#1  
Old 12-08-2015, 03:16 PM
batema's Avatar
batema (Mark)
Registered User

batema is offline
 
Join Date: Jul 2006
Location: Sunshine Coast
Posts: 1,827
Sky flats

I wonder if anyone can offer suggestions for times for sky flats as I am going to have a crack at them tonight. Using a Tec 140 with flattener and SBIG 8300m with filters. I use MaximDl for my capture program but have no idea of a starting time? Maybe 3 seconds but I have no idea.
Reply With Quote
  #2  
Old 12-08-2015, 03:32 PM
Octane's Avatar
Octane (Humayun)
IIS Member #671

Octane is offline
 
Join Date: Dec 2005
Location: Canberra
Posts: 11,159
CCDCommander starts sky flats (dawn) when the Sun is -7 degrees below the horizon (from memory).

So, if you're doing evening sky flats, try about 30-minutes after sunset.

I'm pretty sure there's a sky flat plugin for MaxIm DL that you can get off the web.

H
Reply With Quote
  #3  
Old 12-08-2015, 03:33 PM
gregbradley's Avatar
gregbradley
Registered User

gregbradley is offline
 
Join Date: Feb 2006
Location: Sydney
Posts: 17,877
Hi Mark,

Its not a time so much as an ADU value. The ADU value is displayed usually in a box in CCDsoft. Not sure about Maxim.

I go for between 20,000 and 30,000. 24,000 I find works well. That depends on how bright the light is. If you put a white tshirt over the end of your TEC140 and point it to the opposite side of the sunset and then shoot for that value. It may be 8 seconds. You will see the ADU value fall as twilight progresses so you have to be organised to do all filters LRGB and Ha for example. Ha will take the longest. If you image a 2x2 binning for RGB you may shoot them as 2x2 (it is possible to use 1x1 flats for 2x2 lights in CCstack). The exposure time will be less for 2x2 as its more sensitive.

There may be a bit of trial and error. On an image that is vignetted you can get a flat that overcorrects as its too bright and you can the vignette areas reverse and become too bright! So that correct exposure is a bit touchy. I have also seen flats taken in too dim a light that give the right ADU number but seem pretty even whereas the same ADU level taken in brighter light shows up the vignetting really clearly. You want to be able to see that vignetting in the flat so it fan correct it in the image. Some setups hardly vignette at all - like your TEC140 and 8300 probably will have hardly any vignetting but start using the larger sensors and you see vignetting. Sometimes the vignetting seems extreme yet the flats correct it beautifully. Like magic really.
So if it gets too dark you've missed the boat as a dim flat even if the right ADU won't correct properly. So there is a slight aspect of experience here. But a decent bright flat that is not too bright will work nicely. One way is to trial and error. Get a light ready in CCstack or Maxim. Take one flat and apply it to the light and see what happens. I trial and error and make sure the flat, darks and bias's are actually cleaning up the image rather than just assume that its all good. Sometimes its not.

So create a file for your flats. I usually have the date in the file name so later I know which flats are meant to go with which lights. By the date.

In that file I create a file for each of the filters like Luminance 1x1, red 1x1 etc.

Take several but you probably won't have enough time to do 16 of each, twilight will fail before then depending on how fast your camera is at downloading.

I also take flats with the roll off roof closed and a white cloth over the end of the scope. I can take as many as I want and they are high quality.

Make a master flat using average combine (mean) and either subtract a master bias or not. I don't subtract master bias at this stage but do that when I am calibrating the images at that point. I found it worked better that way with one of my scopes that was sensitive to good flat fielding.

Best to make sure your camera CCD window is clean ( I use photographic lens cleaning clothes and a hurricane blower bulb to clean). Clean your camera and your filters carefully and this will get rid of dust donuts in images that flats handle. Flats assume that the camera is in the same position with regards to the filter wheel and in focus. Rotating around the scope may not be an issue as most scopes (in fact all I have used) are the same at different rotations. The flats are to remove uneven light and uneven CCD sensitivity (really the first as I don't think CCDs are uneven in illumination across the sensor or if they are its miniscule).

It takes about 15-25 minutes to do it. I increase the exposure time as I go to the next filter to compensate for the failing light and keep an eye on that ADU value and try to keep it much the same.

SBIG have even illumination shutters so there probably is no real minimum shutter speed to get an evenly illuminated image. But if you are using a camera that has a leaf type shutter you really need to have a minimum exposure of about 3-4 seconds otherwise you can see the shutter in the flat and it will be useless. After about 4 seconds the shutter effect is gone and the flat is evenly illuminated in the case of leaf shutters.

I know some sensors (KAI) can do video output and you can use that to create flats during the day as well. Richard Crisp has details on how to do that.

A decent set of flats can last a while as it should remain valid until you move something in your image train, or you open up your filters or they get dust on them from usage (best to clean them thoroughly and leave them alone in the filter wheel - not easy if you have a 5 position filter wheel and you also sometimes do narrowband which requires a change.
For this reason I would recommend a 7 or 8 position filter wheel so you don't need to be constantly changing filters which can lead to damage to the filters and dust spots or a set of flats that don't correct a nasty dust spot (they are hard to process out later).

Greg.

Last edited by gregbradley; 12-08-2015 at 03:43 PM.
Reply With Quote
  #4  
Old 12-08-2015, 03:41 PM
codemonkey's Avatar
codemonkey (Lee)
Lee "Wormsy" Borsboom

codemonkey is offline
 
Join Date: Jul 2013
Location: Kilcoy, QLD
Posts: 2,058
Hey Mark,

Not quite what you're asking and maybe not better, but different. I take flats using my laptop screen as a source of light.

I set my background to white, move applications out of the way/minimize them, and then set SGP to start capturing in 10 seconds. That gives me enough time to position the screen over my scope.

From there, I take 25x0.3s exposures. The exposure time will vary depending on your scope, your camera and screen brightness so you'd need to experiment a bit, but I am for the spike on the histogram to be about 1/3 of the way from the left.

Anyway, just food for thought.

Cheers,
Lee
Reply With Quote
  #5  
Old 12-08-2015, 03:41 PM
Octane's Avatar
Octane (Humayun)
IIS Member #671

Octane is offline
 
Join Date: Dec 2005
Location: Canberra
Posts: 11,159
You want to aim for about 33-50% saturation of your camera's sensor for flats:

saturation=well capacity/gain.

On my STL-11000M, it has 50,000 well capacity and 0.8 electron ADU gain:

50,000/0.8=62,500 ADU saturation point.

So, you'd aim for 20,625 to 31,250 ADU.

Substitute values for your 8300-sensor.

I worked out an optimal value, which was around the 42% mark for some reason. Either way, I've put my minimum and maximum ADU values in CCD Commander and it takes care of the rest.

H
Reply With Quote
  #6  
Old 12-08-2015, 04:17 PM
gregbradley's Avatar
gregbradley
Registered User

gregbradley is offline
 
Join Date: Feb 2006
Location: Sydney
Posts: 17,877
Humi that gain should be multiplied not divided but the ADU levels seem right.

Mark if you want to go super in depth then this paper by Richard Crisp whilst perhaps a bit overtechnical has some nice summarisations and principles in it:

http://www.narrowbandimaging.com/inc...3_expanded.pdf

Page 103 actually says the suggested ADU level (ADU= analogue to digital units = number of electrons).

Also the number of flats for 8300 sensor is 5-8 in an earlier page.

Another interesting point in the paper is if your RAW images are brighter then you need brighter flats.

Greg.
Reply With Quote
  #7  
Old 12-08-2015, 08:05 PM
batema's Avatar
batema (Mark)
Registered User

batema is offline
 
Join Date: Jul 2006
Location: Sunshine Coast
Posts: 1,827
Thankyou all and Greg you are a legend. Thankyou for your comprehensive reply. I will check out the link.
Reply With Quote
  #8  
Old 16-08-2015, 02:40 PM
SpaceNoob (Chris)
Atlas Observatory

SpaceNoob is offline
 
Join Date: May 2012
Location: Canberra
Posts: 268
Quote:
Originally Posted by gregbradley View Post
Humi that gain should be multiplied not divided but the ADU levels seem right.

Mark if you want to go super in depth then this paper by Richard Crisp whilst perhaps a bit overtechnical has some nice summarisations and principles in it:

http://www.narrowbandimaging.com/inc...3_expanded.pdf

Page 103 actually says the suggested ADU level (ADU= analogue to digital units = number of electrons).

Also the number of flats for 8300 sensor is 5-8 in an earlier page.

Another interesting point in the paper is if your RAW images are brighter then you need brighter flats.

Greg.

I found this guide quite amazing.

I went in fairly blind after using an 8300 and tried the same methodology with the 16803, it didn't work and the flats were completely under correcting. After reading through the guide and doing some analysis of my own system properties, then adjusting things I ended up with perfect flat fielding.
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 02:42 AM.

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