Help with circles in images

[f33n3y]

Hi all,

I'm a newbie to mono imaging and am having trouble with circular/ring shapes in my subs. The circles are particularly pronounced after stacking my Lum subs but also appear in blue stacks. At the moment, this is making my lum subs unusable and I'm only just managing to put together RGB images.

I'm using an ZWO ASI1600mm with ZWO filter wheel and filters (36mm unmounted).

The attached TIF (LumStack.TIF) is 10x150sec of Lum subs stacked in DSS. The circles aren't very pronounced at this stage but after DBE and STF (see second SunflowerLum.jpg) the circles are very clear to see.

I've also attached a stack of Lum data on Bode's galaxy which shows the same issue.

Does anyone have any ideas what could be causing these?

Thanks in advance

 

LumStack.TIF

[vlaiv]

Are you using flats to correct for dust particles? These circles look like dust particles in imaging train - flats are used to remove those along with any vignetting and uneven sensor response.

[david_taurus83]

Yes, most definitely dust. Probably on your filters. Its difficult to install the filters and keep them clean!! I had to remove mine and clean them as I was unhappy with the dust specks on mine.

[f33n3y]

@vlaiv Yes, I took flats with each filter using the white tshirt method at dawn and also dark flats. I can't leave my scope out overnight so I took it off the mount but kept the camera attached. I was hoping this would be good enough but from what I've read it's best to take the flats immediately after the lights for them to have any real effect.

@david_taurus83 Hmm, sounds like i'll need to give the filters a good clean ? What do you recommend for cleaning them? I was always hesitant to wipe them with anything in case I scratched them but at the moment my lum subs are practically unusable so I think I'll have to!

[vlaiv]

2 minutes ago, f33n3y said:

@vlaiv Yes, I took flats with each filter using the white tshirt method at dawn and also dark flats. I can't leave my scope out overnight so I took it off the mount but kept the camera attached. I was hoping this would be good enough but from what I've read it's best to take the flats immediately after the lights for them to have any real effect.

@david_taurus83 Hmm, sounds like i'll need to give the filters a good clean ? What do you recommend for cleaning them? I was always hesitant to wipe them with anything in case I scratched them but at the moment my lum subs are practically unusable so I think I'll have to!

Hm, I was almost sure that you did not do flat correction.

What does your master flat look like? Same patterns as these? It might be the case of shifted dust, but I think that would be limited in impact - maybe just few specs shifted. As is it looks like no flats applied at all - a lot of particles all over the place.

There could also be the case of flats under correction - if your master flat has dark patterns in exactly the same place and still leaves those in calibrated light subs - it can be due to under correction.

[david_taurus83]

Can you please post up a stretched image of your master flat? They should calibrate out unless its covered in dirt!

[f33n3y]

4 minutes ago, vlaiv said:

Hm, I was almost sure that you did not do flat correction.

What does your master flat look like? Same patterns as these? It might be the case of shifted dust, but I think that would be limited in impact - maybe just few specs shifted. As is it looks like no flats applied at all - a lot of particles all over the place. 

There could also be the case of flats under correction - if your master flat has dark patterns in exactly the same place and still leaves those in calibrated light subs - it can be due to under correction. 

I've attached my master flat TIF and also a jpg of the lum master flat stretched with STF in Pixinsight. Just appears to be very grainy which doesn't seem right ?

 

MasterFlat_ISO0.tif

[vlaiv]

4 minutes ago, f33n3y said:

I've attached my master flat TIF and also a jpg of the lum master flat stretched with STF in Pixinsight. Just appears to be very grainy which doesn't seem right ?

There are several issues with this master flat.

1. 16 bit format - if you can, use 32bit data for your calibration

2. Extremely low ADU values in your master flat. Mean value is only ~100. You want your flats to be around 40000 (in 0-65535 range), or 2/3-3/4 of histogram. You need longer flat exposures for this.

How many flats did you stack? Do at least couple of dozen of them if not more.

3. Master flat shows no dust shadows what so ever, so I'm guessing something went wrong with shooting of flats. Since you mentioned removing scope from the mount and shooting flats in the morning, only thing that I can think of that would create such clean looking master flat would be - you forgot to use the same filter? Or possibly moved focus position when you packed scope away for the night (even if you left your camera attached)?

You need to leave focus position in the right place - like when you took your lights. If you racked in focuser - that would further move dust out of focus (dust shadows are just unfocused dust particles and distance to focal plane determines diameter of rings / doughnuts - if you rack in your focuser you will move them further away from focus plane and make circles large enough not to be registered on sensor at all).

[f33n3y]

4 minutes ago, vlaiv said:

There are several issues with this master flat.

1. 16 bit format - if you can, use 32bit data for your calibration

2. Extremely low ADU values in your master flat. Mean value is only ~100. You want your flats to be around 40000 (in 0-65535 range), or 2/3-3/4 of histogram. You need longer flat exposures for this.

How many flats did you stack? Do at least couple of dozen of them if not more.

3. Master flat shows no dust shadows what so ever, so I'm guessing something went wrong with shooting of flats. Since you mentioned removing scope from the mount and shooting flats in the morning, only thing that I can think of that would create such clean looking master flat would be - you forgot to use the same filter? Or possibly moved focus position when you packed scope away for the night (even if you left your camera attached)?

You need to leave focus position in the right place - like when you took your lights. If you racked in focuser - that would further move dust out of focus (dust shadows are just unfocused dust particles and distance to focal plane determines diameter of rings / doughnuts - if you rack in your focuser you will move them further away from focus plane and make circles large enough not to be registered on sensor at all).

Thanks for this! Before I even think about cleaning the filters I think I'll really need to study flats more as I've clearly not done them right at all

1. Didn't know about using 32 bit flats, I will keep this in mind

2. I did use the flats wizard in Sequence Generator Pro but I must have had the mean ADU value set faaaaar to low!

I took 40 flats with each filter

3. I'm pretty certain I used the same filter (all controllled by SGP) and kept the focuser locked. Sounds to me like the exosure time and ADUs for the flats was my biggest issue, especially seeing as there is no evidence of dust shadows on the master flat.

I think I may invest in a light box/panel so I can take flats asap after the lights and will have a play with SGPs flats wizard in the daytime and see if I can produce a master flat which is showing up dust.

[f33n3y]

Also, it looks like some of the circles line up between the lum and blue subs so it doesn't look to be dust on the filters. Perhaps it is dust on the cover of the ASI1600mm sensor cover?

[vlaiv]

6 minutes ago, f33n3y said:

Also, it looks like some of the circles line up between the lum and blue subs so it doesn't look to be dust on the filters. Perhaps it is dust on the cover of the ASI1600mm sensor cover?

If dust shadows align between filters then yes - dust is on something else - like flattener / coma corrector if you use one, or indeed on camera chamber window. Smaller the circle, closer it is to sensor so you can figure out where it is by that.

There is also a formula for calculating exact distance of dust depending on pixel size, dust shadow diameter and focal length of scope (or F/ratio, forgot which one it is).

http://www.wilmslowastro.com/software/formulae.htm#Dust

 

[carastro]

Why does the OP need to do flats as 32 bit, I have never done mine at anything except 16bit?

Carole 

[vlaiv]

1 minute ago, carastro said:

Why does the OP need to do flats as 32 bit, I have never done mine at anything except 16bit?

Carole 

Individual subs need not be anything more than is provided by camera - in this case they will be 12bit - but recorded as 16bit (12bit MSB with 4LSB bits 0 - effectively value multiplied with 16).

However I recommend that all processing including master calibration frames be done in 32bit.

This is simply because using 16 bit subs and stacking any number of such subs and recording result in 16 bit - just leads to error due to the rounding. In order to record stack of 2 frames - 16 bit each, you need 17bit precision, for 4 frames stacked you need 18bit precision and so on (number of stacked frames log base 2 represents number of additional bits needed).

For most purposes 32bit precision is enough to cover all processing without additional error introduced by rounding of due to lower bit format.

[Thalestris24]

Hi

I've had similar problems with my flats recently but the defects only show with very strong stretching of the stacked image. I've tried making synthetic flats and using ABE in Pixinsight but there doesn't seem to be way of fixing the problem after the fact. I think the trouble may lie with the flats being slightly under exposed (relative to my lights which tend to be 'exposed to the right' because of skyglow) so will try making flats using longer exposures. Other than that, it could be something peculiar to Glasgow ha ha.

Louise

 

[f33n3y]

10 hours ago, Thalestris24 said:

Hi

I've had similar problems with my flats recently but the defects only show with very strong stretching of the stacked image. I've tried making synthetic flats and using ABE in Pixinsight but there doesn't seem to be way of fixing the problem after the fact. I think the trouble may lie with the flats being slightly under exposed (relative to my lights which tend to be 'exposed to the right' because of skyglow) so will try making flats using longer exposures. Other than that, it could be something peculiar to Glasgow ha ha.

Louise

 

Good to hear that someone else is having a similar issue (in Glasgow too haha). All comments so far seem to suggest poor flats so think that'll be the first thing I try and sort before trying a clean of filters and chamber window.

Does anyone have any examples of a good master flat taken with the 1600mm they just so I can see how it appears?

[david_taurus83]

Here's one of my stretched Ha flats. You should try to take your flats again and increase the ADU. Theres a good chance the dust hasn't moved. I cleaned mine with a lens cleaning kit and a little blower.

Edited April 28, 2019 by david_taurus83

[f33n3y]

9 minutes ago, david_taurus83 said:

Here's one of my stretched Ha flats. You should try to take your flats again and increase the ADU. Theres a good chance the dust hasn't moved. I cleaned mine with a lens cleaning kit and a little blower.

Thanks, that's great! I've long detached the imaging train since imaging the above targets but hopefully by achieving focus again and retaking the flats I might be able to salvage the Lum data a bit

[wornish]

Learned a lot from this thread thanks.

There is a good tutorial here on making synthetic flats using Pixinsight which people may find of use.

 

Edited April 28, 2019 by wornish

[Thalestris24]

3 minutes ago, wornish said:

Learned a lot from this thread thanks.

There is a good tutorial here on making synthetic flats using Pixinsight which people may find of use.

 

I have tried that but it didn't work very well for me.

Louise

[f33n3y]

On 28/04/2019 at 10:31, david_taurus83 said:

Here's one of my stretched Ha flats. You should try to take your flats again and increase the ADU. Theres a good chance the dust hasn't moved. I cleaned mine with a lens cleaning kit and a little blower.

Would the below light box be suitable for taking flats with the 200PDS?

HUION A3 Size Super Bright LED Tracing Light Box with Adjustable Brightness Function https://www.amazon.co.uk/dp/B00DNPCZVG/ref=cm_sw_r_cp_apa_i_iPs0Cb2DAZ8H9

[david_taurus83]

1 hour ago, f33n3y said:

Would the below light box be suitable for taking flats with the 200PDS?

HUION A3 Size Super Bright LED Tracing Light Box with Adjustable Brightness Function https://www.amazon.co.uk/dp/B00DNPCZVG/ref=cm_sw_r_cp_apa_i_iPs0Cb2DAZ8H9

Possibly ok as a light source. A few sheets of paper to diffuse the light as well. The problem with Newtonian flats is you need to make sure that only light reflected from the primary onto the secondary makes it onto your camera sensor. If you stick that panel directly on top of the tube then you risk lighting up the focuser drawtube which can cause uneven illumination. Though some people take t shirt flats this way no problem.