#1  
Old 22-01-2012, 10:22 AM
pvelez's Avatar
pvelez (Pete)
Registered User

pvelez is offline
 
Join Date: Aug 2006
Location: Sydney
Posts: 1,250
LRGB - stray coloured pixels

With all the lousy weather in Sydney lately, I have been collecting data from the iTelescope (previously GRAS) service from New Mexico. Its a great system and lets me image in daylight (from Sydney anyway).

I have been collecting LRGB data with a Tak FSQ106 - all very nice.

However, after aligning and combining with Maxim, I find stray coloured pixels. In a dark field, there will be a single red or green pixel (rarely blue for some reason).

The frames are pre-calibrated. iTelescope also supplies the calibration frames and I had a go at calibrating the lights myself - same result.

Clearly there are a few pixels that show up as white in a R or G frame and not in any others. When combined and stretched I have a nasty brightly coloured pixel.

I suspect it is a function of my combine process. I use the Sigma Clip option with 3 passes and a sigma factor of 3.00. I have Ignore Black Pixels and Ignore pixels over 64,999 checked.

Any thoughts on what I am doing wrong?

Pete
Reply With Quote
  #2  
Old 22-01-2012, 02:33 PM
pvelez's Avatar
pvelez (Pete)
Registered User

pvelez is offline
 
Join Date: Aug 2006
Location: Sydney
Posts: 1,250
I've spent some time playing with this today. I can reduce the impact of the stray coloured pixels when combining using the HST option in Combine and setting the Luminence factor at 100%.

However, there seems to be a consistent mismatch between the R and G/B frames when combining. This results in heaps of colour noise. Have a look at the screenshot attached. This is after DD in Maxim and then a manual stretch in Photoshop. The red frames don't line up with the G/B (which appear aqua) - they seem shifted down and slightly to the right. This is odd as it appears in both Astrometric and Auto-Correlation alignment modes. More oddly (to me anyway) is that the stars display no shadowing - ie there is no difference in colour in the star images themselves.

This image comprises about 1 hour of L and 50 minutes each of R, G and B.

So what am I missing? Is it a processing issue or a combine issue?

I'd welcome any comments.

Pete
Attached Thumbnails
Click for full-size image (Screen shot 2012-01-22 at 3.26.07 PM.jpg)
65.9 KB42 views
Reply With Quote
  #3  
Old 22-01-2012, 02:41 PM
pvelez's Avatar
pvelez (Pete)
Registered User

pvelez is offline
 
Join Date: Aug 2006
Location: Sydney
Posts: 1,250
Attached is a close up which I think illustrates the issue well.

I can deal with the darker areas - its the nebulosity that I have no idea how to process.

I'm mighty confused.

Pete
Attached Thumbnails
Click for full-size image (Screen shot 2012-01-22 at 3.36.16 PM.png)
134.0 KB34 views
Reply With Quote
  #4  
Old 22-01-2012, 05:24 PM
avandonk's Avatar
avandonk
avandonk

avandonk is offline
 
Join Date: Aug 2005
Location: Melbourne
Posts: 4,786
Pete are you dithering your subs? This really helps to eradicate any sensor irregularities.

The problem with very good tracking is the noise in your sensor stacks beautifully to enhance it!

See here

http://www.iceinspace.com.au/forum/s...ight=dithering



Bert
Reply With Quote
  #5  
Old 22-01-2012, 06:23 PM
pvelez's Avatar
pvelez (Pete)
Registered User

pvelez is offline
 
Join Date: Aug 2006
Location: Sydney
Posts: 1,250
Thanks Bert

I use dithering as a matter of course with my own images - I hadn't found the dither alternative in the iTelescope planner till this afternoon.

Your post was quite clear - now I know what to do. A pity I can't resurrect what I have already

Pete
Reply With Quote
  #6  
Old 22-01-2012, 10:13 PM
OzRob's Avatar
OzRob (Rob)
Registered User

OzRob is offline
 
Join Date: May 2009
Location: Thailand
Posts: 446
It looks like you have the answer already but I have an unrelated suggestions.

If you are interested in continuing using these type of services you may want to check out Lightbuckets. They have gone through a bit of an upheaval recently but I found them a lot easier to deal with and more responsive when there is a problem. As an added advantage they are cheaper and you do no have to be on a plan. For example they have a CDK 17 available at $55 per hour. The same scope/camera on GRAS/iTelescope setup is $145 on the lowest plan.

This is the first I have heard of GRAS changing their name. I wonder is there is a law suit on the horizon from a certain phone maker....lol
Reply With Quote
  #7  
Old 23-01-2012, 07:02 AM
rat156's Avatar
rat156
Registered User

rat156 is offline
 
Join Date: Aug 2005
Location: Melbourne
Posts: 1,694
Hi Pete,

Another way around this problem is to use the "reject hot/cold pixels" command in CCDStack on each of your subs, then you have to interpolate the rejected pixels (0.2, 3 iterations). I find this cleans up most of the strays, then a little PS work later fixed the few that survive.

Cheers
Stuart
Reply With Quote
  #8  
Old 23-01-2012, 09:15 AM
Peter Ward's Avatar
Peter Ward
Galaxy hitchhiking guide

Peter Ward is offline
 
Join Date: Dec 2007
Location: The Shire
Posts: 8,090
Looks like RBI to me, as the noise is "paired" (ie red dot has corresponding green dot).

In short... nothing you are doing, but something Lightbuckets need to fix.
Reply With Quote
  #9  
Old 23-01-2012, 09:43 AM
pvelez's Avatar
pvelez (Pete)
Registered User

pvelez is offline
 
Join Date: Aug 2006
Location: Sydney
Posts: 1,250
Thanks Peter

what is RBI? I'm not familiar with that expression.

Peter
Reply With Quote
  #10  
Old 23-01-2012, 09:44 AM
pvelez's Avatar
pvelez (Pete)
Registered User

pvelez is offline
 
Join Date: Aug 2006
Location: Sydney
Posts: 1,250
Quote:
Originally Posted by rat156 View Post
Hi Pete,

Another way around this problem is to use the "reject hot/cold pixels" command in CCDStack on each of your subs, then you have to interpolate the rejected pixels (0.2, 3 iterations). I find this cleans up most of the strays, then a little PS work later fixed the few that survive.

Cheers
Stuart
Cheers Stuart

I haven't used CCD Stack before - it will be worth it if I can save the data I have.

Pete
Reply With Quote
  #11  
Old 23-01-2012, 09:52 AM
pvelez's Avatar
pvelez (Pete)
Registered User

pvelez is offline
 
Join Date: Aug 2006
Location: Sydney
Posts: 1,250
Quote:
Originally Posted by pvelez View Post
Thanks Peter

what is RBI? I'm not familiar with that expression.

Peter
Disregard this - I did my homework - a search through old threads which had the answer.

Looks like its worth following up with iTelescope.

Pete
Reply With Quote
  #12  
Old 23-01-2012, 11:10 AM
Paul Haese's Avatar
Paul Haese
Registered User

Paul Haese is offline
 
Join Date: Jan 2009
Location: Adelaide
Posts: 9,937
This looks like poor calibration to me. LB maybe need to sort out their calibration files. I sometimes get this every 6 months or so when calibration files need to be redone.
Reply With Quote
  #13  
Old 23-01-2012, 11:32 AM
pvelez's Avatar
pvelez (Pete)
Registered User

pvelez is offline
 
Join Date: Aug 2006
Location: Sydney
Posts: 1,250
Quote:
Originally Posted by Paul Haese View Post
This looks like poor calibration to me. LB maybe need to sort out their calibration files. I sometimes get this every 6 months or so when calibration files need to be redone.
Thanks paul

I've sent a query off - will let you know what the response is.

Pete
Reply With Quote
  #14  
Old 26-01-2012, 01:04 PM
RobF's Avatar
RobF (Rob)
Mostly harmless...

RobF is offline
 
Join Date: Jul 2008
Location: Brisbane, Australia
Posts: 5,716
Like Stuart says, if the problem pixels are relatively bright, you may be able to get good software to pull them out based on brightness and size criteria. (Pixinsight has a nice function for this that does a reasonable job).
Reply With Quote
  #15  
Old 27-01-2012, 07:09 AM
gregbradley's Avatar
gregbradley
Registered User

gregbradley is offline
 
Join Date: Feb 2006
Location: Sydney
Posts: 17,877
What was the equipment you were using?

At first glance it looks a bit like the images are not aligned properly.

Median combine is the best method for combine as it will get rid of many artifacts.

Sigma clip may be fine if you have enough subs for the maths to work. It assumes quite a few subs being worked over.

Greg.
Reply With Quote
  #16  
Old 27-01-2012, 07:18 AM
pvelez's Avatar
pvelez (Pete)
Registered User

pvelez is offline
 
Join Date: Aug 2006
Location: Sydney
Posts: 1,250
Quote:
Originally Posted by gregbradley View Post
What was the equipment you were using?

At first glance it looks a bit like the images are not aligned properly.

Median combine is the best method for combine as it will get rid of many artifacts.

Sigma clip may be fine if you have enough subs for the maths to work. It assumes quite a few subs being worked over.

Greg.
Greg

the scope was an FSQ106 and CCD was a SBIG STL11000M.

The odd thing is that the stars are well aligned - I tried both Astrometric and Star Matching in Maxim - it doesn't show up well in the screenshots I posted but the stars are fine, its the imperfections in the sensor response to the R vs G/B frames that don't match. I also tried combining with median, sum and average for the same result.

Word from iTelescope is that the darks may be as old as October - they plan on redoping them and having another crack at calibration with new master darks. I'm hoping that sorts it.

Pete
Reply With Quote
  #17  
Old 27-01-2012, 05:07 PM
gregbradley's Avatar
gregbradley
Registered User

gregbradley is offline
 
Join Date: Feb 2006
Location: Sydney
Posts: 17,877
I'd be interested in finding out how it goes with the new darks. I wouldn't've thought darks from October particularly old. CCDs age but not that quickly.

It'd be interesting to see.

Greg.
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 01:25 AM.

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