Page 1 of 1
Vignetting and Flat Frames
Posted: Sun Jun 28, 2020 8:29 am
by Graeme1858
The image below of M81 was stacked from 10 subs with 5 darks and 20 flat frames. It has been stretched in curves but not levelled, then exported.
Clipping the left hand side of the levels graph and cropping gets rid of most of the vignetting but it looks like the Flat frames could be improved.
The histogram shows the peaks in the right hand side of the graph so the frames are not under exposed. They're done in
APT with my
DSLR camera on Av mode.
I did think that taking the dew shield off before placing the light box could cause the vignetting problem but that wouldn't account for the dust bunnies that are still on show.
Regards
Graeme
Re: Vignetting and Flat Frames
Posted: Sun Jun 28, 2020 4:48 pm
by STEVE333
Hi Graeme - When taking flats you want to have the exact same optical setup as when taking the pictures. So, when taking the Flats, leave the Dew Shield in the same position it was in when taking the Light frames.
I found Flats to be tricky with a DSLR because you shoot all three color channels at the same time. Because all three channels respond differently to the "flats source" it can be difficult to achieve a good exposure for all three channels simultaneously. Just make sure none of the channels are saturated!
Another BIG thing. Because your sensor is CMOS you DO NOT want to use Bias frames in calibration. Try the following:
1) Take All the Flat frames at the same exposure time (Tflat).
2) Take Flat Darks at the same exposure time as the Flats (Tflat).
2a) Create a MasterFlatDark (by combining the Flat Darks).
3) Create a MasterFlat using only the MasterFlatDark (no Bias frames) for calibration.
4) Create a MasterDark by just combining the Dark frames. Do NOT use any Bias frames.
5) Calibrate the Light frames using the MasterFlat and MasterDark only.
This should produce excellent calibrated Light frames which can then be registered and combined in the usual way.
Hope this helps.
Steve
Re: Vignetting and Flat Frames
Posted: Sun Jun 28, 2020 8:28 pm
by Graeme1858
Thanks Steve
It occurred to me that the dew shield might be better left in place after I saw a problem. I'll do that next time.
My light box has a zig zag of 1m length of soft white LEDs. The histogram shows the three colours are at different intensity but none are saturated.
I've not used any Bias frames since you advised not to with a CMOS a little while ago! And after I posted this thread it occurred to me that you might have already answered this question in your previous thread! So, if you did, thanks for coming back again!
1) APT requires a Flat Frame Plan to have the camera on Av. It then captures frames at 1/25 sec or 1/30 sec and they do vary sometimes during the plan but only ever between these two settings. This then produces a histogram showing the three peaks at different levels but all between the middle and right sections. Should I instead be selecting a shutter speed?
2) I've never done flat darks! I'll have to get into that.
2a) Let DSS do that automatically? Or manually stack just the flat darks in DSS?
3) Ditto
4) Ditto
5) Understood
Thanks for your help Steve.
Regards
Graeme
Re: Vignetting and Flat Frames
Posted: Sun Jun 28, 2020 10:40 pm
by STEVE333
You are most welcome Graeme.
Graeme1858 wrote: ↑Sun Jun 28, 2020 8:28 pm
Thanks Steve
I've not used any Bias frames since you advised not to with a CMOS a little while ago! And after I posted this thread it occurred to me that you might have already answered this question in your previous thread! So, if you did, thanks for coming back again!
1) APT requires a Flat Frame Plan to have the camera on Av. It then captures frames at 1/25 sec or 1/30 sec and they do vary sometimes during the plan but only ever between these two settings. This then produces a histogram showing the three peaks at different levels but all between the middle and right sections. Should I instead be selecting a shutter speed?
Yes, select a shutter speed. You could let
APT determine the proper exposure time, then, setup a capture plan for Flats using only that exposure time.
Graeme1858 wrote: ↑Sun Jun 28, 2020 8:28 pm
2) I've never done flat darks! I'll have to get into that.
APT allows you to setup a plan for
Dark Flats. Just use the exact same exposure time as used for the Flats.
Graeme1858 wrote: ↑Sun Jun 28, 2020 8:28 pm
2a) Let DSS do that automatically? Or manually stack just the flat darks in DSS?
3) Ditto
4) Ditto
I think
DSS will do all of this automatically. I haven't used
DSS for some time now, so, I don't remember if there are any special settings required to make it do this.
Best of luck. If you don't mind, would you let me know how this works out?
Cheers,
Steve
Re: Vignetting and Flat Frames
Posted: Mon Jun 29, 2020 2:31 am
by Juno16
Very interesting!
Thanks Graeme for starting this thread. Thanks Steve [mention]STEVE333[/mention] for the interesting fact concerning dark flats instead of bias calibration frames for dslr’s. I’ve always shot and stacked bias calibration frames with my dslr. After reading your post #2 above, my eyes were wide open!
Steve, what could be the result of stacking bias calibration frames instead of dark flat frames?
Thanks Steve!
Re: Vignetting and Flat Frames
Posted: Mon Jun 29, 2020 3:09 am
by JayTee
Everyone,
I don't know if you all have seen the
DSS processing flow charts from the
DSS manual. I thought I would share them to generate a bit more discussion.
Cheers,
JT
Re: Vignetting and Flat Frames
Posted: Mon Jun 29, 2020 9:02 am
by STEVE333
Perfect JT - Thanks so much for sharing this.
Alternate Calibration Process II (no Bias frames) is exactly what is required for calibrating CMOS cameras.
That illustration makes it very clear.
Steve
Re: Vignetting and Flat Frames
Posted: Mon Jun 29, 2020 10:59 am
by Juno16
JayTee wrote: ↑Mon Jun 29, 2020 3:09 am
Everyone,
I don't know if you all have seen the DSS processing flow charts from the DSS manual. I thought I would share them to generate a bit more discussion.
Cheers,
JT
STEVE333 wrote: ↑Mon Jun 29, 2020 9:02 am
Perfect JT - Thanks so much for sharing this.
Alternate Calibration Process II (no Bias frames) is exactly what is required for calibrating CMOS cameras.
That illustration makes it very clear.
Steve
Very interesting gentlemen!
Perhaps this is why I see an unusual noise or gradient like pattern when my image is “overstretched” in
Startools. Kind of looks like vignetting but a different pattern. Once the region of interest is selected and the stretching is reduced, this pattern kind of disappears, but I am sure it is still there, just much less pronounced. I thought it was bad flats!
Thanks guys!
Re: Vignetting and Flat Frames
Posted: Mon Jun 29, 2020 4:07 pm
by Graeme1858
STEVE333 wrote: ↑Sun Jun 28, 2020 10:40 pm
Best of luck. If you don't mind, would you let me know how this works out?
Cheers,
Steve
Certainly will Steve, in fact I'll be posting up chapter and verse!
Thanks for your help.
Regards
Graeme
Re: Vignetting and Flat Frames
Posted: Tue Jun 30, 2020 3:05 am
by bobharmony
Juno16 wrote: ↑Mon Jun 29, 2020 2:31 am
Very interesting!
Thanks Graeme for starting this thread. Thanks Steve @STEVE333 for the interesting fact concerning dark flats instead of bias calibration frames for dslr’s. I’ve always shot and stacked bias calibration frames with my dslr. After reading your post #2 above, my eyes were wide open!
Steve, what could be the result of stacking bias calibration frames instead of dark flat frames?
Thanks Steve!
I understand the different flows available in
DSS (thanks for including them here JT). What I am missing is what the negative impact would be from using a master bias frame for calibration instead of dark flats. Is it the different noise patterns that are generated by a 1/8000 bias frame and a 1/20 dark flat frame? I would like to understand the "why" behind the statement.
Bob
Re: Vignetting and Flat Frames
Posted: Tue Jun 30, 2020 3:29 am
by starfield
With my CMOS camera I never had any luck with bias and just use Dark Flats. This has worked pretty consistently.
One change I made recently was to start creating my master flat in SharpCap. It also can be set to capture dark flats at the end and work this into to the master flat. When I remember to that I can skip adding in the dark flats in DSS. Of course, I've started playing with SGP so may need to go back to capturing the dark flats.
Re: Vignetting and Flat Frames
Posted: Tue Jun 30, 2020 5:45 am
by STEVE333
bobharmony wrote: ↑Tue Jun 30, 2020 3:05 am
Juno16 wrote: ↑Mon Jun 29, 2020 2:31 am
Very interesting!
Thanks Graeme for starting this thread. Thanks Steve @STEVE333 for the interesting fact concerning dark flats instead of bias calibration frames for dslr’s. I’ve always shot and stacked bias calibration frames with my dslr. After reading your post #2 above, my eyes were wide open!
Steve, what could be the result of stacking bias calibration frames instead of dark flat frames?
Thanks Steve!
I understand the different flows available in
DSS (thanks for including them here JT). What I am missing is what the negative impact would be from using a master bias frame for calibration instead of dark flats. Is it the different noise patterns that are generated by a 1/8000 bias frame and a 1/20 dark flat frame? I would like to understand the "why" behind the statement.
Bob
Sorry for taking so long to respond to these questions. I missed them. A little tired from trying to capture some new NB data on the Western Veil.
Apparently the Read Noise "pattern" varies as the exposure time varies. Thus, the "noise pattern" for a very short Bias frame will not match the "noise pattern" for the longer exposure Flat frames. Using a Flat Dark (or Dark flat) with an exposure time equal to the Flat Frames will exactly match the Read Noise "pattern" which is what is required.
The same reasoning is why uncalibrated Dark Frames are subtracted from the Lights. The uncalibrated Dark Frames already have the same Read Noise "pattern" as the Light frames because they have exactly the same exposure.
I have to thank Adam Block for explaining all of this to us mere mortals.
Hope this makes sense.
Steve