#1  
Old 13-12-2011, 12:26 PM
RobF's Avatar
RobF (Rob)
Mostly harmless...

RobF is offline
 
Join Date: Jul 2008
Location: Brisbane, Australia
Posts: 5,716
Flats - can I resize binnedx2 flats?

I've made a concerted effort lately to sort out my flats for the QHY9, where the main issue is getting decent ADU counts (no higher than say 23000 usually for this camera I believe) for each filter while keeping exposure time >3secs to avoid shutter artefacts.

When I go to shoot flats for binnedx2 the exposures need to be quite a bit shorter however, potentially less than 3 secs as I can't dim my light box enough. I've thought about adding layers of T Shirt or changing wiring, but is it really going to matter if I shot all my flats binnedx2 (assuming I made box dim enough) then resized them up for unbinned work? Surely the vignetting change across the FOV is fairly smooth?
Reply With Quote
  #2  
Old 13-12-2011, 12:51 PM
Peter.M's Avatar
Peter.M
Registered User

Peter.M is offline
 
Join Date: Sep 2011
Location: Adelaide
Posts: 947
In anticipation of buying my new camera I have been thinking about everything that might come up when I first get to use it. I was going to ask if people use one set of flats for each filter used? or shoot one set of flats per night and use it for all the exposures. I think that vignetting would be constant across the sensor, but I would have thought that downsizing the 1X1 bin to 2X2 would produce a better result. In my experience upsizing images will lead to pixelation which may cause bad corrections.

The only bad thing about doing it this way that I can see is you do not account for any dust on individual filters. But depending on how clean they are it may be quicker.
Reply With Quote
  #3  
Old 13-12-2011, 02:05 PM
RobF's Avatar
RobF (Rob)
Mostly harmless...

RobF is offline
 
Join Date: Jul 2008
Location: Brisbane, Australia
Posts: 5,716
I seem to see more defined vignetting on some filters Peter, which I don't really understand to be honest. Again, would be interested in people's thoughts on flats for each filter. The problem with 1x1 flats, especially if you start doing >10 for multiple filters is the 30 odd sec read/download time. 2x2 come down very quickly. I don't relish 20-30 mins of shooting flats last thing before packing up (as well as waiting for a proper sensor warm up time), lazy sod I am
Reply With Quote
  #4  
Old 14-12-2011, 06:16 AM
leon's Avatar
leon
Registered User

leon is offline
 
Join Date: Apr 2006
Location: Warrnambool
Posts: 12,430
Why on Earth would you shoot flats for 20-30 minutes, when the best time would be 30 secs at tops for the lot.

Leon
Reply With Quote
  #5  
Old 14-12-2011, 06:44 AM
gregbradley's Avatar
gregbradley
Registered User

gregbradley is offline
 
Join Date: Feb 2006
Location: Sydney
Posts: 17,877
Leon he means how long he spent taking all the flats not each individual exposure.

I just tried it - yes you can but the results are not quite as good as doing a proper match.

Using 1x1 flats on 2x2 images works too I am pretty sure.

Exact matches get the best results for both flats and darks.

Make sure your flats are flat dark subtracted.

Greg.
Reply With Quote
  #6  
Old 14-12-2011, 06:47 AM
[1ponders]'s Avatar
[1ponders] (Paul)
Retired, damn no pension

[1ponders] is offline
 
Join Date: Nov 2004
Location: Obi Obi, Qld
Posts: 18,778
Sorry Rob but no can do. Flats need to be taken at the same binning and with the same filter as the lights. The time for flats though can be very short (some say less than a sec, others recommend up to 10 sec and then take flat darks) not like darks that need to be the same length as your lights (or close if you are taking bias frames and scaling).

One of the things flats do is to 'normalize' any quantum efficiencies between pixel across the frame. You can't do that if you take 4 pixels, make them one big pixel and then try to split that again back into 4 pixels.
Reply With Quote
  #7  
Old 14-12-2011, 07:12 AM
gregbradley's Avatar
gregbradley
Registered User

gregbradley is offline
 
Join Date: Feb 2006
Location: Sydney
Posts: 17,877
Not quite true Paul.

I have actually done it and it works reasonably well. The same is true in reverse a 1x1 flat used on 2x2 images. In CCDstack the reverse works automatically where you can use 1x1 flats on 2x2 images and it works fine.

Vignetting is a gradual thing not sharp so there are no defined hard edges which is probably why it works.

Its not an exact match so you may be left with more pronounced gradients though.

I just did a test and it worked and it even took out a dust donut cleanly but there seemed to be a more pronounced gradient (the correctly flatted image also had a gradient it just seemed a bit worse with the wrong flat).

It could get you out of trouble if you missed taking flats but its not best practice.

If you don't feel like taking flats at night you can take them during the day.

I took flats the next day with a desert storm telescope cover over the scope and a white t-shirt in front of the lens.
It worked fine. I was surprised.

I usually take flats during the day with my observatory roof closed. The interior of the room is painted matt black and I install the
soft white cover for my CDK my wife made. I make sure the scope is not directly pointing at a bright area where the roof meets the walls
and flats work fabulously. Its very convenient.

Greg.
Reply With Quote
  #8  
Old 14-12-2011, 08:41 AM
[1ponders]'s Avatar
[1ponders] (Paul)
Retired, damn no pension

[1ponders] is offline
 
Join Date: Nov 2004
Location: Obi Obi, Qld
Posts: 18,778
Thanks for that Greg. Always nice to learn something new. I agree that its not best practice though. I wouldn't want to do it if I was doing anything more serious that general AP images. My comment re QE inequality across a chip would still hold true though for more scientific style imaging (photometry etc) though.
Reply With Quote
  #9  
Old 14-12-2011, 04:40 PM
leon's Avatar
leon
Registered User

leon is offline
 
Join Date: Apr 2006
Location: Warrnambool
Posts: 12,430
Opp's sorry, I probably should have read that a bit better, my apologies.

Leon
Reply With Quote
  #10  
Old 14-12-2011, 07:14 PM
gregbradley's Avatar
gregbradley
Registered User

gregbradley is offline
 
Join Date: Feb 2006
Location: Sydney
Posts: 17,877
Quote:
Originally Posted by [1ponders] View Post
Thanks for that Greg. Always nice to learn something new. I agree that its not best practice though. I wouldn't want to do it if I was doing anything more serious that general AP images. My comment re QE inequality across a chip would still hold true though for more scientific style imaging (photometry etc) though.
Yes I agree. Its more of a get out of jail free card.

With my CDK17 I need excellent flat procedures and gradient handling procedures and it would not be too good for that either. I suspect it makes gradient handling a bit harder as you have pointed out.

Greg.
Reply With Quote
  #11  
Old 14-12-2011, 09:27 PM
RobF's Avatar
RobF (Rob)
Mostly harmless...

RobF is offline
 
Join Date: Jul 2008
Location: Brisbane, Australia
Posts: 5,716
Thanks guys. Sometime you need someone to tell you to wake up and behave. I would have thought between bias and darks decent software would sort out between pixel differences, but good point that to be technically correct need to properly correct for this. I guess I was thinking more in terms of vignetting correction. It wasn't until I considered how much faster 2x2 shooting is I even started to consider it but probably silly to shoot hours of lights without decent matching cal frames.

I'm having more trouble than I thought dimming my light box even with a decent dimmer in the circuit now. May have to rewire with LEDs is the real issue.
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 03:35 PM.

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