Jump to content




Please help with my flats nightmare


Recommended Posts

Hi Experts,

I have been imaging for a year or two, gradually increasing my skill levels (or so I thought). I use a SkyWatcher ED100 on an HEQ5-Pro, and an old Canon 350D for imaging (unmodified). I'm also guiding with a QHY 5-II through the SW finder scope. Everything has been fairly ok until a week or so ago when I removed the SW LPF that I have been using. The individual subs appeared to have more detail in them with a 300s exposure than previously, and I thought I would be able to process out any light pollution using PixInsight. I use DSS to stack and combine (I really must get around to learning how PI does the integration). However it hasn't worked out like that.

I realised that having changed the imaging train, I would need to take new flats. However no matter how I take the flats (using a table over the end of the scope, using an LCD monitor showing white, or using the sky through a white t-shirt), I get disastrous stacks from DSS, badly over corrected with a lot of extraneous colour (see first attachment below). I assume it is down to the MasterFlat that DSS creates and uses. The MasterFlat used to look like the second attachment, but now look like the third attachment. 

Have I simply over exposed now that I don't have the LPF reducing the light getting to my camera, or have I tripped up another way? In case it helps, here's also a single light sub as the fourth attachment, and a single flat  as the fifth attachment that went on to produce the horror that is the masterflat with the red ring!

Any thoughts welcome - this is driving me nuts and there could be clear skies coming up!


Over corrected.JPG

Previous masterflat.JPG

Current masterflat.JPG

Single sub.JPG

Single flat.JPG

Link to comment
Share on other sites


When you took the new flats you would have had to make a change to the exposure time due to the removal of the LPF, were the three histogram peaks, corresponding to the R, G and B pixel groups of the camera still fully contained within the X axis of the histogram chart?

 For a DSLR the usual method for setting flats exposure is to adjust the exposure time so that the three peaks are positioned around the 2/3 distance towards the right hand side of the histogram. But this is not always so easy to achieve using various light sources since the colour of the light source will affect the relative positions of the peaks. Using a sky flat, the colour is predominantly blue and the blue peak will be widely separated from the red and green, using an artificial tungsten light source the red peak may be over to the right side, an LED panel may only show two peaks where the red and blue are overlapping precisely and the green is moved to the left etc etc, the various combinations are countless.

The important thing is that none of the visible peaks are pushed so far over to the right hand or the left hand side of the histogram so that the peaks are cut-off.

You should see one, two or three full peaks clearly visible and a valley on both sides of the histogram, adjust exposure time for the flats so that the histogram peaks are as near to the two thirds distance to the right as possible with no cut-off and a clear valley visible at both ends. this should give you a good flat exposure.

I see from the header of your posted flats single-frame that the exposure time was 180 seconds, assuming this was different to the last set of flats you took with the LPF fitted did you remember to shoot matching dark frames? At this length of exposure time matching darks will most certainly be needed but if 180 seconds was used this time does seem rather long unless the light source is really dim?

If you are not taking matching darks for the flats then you need to be using a bright enough light source for the flats that the exposure time is short, less than a second or two for a tungsten lamp or sky flat and use bias frames re-named as darks-for-flats to make a master dark specifically for the flats calibration group.

If you are using a laptop, iPad or LED draughting panel with variable dimming etc as the light source then an exposure time of around 5 seconds is required to avoid frequency banding in the flats, use neutral density absorbers between the light source and the camera to achieve the 2/3rd's histogram position (plain printer paper or neutral density lighting acetates are good for this) and use matching time dark frames.


Lastly, avoid taking flats in the daytime, light leaks around the focuser draw tube will introduce stray light into the system and distort the flats, if unavoidable work in a darkened room with a thick opaque cover over the telescope and camera and arrange that the the light from the light source is only entering the telescope and not spilling out everywhere else, cut-out a cardboard mask to fit around the telescope aperture that masks off the unused parts of the flats screen being used. Sky flats should only be done in the very short period of twilight just before dawn and just after sunset when the sky is still just glowing but the local surroundings are dark. Daytime sky flats with the sun above the horizon will not work properly if there is even a tiny bit of light entering the telescope where it shouldn't. At this time of year there is a window of around only 30 minutes either side of sun up or sun down where sky flats are a serious possibility.

And finally of course, make sure the DSLR viewfinder mask is fitted when taking all images, lights, flats, darks and bias. Depending on the model DSLR this will be a tiny clip-on cover, usually attached somewhere on the camera strap, or somewhere in the accessory pack otherwise it may be a mechanical mask inside the camera operated by a small lever next to the viewfinder aperture. Light may leak into the camera through the viewfinder port and contaminate the images when taking long-time exposures. This is not a problem when using the camera normally since your eye is covering the port but when the camera is on a tripod or mounted on a telescope the port is uncovered and this can affect automatic light metering in the daytime and allow light leaks around the frame edges. This is why most DSLR's include the mask and it should be used for astro photography.

Hope the above may point towards the problem.

Edited by Oddsocks
Link to comment
Share on other sites

1 hour ago, Oddsocks said:


When you took the new flats you would have had to make a change to the exposure time due to the removal of the LPF, were the three histogram peaks, corresponding to the R, G and B pixel groups of the camera still fully contained within the X axis of the histogram chart?


Hope the above may point towards the problem.

I took the flats using AV mode on the Canon as I have always done. The exposure length was 1/180s which is slightly shorter than previously but as expected (the LPF would have required a slightly longer exposure due to the restriction in light I assume). The histogram peaks were well within the histogram. I did take the flats in the daytime, but I have done that always in the past as well without any issues. I stopped using darks a year or so ago, since they seemed to make little if any difference once dithering in my guiding setup. 

Everything you have said is good advice, but I'm struggling to work out why the master flat is coming out so differently now. I suppose I might have to reintroduce the LPF and see if things work as they used to? I'm also wondering if I have a silly setting in DSS but for the life of me I can't find it!


Link to comment
Share on other sites

22 minutes ago, Stub Mandrel said:

I take my flats the way you do. Sometimes things go bad, like you have got, and retaking the flats improves things.

Has your focus shifted?

I've retried taking the flats many times and still get the same odd results. I'm pretty sure focus hasn't moved much - the flats looks very similar to previously. I'd like to think that a small shift in focus wouldn't produce the bizarre results (although in AP nothing would surprise me now).

Link to comment
Share on other sites

17 minutes ago, dph1nm said:

You have rembered to subtract the bias for the new master flat I take it? It certainly doesn't look right with that red background. I suspect DSS has got confused somewhere.


I've left it all to DSS - just providing it with lights, flats and bias. However I do agree that DSS has got confused! Is there any way to reset the settings? I've run out of things to try!

Link to comment
Share on other sites

I agree, I think DSS has somehow got the wrong settings or got confused. 

In your first picture, the dust bunnies have been made to look as though they are 3D pitted.  I had this happen to me once with DSS, and I never did find out why.   I no longer use DSS I use Astroart, but I now own a mono camera, so would not know how to use Astroart for a DSLR.

I wonder whether it would be a useful exercise for you to upload some light frames, some flats and some bias and let us guys try to stack it with whatever we use.  That way you know if it's DSS (or it's settings) that's the problem, and whether the captured data is in fact OK.


Link to comment
Share on other sites

1 hour ago, Stub Mandrel said:

Options -> Load -> Restore Default Settings


I did the restore default settings, and then went through the recommended settings. The one that I didn't recognise says:


  • You are using bias frames
  • >Set the black point to 0 to improve calibration

I wasn't sure whether I had set this before, so had a quick Google. That took me back to a four-year-old thread 

where "dph1nm" said:

  • "Using this (with bias frames) kills my Canon 1000D reductions (flats no longer work), so I always have this switched off."

And sure enough, switching this off has got my stacking almost back to normal. Now I think I have to take the flats in the right way since they aren't correcting 100%, but the bizarre colour correction is largely sorted out. :hello2:

Thanks for your help and suggestions!


Link to comment
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

  • Recently Browsing   0 members

    • No registered users viewing this page.
  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue. By using this site, you agree to our Terms of Use.