture.  So an 8" f/8 scope may end up with a 6" f/10.4 guidestar.  You can then go ahead and add a reducer and it will make it faster, but with that same, smaller, aperture - so maybe it is a 6" f/8 guidestar.

 

But if you had used a slower scope in the first place, the light cone might fit into the prism and you could have an 8" f/10 scope with an 8" f/10 guidestar.  Although that may seem "slow" - it is full aperture and there won't be any distortions caused by having the cone clipped on the way to the sensor.  So that 8" f/10 guidestar may be perfectly fine - in terms of being bright and a good size - i.e. not too small and not too big on the scale of the pixels.

 

The key to getting the light cone to fit into the prism is to have the guide sensor as close to the prism as possible.  If you don't and the sensor is far away, you may be able to bring the sensor closer to the prism by inserting a reducing lens - but you won't gain any of the light lost entering the prism.

 

So I doubt that a reducing lens would really be beneficial if the oag is set up right in the first place.  And no matter what the oag should have a wide view of the prism - and a good size prism.  If the oag restricts light from the pupil, that is a big problem right there.  And on the topic of this thread - the last thing I would want to do if I have light loss, is insert a filter that throws away even more photons.

 

Frank


Edited by freestar8n, 24 July 2021 - 06:34 PM.


#16 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 10,628
  • Joined: 12 Oct 2007
  • Loc: Melbourne, Australia

Posted 24 July 2021 - 06:45 PM

I tried using an IR pass on my guidescope since it's cheap, and I figured would have a lot of CA. My reasoning was that with the 462mc having 100% QE in IR, and IR being less affected by seeing, that it would give more stable guidestars. I honestly can't say if it helped, as I did it early on durinng my initial testing with the guidescope, but it definitely worked. I was getting sub-arcsecond RMS values, and my scale for the imaging scope was 2.25"/pixel. On a camera less sensitive to IR though, I could easily see it being a hinderance. I'm probably going to leave it off moving forward though since I removed it for use with my SCT and OAG.

If you're just using a guidescope then you may have plenty of guidestars to choose from, and the main thing needed is a small star spot that allows a good centroid for calculation.  Typical guidescopes have a lot of chromatic aberration - so a filter may indeed help, but it wouldn't have anything to do with seeing effects and wavelength dependence.

 

So for any situation where you have plenty of bright guidestars and you know the optics have chromatic aberration, it would make sense to try different filters and see which ones end up with small and bright guidestar spots.  But many systems are optimized for color in the green - so it's likely a green filter would be a good choice.

 

Certain cameras may have good response in the IR - but refractors aren't usually optimized for IR performance, and most guidestars you see are emitting very strongly in the center of the visual band.

 

Frank



#17 Oort Cloud

Oort Cloud

    Ranger 4

  • -----
  • Posts: 373
  • Joined: 19 Nov 2020
  • Loc: New Jersey, USA

Posted 24 July 2021 - 07:40 PM

If you're just using a guidescope then you may have plenty of guidestars to choose from, and the main thing needed is a small star spot that allows a good centroid for calculation. Typical guidescopes have a lot of chromatic aberration - so a filter may indeed help, but it wouldn't have anything to do with seeing effects and wavelength dependence.

So for any situation where you have plenty of bright guidestars and you know the optics have chromatic aberration, it would make sense to try different filters and see which ones end up with small and bright guidestar spots. But many systems are optimized for color in the green - so it's likely a green filter would be a good choice.

Certain cameras may have good response in the IR - but refractors aren't usually optimized for IR performance, and most guidestars you see are emitting very strongly in the center of the visual band.

Frank


Thanks for the info...when I go back to the frac, maybe I'll try some color filters too...this particular camera is actually more sensitive to IR than visible light, so that was another reason I tried the IR filter. The thing I wasn't sure about was whether stars output more or less in IR than visible, so visible could still be a stronger signal coming into the guidescope but since the camera is color, the CFA would attenuate that significantly if it was a single color (more if R/B than G, obviously). In IR however, all 3 color pixels have 100% QE. This particular sensor really is a special case in that regard, based on the response curves published by ZWO.

#18 rgsalinger

rgsalinger

    Cosmos

  • *****
  • Moderators
  • Posts: 8,834
  • Joined: 19 Feb 2007
  • Loc: Carlsbad Ca

Posted 24 July 2021 - 07:52 PM

"QE peak: Over 80%" for the 462 in the IR region. That's from the OPT website. On the ZWO website, ZWO shows relative 100 percent. Relative to what they don't say in the manual. I'm sure it's a good camera but I really doubt that using it as a guide camera with a filter is going to do much for you. Would be interested in seeing the numbers. I'm just theorizing since I don't own one to test.  


  • freestar8n likes this

#19 Oort Cloud

Oort Cloud

    Ranger 4

  • -----
  • Posts: 373
  • Joined: 19 Nov 2020
  • Loc: New Jersey, USA

Posted 24 July 2021 - 08:08 PM

"QE peak: Over 80%" for the 462 in the IR region. That's from the OPT website. On the ZWO website, ZWO shows relative 100 percent. Relative to what they don't say in the manual. I'm sure it's a good camera but I really doubt that using it as a guide camera with a filter is going to do much for you. Would be interested in seeing the numbers. I'm just theorizing since I don't own one to test.

On the EQ6r, I get as low as .3-.4" RMS using it with OAG and 6" SCT (it's a struggle though; the tiny light cone on that scope and the square sensor of the 533 is a bad combination), and with the 32mm guidescope, it's consistently .8"-1" RMS. Overall, I'm pretty happy with it.

Edit: Not sure about the specs you read, I was just comparing the response curve graphs from various sensors...

Edited by Oort Cloud, 24 July 2021 - 08:10 PM.


#20 freestar8n

freestar8n

    Vendor - MetaGuide

  • *****
  • Vendors
  • Posts: 10,628
  • Joined: 12 Oct 2007
  • Loc: Melbourne, Australia

Posted 24 July 2021 - 09:02 PM

"QE peak: Over 80%" for the 462 in the IR region. That's from the OPT website. On the ZWO website, ZWO shows relative 100 percent. Relative to what they don't say in the manual. I'm sure it's a good camera but I really doubt that using it as a guide camera with a filter is going to do much for you. Would be interested in seeing the numbers. I'm just theorizing since I don't own one to test.  

Yes it's hard for me to tell just how responsive that sensor is - but it does sound like it is different and has higher response in the IR than other sensors. It's also very unusual that response in all channels is about the same in the IR, so it stops acting like an OSC camera there and becomes more mono.  This may be a good thing when guiding because the bayer array can mess with the centroid calculation - and that impact would be greater with a guidescope because each pixel is large in arc-sec compared to oag.

 

But at the same time most guidestars will be fainter in IR than toward the green because, although there are stars that predominantly emit in the IR, they are very faint and cool, so you only would see them as a guidestar if they are fairly close.  The main ones you see are hot and bright - and stronger in green than IR.

 

And there are absorption bands in the atmosphere that will lose light in IR.  Here is a spectrum of light from the sun received from the ground through the atmosphere, and you can see that it peaks in the visible and drops off in IR, with bands of low transmission.  Note that this plot specifically shows photon flux rather than Watts.  Watts is more commonly shown for blackbody plots - but what matters for a guide sensor is photon count - and the plot for watts is different from the plot for photon count.

 

Anyway - there are many things going on here, and for that particular sensor with a guidescope - some filtering may help - and IR may result in better centroid due to the sensor becoming mono.  But the IR benefit wouldn't be related to seeing - and I wouldn't use any filter with oag if the signal is faint.

 

Frank

 

The-photon-flux-spectrum-of-solar-radiat



#21 Oort Cloud

Oort Cloud

    Ranger 4

  • -----
  • Posts: 373
  • Joined: 19 Nov 2020
  • Loc: New Jersey, USA

Posted 24 July 2021 - 09:07 PM

Yes it's hard for me to tell just how responsive that sensor is - but it does sound like it is different and has higher response in the IR than other sensors. It's also very unusual that response in all channels is about the same in the IR, so it stops acting like an OSC camera there and becomes more mono. This may be a good thing when guiding because the bayer array can mess with the centroid calculation - and that impact would be greater with a guidescope because each pixel is large in arc-sec compared to oag.

But at the same time most guidestars will be fainter in IR than toward the green because, although there are stars that predominantly emit in the IR, they are very faint and cool, so you only would see them as a guidestar if they are fairly close. The main ones you see are hot and bright - and stronger in green than IR.

And there are absorption bands in the atmosphere that will lose light in IR. Here is a spectrum of light from the sun received from the ground through the atmosphere, and you can see that it peaks in the visible and drops off in IR, with bands of low transmission. Note that this plot specifically shows photon flux rather than Watts. Watts is more commonly shown for blackbody plots - but what matters for a guide sensor is photon count - and the plot for watts is different from the plot for photon count.

Anyway - there are many things going on here, and for that particular sensor with a guidescope - some filtering may help - and IR may result in better centroid due to the sensor becoming mono. But the IR benefit wouldn't be related to seeing - and I wouldn't use any filter with oag if the signal is faint.

Frank

The-photon-flux-spectrum-of-solar-radiat


Thanks for the thorough reply...FWIW, I did remove the filter for OAG use last night, since the scope is also a reflector, and I did notice brighter guidestars afterward.
  • freestar8n likes this


CNers have asked about a donation box for Cloudy Nights over the years, so here you go. Donation is not required by any means, so please enjoy your stay.


Recent Topics






Cloudy Nights LLC
Cloudy Nights Sponsor: Astronomics