Go Back   IceInSpace > Equipment > Astrophotography and Imaging Equipment and Discussions

Reply
 
Thread Tools Rate Thread
  #1  
Old 25-08-2015, 04:07 PM
rogerg's Avatar
rogerg (Roger)
Registered User

rogerg is offline
 
Join Date: Feb 2006
Location: Perth, Western Australia
Posts: 4,563
Question Dealing with inconsistency of LRGB frames vs OSC

Hi everyone,

So, as many of you know I don’t have the best results from LRGB imaging.

I currently have on loan a ST8300c camera and am very happy with the results. One of the things, probably the main thing, which is easier to use about the OSC camera is the consistency of image quality between RGB frames. For example, with my ST8-XME when I do 20xL + 10xRGB frames at 300s, the difference in time between frames means there’s variation in atmospheric conditions and focus (even if refocused regularly as I do) resulting in slight offsets of colour channels. I’ve found this very hard to overcome and produce nice images that for example don’t have a bloat of colour around stars. This problem doesn’t exist for the OSC because RGB frames are taken concurrently, even if they are later de-bayered for processing.

Of course the problem is less when at more lenient resolutions (2”/pixel for example vs 0.84”/pixel). At 0.84”/pixel it’s common that each re-focus will result in a very subtly different focus point and quality of focus, hence the variation I notice. At 500mm focus is always perfect and atmospheric variations so small as to not be noticeable, so there’s no variation.

In considering the decision of Mono vs OSC for my next camera, how do people mitigate this difference in image quality between channels with LRGB imaging to not have the dodgy results I achieve with the Mono+LRGB approach?

Thanks,
Roger.
Reply With Quote
  #2  
Old 25-08-2015, 04:25 PM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
 
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
it's a real problem. You might try to expose red at lower angles above the horizon, followed by green and then blue when high up - the seeing varies with altitude and wavelength and this approach can help get similar PSFs in the three channels. Otherwise, it is possible to adjust colours using a combination of deconvolution to tighten up channels or, if all else fails, blurring will bring tighter channels back to the norm.

Or, you could just shrug and say that that is what the data is like - and accept a bit of colour edging.
Reply With Quote
  #3  
Old 25-08-2015, 08:47 PM
PRejto's Avatar
PRejto (Peter)
Registered User

PRejto is offline
 
Join Date: Jan 2011
Location: Rylstone, NSW, Australia
Posts: 1,397
I've tried to use deconvolution to tighten up blue but the results are not very satisfying and I've given that up. But, recently on a different forum someone posted that the answer to blue bloat is just the opposite; namely one should "convolve" red and green to match blue. I have not tried this but I think it actually might work. I think shrinking the blue may actually remove the part of the blue (the halo) that actually shows and you don't end up with blue being better. But, perhaps convolving (enlarging??) red and green might mean the halo isn't so prominent and blends with the larger FWHM. Anyone ever tried this?

Peter
Reply With Quote
  #4  
Old 25-08-2015, 09:07 PM
Shiraz's Avatar
Shiraz (Ray)
Registered User

Shiraz is offline
 
Join Date: Apr 2010
Location: ardrossan south australia
Posts: 4,918
Quote:
Originally Posted by PRejto View Post
I've tried to use deconvolution to tighten up blue but the results are not very satisfying and I've given that up. But, recently on a different forum someone posted that the answer to blue bloat is just the opposite; namely one should "convolve" red and green to match blue. I have not tried this but I think it actually might work. I think shrinking the blue may actually remove the part of the blue (the halo) that actually shows and you don't end up with blue being better. But, perhaps convolving (enlarging??) red and green might mean the halo isn't so prominent and blends with the larger FWHM. Anyone ever tried this?

Peter
Hi Peter. convolution is the process used in simple filters - Gaussian blur for example is a convolution filter that will expand the FWHM in much the same way as seeing, so should be best if the channel mismatch is due to seeing variations. It works, but spreads the colour and reduces saturation on small regions such as Ha spots in a galaxy. What did you find went wrong with deconvolution? - applied to linear data, it should provide tighter stars of about the right profile.

for interest, tried morphological erosion to reduce blue stars with horrible results (a few green stars for example) - deconvolution on linear data was much better

Last edited by Shiraz; 26-08-2015 at 07:47 AM.
Reply With Quote
  #5  
Old 26-08-2015, 02:05 PM
rogerg's Avatar
rogerg (Roger)
Registered User

rogerg is offline
 
Join Date: Feb 2006
Location: Perth, Western Australia
Posts: 4,563
Well, I'm heartened that I'm not the only one with this problem, and it's perhaps even common

I haven't tried deconvolution, sounds like something I'll need to invest some time in to learn the tricks of, for using the ST8 in LRGB. Of course my problem with the ST8 is worsened by the fact it's NABG.
Reply With Quote
  #6  
Old 26-08-2015, 02:42 PM
alistairsam's Avatar
alistairsam
Registered User

alistairsam is offline
 
Join Date: Nov 2009
Location: Box Hill North, Vic
Posts: 1,837
I'm going with the approach of tandem imaging.
mono on a 10inch F4 and rgb with an osc on a WO81 piggybacked on the main scope .

haven't decided on the osc yet but with what I have in mind, the image scale and fov are similar.
yet to try it. hopefully in a week or two.

I'm guiding with an oag on the 10inch so hopefully guide errors shouldn't be too pronounced on the wider refractor .
im using losmandy dovetails and custom machined rings to keep diff flexure to a minimum.
else imaging red first works as well.

Cheers
Alistair
Reply With Quote
  #7  
Old 26-08-2015, 02:56 PM
rogerg's Avatar
rogerg (Roger)
Registered User

rogerg is offline
 
Join Date: Feb 2006
Location: Perth, Western Australia
Posts: 4,563
Quote:
Originally Posted by alistairsam View Post
I'm going with the approach of tandem imaging.
mono on a 10inch F4 and rgb with an osc on a WO81 piggybacked on the main scope .

haven't decided on the osc yet but with what I have in mind, the image scale and fov are similar.
yet to try it. hopefully in a week or two.
Yes I've often considered the same and currently this borrowed 8300c is on my Megrez 90 which is mounted atop my 12" SCT. If I were to buy a OSC I'm very tempted to go for the Sony ICX-814 chip (cameras such as the Starlight Xpress Trius SX-814), because of the chip's increased sensitivity and smaller pixels to suit the short 500mm focal length of my Megrez90 (with reducer/flattener incorporated).

anyway, we digress

Roger.
Reply With Quote
  #8  
Old 26-08-2015, 04:32 PM
PRejto's Avatar
PRejto (Peter)
Registered User

PRejto is offline
 
Join Date: Jan 2011
Location: Rylstone, NSW, Australia
Posts: 1,397
Quote:
Originally Posted by Shiraz View Post
What did you find went wrong with deconvolution? - applied to linear data, it should provide tighter stars of about the right profile.
Hi Ray,

Thanks for your comments. I was specifically trying to reduce blue bloat.. I applied decon in CCDStack on blue linear data to the point where the 3 RGB frames had about equal FWHM values. The result was not better.. It has been a while since I've tried it so I don't specifically remember exactly how/why it didn't work.

Peter
Reply With Quote
  #9  
Old 26-08-2015, 08:56 PM
gregbradley's Avatar
gregbradley
Registered User

gregbradley is offline
 
Join Date: Feb 2006
Location: Sydney
Posts: 17,877
A bit more data would help to analyse your problem Roger.

What LRGB filters did you use for mono filtered imaging? Not all filters are parfocal and not all filters are 1:1:1 combine ratio. This is one main reason I use Astrodon. If you focus on the luminance you will find your other RGB filters and even narrowband are also in focus. There may be some tiny difference but its negligible.

What camera are you using? As Peter has mentioned the Sony sensors are more than twice as sensitive than most Kodak sensors in O111 and do very well in blue. So chromatic aberration (which is often in the blue channel) will show up more than say with a less blue sensitive camera that uses a Kodak sensor.

Secondly what software do you use to callibrate and combine your images? Some combine methods work better than others.

Another point is what angles do you image at? I tend to image objects that are near the zenith or slightly to the east and then image them down to maybe 45 degrees. Below that most subs start to deteriorate in quality.

What scope is it? Does it have a corrector? Is there mirror shift? chromatic aberration?

How do you autoguide? Is there differential flexure which may get worse at certain angles?

How often do you refocus and how sensitive is your scope to temperature shifts. How much does your environment shift in temperature?

Are you imaging at a dark site or under light pollution? Is your seeing very variable or tends to be reasonably stable?

As you can see there could be several factors.

Greg.
Reply With Quote
  #10  
Old 27-08-2015, 09:21 AM
rogerg's Avatar
rogerg (Roger)
Registered User

rogerg is offline
 
Join Date: Feb 2006
Location: Perth, Western Australia
Posts: 4,563
Yes all fair points Greg, lots of factors involved and my aging equipment doesn't help

Quote:
Originally Posted by gregbradley View Post
What LRGB filters did you use for mono filtered imaging? Not all filters are parfocal and not all filters are 1:1:1 combine ratio. This is one main reason I use Astrodon. If you focus on the luminance you will find your other RGB filters and even narrowband are also in focus. There may be some tiny difference but its negligible.
I use parfocal Astronomik Type II (not typeIIc) so they are aging and I would love a set of 1:1:1 astrodons, my astronomik's are FAR from 1:1:1. This would simplify things a lot for me probably.

Quote:
Originally Posted by gregbradley View Post
What camera are you using?
ST8-XME. The NABG KAF sensor probably doesn't help things by being NABG becaus any blooming makes the problem all that much harder to deal with.

Quote:
Originally Posted by gregbradley View Post
Secondly what software do you use to callibrate and combine your images? Some combine methods work better than others.
MaximDL for calibration.

Quote:
Originally Posted by gregbradley View Post
Another point is what angles do you image at? I tend to image objects that are near the zenith or slightly to the east and then image them down to maybe 45 degrees. Below that most subs start to deteriorate in quality.
Same as you - eastern hemisphere, Zenith down to about 45 degrees.

Quote:
Originally Posted by gregbradley View Post
What scope is it? Does it have a corrector? Is there mirror shift? chromatic aberration?
Major problem #1 is probably the scope. Aging 12" SCT Meade. Stars are mushy. But, work with what I have.

Quote:
Originally Posted by gregbradley View Post
How do you autoguide? Is there differential flexure which may get worse at certain angles?
Typically don't autoguide, exposures 300s or shorter on my PME don't typically exhibit trailing.

Quote:
Originally Posted by gregbradley View Post
How often do you refocus and how sensitive is your scope to temperature shifts. How much does your environment shift in temperature?
Refocus every hour, but as I say, the exact accuracy of the focus varies and this is a problem. At 0.84"/pixel on the 12" I don't think I can get around this, there is always going to be some slight variance. Having said that, I should set FocusMax back to doing 3-5 image samples averaged, it might be on 1 image samples.

Quote:
Originally Posted by gregbradley View Post
Are you imaging at a dark site or under light pollution? Is your seeing very variable or tends to be reasonably stable?
Seeing is very variable, and this would feed in to the variability of the focus accuracy throughout any given night.

So, in summary, I'm aware of the source of much of the error, but the question was around how much of a problem this is for other people and if it is how they mitigate it, as it's rare for me to see the problem in other people's images. Seems deconvolution might go a long way towards helping me. I then wonder how much work this is, so will be interesting to try it and see how it goes. Of course the comparison is drawn because it's simply not a problem for the OSC.

Thanks,
Roger.
Reply With Quote
  #11  
Old 27-08-2015, 12:16 PM
marc4darkskies's Avatar
marc4darkskies (Marcus)
Billions and Billions ...

marc4darkskies is offline
 
Join Date: Mar 2007
Location: Quialigo, NSW
Posts: 3,141
Yep, differences in stellar profiles are almost inevitable if you image across different nights (seeing) or you have to refocus often or at a filter change. These days I don't have to refocus but of course seeing across multiple nights is still a pain.

My standard workflow includes zooming in on a non-saturated star that's as close to G2V as possible and decon'ing until the stellar profiles in R, G & B for that star are about the same. It's usually possible to spot G2V (or close to) stars in good images of the same field from other people - they should appear white. You can also wing it and examine your own raw RGB combine for non-saturated white-ish stars but that requires that you do a reasonably accurate combine. I've had luck doing my raw RGB combines by looking only at an isolated star field in the frame (ie zoomed in). Green stars are a no-no, background should be neutral and star colours should appear natural (eg red, white, blue and yellow). After you get that about right, decon to normalise on a white-ish star. The limitation of decon'ing of course is the amount of decon you need to do. If the seeing was particularly bad for one filter but not the others, decon artifacts may be unsightly.

Cheers, Marcus

PS: I've even been known to blur a sharp colour channel to achieve consistent stellar profiles since I usually layer a slightly blurred RGB layer to luminance anyway (in Photoshop).
Reply With Quote
Reply

Bookmarks

Thread Tools
Rate This Thread
Rate This Thread:

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +10. The time is now 06:19 AM.

Powered by vBulletin Version 3.8.7 | Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
Advertisement
Testar
Advertisement
Bintel
Advertisement