Lightbox for Flats

[Marmo720]

10 hours ago, bottletopburly said:

Flat frames should be taken at same  iso setting as lights , I use an EL panel with a couple of sheets of A4 paper sandwiched between frosted acrylic absolutely no problems with flats now , turn dial on camera to AV mode before taking flats it will work out correct exposure and I take 25 flats each session .

Thanks. Not sure why I am getting bright centre and dark edges then as I currently take flats at same settings as lights. Maybe using too many flats? I currently use 50 flats while I see most people using closer to half that. I would think DSS should be able to take care of too many flats if that was the case?

Forecast is clear tonight so will try as many different things as possible. Now that I have a working light box, I can just take a bunch of flats at each ISO and see what works.

[Laurin Dave]

Do your flats work? Ie do they correct vignetting and dust bunnies..  if so all is well..  they generally  look worse than they are (ie bright centre dark corners) because they are auto stretched so much.  Measure the difference in intensities in the middle and the corners of an Unstretched flat I suspect it’ll be less than you’d think

[Marmo720]

1 hour ago, Laurin Dave said:

Do your flats work? Ie do they correct vignetting and dust bunnies..  if so all is well..  they generally  look worse than they are (ie bright centre dark corners) because they are auto stretched so much.  Measure the difference in intensities in the middle and the corners of an Unstretched flat I suspect it’ll be less than you’d think

The flats alone look fine to me in that I can see the slight vignetting and and dust bunnies that would be subtracted from the lights. Unfortunately, once I have stacked them in DSS with the lights, the final image looks bright at the centre and darker on the outsides. To the point where there is a massive bright halo around my target. 

In saying that, I sent the final tiff to @carastro and @RolandKol, and both produced fantastic looking images after some processing so maybe that is where my issue is? I just don't remember seeing such a huge halo back in the summer when I was using ISO100 for flats and found I could process easier. Hopefully skies stay clear tonight and I can produce a bunch of different flats to see what works.

[carastro]

It's sounding suspiciously like DSS is not applying the flats properly.   Do you have a master flat afterwards (should be filed in the same directory as the flat files.   If so can you post it on here for us to take a look at?

Finally are you using bias frames?  I hear tell that they help the flats to be applied properly - though personally I find they create problems using my mono camera, but I think yours is a DSLR, so lets take a look at the master flat.

Carole 

[RolandKol]

2 hours ago, Marmo720 said:

The flats alone look fine to me in that I can see the slight vignetting and and dust bunnies that would be subtracted from the lights. Unfortunately, once I have stacked them in DSS with the lights, the final image looks bright at the centre and darker on the outsides. To the point where there is a massive bright halo around my target. 

In saying that, I sent the final tiff to @carastro and @RolandKol, and both produced fantastic looking images after some processing so maybe that is where my issue is? I just don't remember seeing such a huge halo back in the summer when I was using ISO100 for flats and found I could process easier. Hopefully skies stay clear tonight and I can produce a bunch of different flats to see what works.

Nope mate...

As I mentioned before, I was made to crop the most of the image ;(
As Carole said, it looks like flats are not applied properly, - so maybe you have a mistake in your staking routine?
So if possible, drop the files into the share drive (all lights and new/old flats, also Master-bias, Master-dark (to save download/upload times).
I will try to stack them to check if the result is the same, I have nothing to do anyway as my Filters are not yet arrived!!! (getting furious!!!!  but maybe that's the reason, you guys, have some clear skies...  )

[Marmo720]

Thanks both.

I will upload the master flats for different ISO as well as the lights/darks/bias after the session tonight. Hopefully can also tell me if there is gradient from the lightbox or if that is OK.

Can't believe your filters haven't arrived yet Roland. Any hope of them arriving this side of Christmas before all the holiday posts?

 

[carastro]

Hopefully my eyes will have recovered by this evening after going to the eye clinic today.  Had eye drops to dilate the pupils 4.5 hours ago and still can't see properly.

Carole 

[RolandKol]

19 minutes ago, carastro said:

Hopefully my eyes will have recovered by this evening after going to the eye clinic today.  Had eye drops to dilate the pupils 4.5 hours ago and still can't see properly.

Carole 

Sad to hear it...
I hope you will get better soon

[Marmo720]

21 minutes ago, carastro said:

Hopefully my eyes will have recovered by this evening after going to the eye clinic today.  Had eye drops to dilate the pupils 4.5 hours ago and still can't see properly.

Carole 

Didn't know - I hope you recover fast and well. Are you planning to image tonight if your eyes are better? 

 

[carastro]

Now 3.15 and eyes are seeing a bit better.   I have people here this evening but if it is clear after they have gone I might go out.

Carole 

[carastro]

Eyes seem OK now.  Didn't go out as the "visitors" didn't leave until about 10.30 and was too tired by then. 

Carole 

[Marmo720]

10 hours ago, carastro said:

Eyes seem OK now.  Didn't go out as the "visitors" didn't leave until about 10.30 and was too tired by then. 

Carole 

I am glad to hear your eyes are OK now. 

I spent the entire night out capturing last night in the freezing cold. Went to bed at 3am and woke up at 6am for work. So very tired at the moment but hopefully worth it. 

I captured 90x60sec subs of M31 at ISO800. Took new bias and darks too. I also took 53x60sec subs at ISO800 of the Rosette Nebula before the clouds stormed in.

For Flats, I reduced the panel brightness to its dimmest and took:

- 10x ISO 100 with panel in normal position.

- 10x ISO 100 with panel in inverted position.

- Same as above 2 but with ISO 200.

- 20x ISO 400 and ISO 800 in the normal panel position.

 

I inverted the panel to check for any gradient from the LED lights being on one side. Visually, I couldn't see any at all and the flats looked the same regardless of panel orientation. Hopefully you guys can spot any gradient once I upload the masters for each setup.

I did a quick stack on the first set of IS100x10 Flats with my M31 this morning before rushing to work and I could still see the brighter centre and darker outsides after stacking. I will do stacks with all the different flats to see if any work and/or if you guys can spot where my issues are.

Thanks,

Mo

[Marmo720]

Hi all,

Below are links to my files from last week. Not having luck with stacking and cleaning the image so hope the community wisdom will help spot what I might be doing wrong and how to improve it. 

My camera is astromod Canon 600D, I have a field flattener on that and using an Altair Wave 80. I am using APT for capture, PHD2 for guidance with ASI 224 and I am applying dithering. Finally, I use DSS for stacking and a quick check on GIMP or Photoshop shows bright white halo in the middle and darker outsides. 

 

Many thanks in advance for any help.

 

Lights: https://we.tl/t-9O4TS659iR?src=dnl

Flats @ ISO100: https://we.tl/t-7UAh1fkHKr?src=dnl

Flats @ ISO800: https://we.tl/t-GtRCNFfw5X?src=dnl

Darks: https://we.tl/t-WTV5rQXlVg?src=dnl

Bias: https://we.tl/t-8RNquuP4ZE?src=dnl

[RolandKol]

6 hours ago, Marmo720 said:

Hi all,

Below are links to my files from last week. Not having luck with stacking and cleaning the image so hope the community wisdom will help spot what I might be doing wrong and how to improve it. 

My camera is astromod Canon 600D, I have a field flattener on that and using an Altair Wave 80. I am using APT for capture, PHD2 for guidance with ASI 224 and I am applying dithering. Finally, I use DSS for stacking and a quick check on GIMP or Photoshop shows bright white halo in the middle and darker outsides. 

Many thanks in advance for any help.

Lights: https://we.tl/t-9O4TS659iR?src=dnl

Flats @ ISO100: https://we.tl/t-7UAh1fkHKr?src=dnl

Flats @ ISO800: https://we.tl/t-GtRCNFfw5X?src=dnl

Darks: https://we.tl/t-WTV5rQXlVg?src=dnl

Bias: https://we.tl/t-8RNquuP4ZE?src=dnl

Hi Mo,

is there any particular reason why your different sub types have the different sizes???

Very strange....
Lights 24mb per sub,

Darks/Bias - 19mb

Flats -22mb

I guess, you may have incorrect settings on APT for calibration subs..

Edited December 20, 2019 by RolandKol

[carastro]

Only just twigged Marmo is the same person I have been in E mail discussion with where I know you as Mohamed.  Lol

Just to say guys, I have spent a couple of days fiddling around with Mo's data, and I am foxed.  I think the flats are faulty, but the histogram is in the right place on the 800iso flats.  On the 100iso flats the histogram seemed too far to the left, but whatever way I stack them DSS does not seem to be applying the flats.  

I just hope some-one on here can work out what is going on.  I am also rather out of date with DSLR imaging, and what should look right with the flats, but I have to say on trying to stretch the master, I can't really get Vignetting or dust to show up (except for one dark bit of dust).  But if the flats are faulty yet the histogram is in the right place, I just don't know what to suggest.

As a start I would forget about doing the APT flats plan, and just use "Shoot" and try out different lengths" until the histogram looks right, I have another chum who had to abandon the "Flats plan" with a DSLR in APT. 

MO, your master flat should look something like this once heavily stretched.  I could not get your flats to look like this.

Carole 

Edited December 20, 2019 by carastro

[RolandKol]

31 minutes ago, RolandKol said:

Lights 24mb per sub,

Darks/Bias - 19mb

Flats -22mb

I guess, you may have incorrect settings on APT for calibration subs..

Just checked my old Canon files,

they have different sizes also...

Please ignore my previous msg...
 

Just a quick observation (loud thinking)...

Flats should be a fracture of the second, - somewhere around one third of the sec .
It can be a second also, but 10s is a bit too much in terms of efficency... it takes too long... and short flats usually work OK.

Not sure if such a dim light, which enables 10sec Flats subs, is an issue, - but it may be.

[RolandKol]

P.S. Will try stacking them on Monday...
Flats actually do look matching the lights, which have almost no vignetting...

I would guess, - result should be OK...

lets see

[Marmo720]

3 hours ago, carastro said:

Only just twigged Marmo is the same person I have been in E mail discussion with where I know you as Mohamed.  Lol

Haha now I understand why you were asking about my gear setup. Confused me considering you pretty much helped me choose and setup my entire gear

Thank you so much for spending some of your time trying to get this resolved and I appreciate the help. Absolutely confused at the moment and hoping it is an easy fix.

@RolandKol thank you for the help too. I did think the flats took a lot longer this time than before and the lightbox is set to lowest with 6 sheets of white paper. I have in the past used ipad at much higher brightness and a t-shirt, and the flats would be much quicker but the histogram was about 2/3 towards the right side instead of where it is now. 

 

[vlaiv]

I'm late to the party and did not read most of what has been written so far (sorry about that), so I might mention something that has already been addressed, but here are my finding so far:

- Flats are severely under exposed - both ISO800 and ISO100. Not sure if that is the cause of the problem - might not be, but it is something that you want to correct.

Flats have proper looking histogram, but I believe they are well under exposed. They should be at about 3/4 to the right of histogram, but on scale 0-16384 (14bit DSLR) it looks like this:

Otherwise, when zoomed in on relevant part, it looks fine:

Three peaks as it should have, nothing clipping, etc ...

- Second observation for all those who will attempt to process above data - don't even try to use ISO100 flats. Master flat should be calibrated with flat darks, or in this case, master bias will be enough. Both set of files (flats and bias) need to be shot at same ISO settings. Above included files contain ISO800 bias and both ISO800 flats and ISO100 flats. ISO800 flats will be fine to use with bias, but not ISO100 flats because of mismatch in iso setting.

Third observation is that something is seriously wrong with flats and I'm not still sure what it is, here is what flat fielding looks like (single sub after calibration with single dark, flat and bias):

This clearly shows that there was under correction of flat calibration in corners, but there was over correction in the central part - dust doughnut is brighter than it should be.

In most cases one has over or under correction over whole field and that is consequence of either improper calibration or some issues with darks or bias, or maybe some light leak, but this is quite a bit different.

My guess is that it has something to do with the way flats were taken, or the flat panel itself (although I'm not sure what it could be).

Can I ask what scope is this and what the setup looks like and under what conditions where flats taken?

In particular - was there change in anything in setup between lights and flats - for example - focus position?

[Marmo720]

On 21/12/2019 at 12:40, vlaiv said:

I'm late to the party and did not read most of what has been written so far (sorry about that), so I might mention something that has already been addressed, but here are my finding so far:

- Flats are severely under exposed - both ISO800 and ISO100. Not sure if that is the cause of the problem - might not be, but it is something that you want to correct.

Flats have proper looking histogram, but I believe they are well under exposed. They should be at about 3/4 to the right of histogram, but on scale 0-16384 (14bit DSLR) it looks like this:

Otherwise, when zoomed in on relevant part, it looks fine:

Three peaks as it should have, nothing clipping, etc ...

- Second observation for all those who will attempt to process above data - don't even try to use ISO100 flats. Master flat should be calibrated with flat darks, or in this case, master bias will be enough. Both set of files (flats and bias) need to be shot at same ISO settings. Above included files contain ISO800 bias and both ISO800 flats and ISO100 flats. ISO800 flats will be fine to use with bias, but not ISO100 flats because of mismatch in iso setting.

Third observation is that something is seriously wrong with flats and I'm not still sure what it is, here is what flat fielding looks like (single sub after calibration with single dark, flat and bias):

This clearly shows that there was under correction of flat calibration in corners, but there was over correction in the central part - dust doughnut is brighter than it should be.

In most cases one has over or under correction over whole field and that is consequence of either improper calibration or some issues with darks or bias, or maybe some light leak, but this is quite a bit different.

My guess is that it has something to do with the way flats were taken, or the flat panel itself (although I'm not sure what it could be).

Can I ask what scope is this and what the setup looks like and under what conditions where flats taken?

In particular - was there change in anything in setup between lights and flats - for example - focus position?

Thank you so much for taking the time to provide feedback. 

Flats are probably underexposed due to me setting the lightbox on lowest setting and using 6 white papers. Each flat took significantly longer than I thought it would take but as the histogram looked correct, I assumed it was fine. In the past, I used an iPad at about 75% brightness and a t-shirt. So I guess change the brightness of the lightbox to being mid-high might be better and will try that next. I will also try with the iPad just in case as that used to work fine in the past.

I didn't realise a single flat fielding would show the impact so much so I could spend the next clear night just checking my sub/lights/darks/bias using single files before wasting another evening of lights that I can't make use of.

There wasn't any change between lights/flats besides moving the scope to better position for the flats. Don't think there was any focus shift as I followed up the M31 Flats with M42 lights and they also have the same issue.

My setup train is currently Altair Wave 80 ED APO -> Altair 1.0x field flattener -> Canon astromod 600D. My mount is the is the Skywatcher HEQ5 Pro and I am using the ASI 224 MC for guiding. Software is APT for capture, PHD2 for guiding and Stellarium for target selection.

I don't currently have any dew heaters and my original thought was maybe dew had covered centre part of the scope lens but I would think that would be obvious from lights and flats? Is this something could explain my images? 

Forecast is clear skies after midnight on Christmas Eve so maybe I will get a gift from Santa and I can have clear images too  

Thanks,

Mo

[vlaiv]

18 minutes ago, Marmo720 said:

I didn't realise a single flat fielding would show the impact so much so I could spend the next clear night just checking my sub/lights/darks/bias using single files before wasting another evening of lights that I can't make use of.

In fact you can do full calibration with single sub of each - one light, one dark, one flat and one flat dark.

The only reason we use multiple of each is to reduce noise, but if you are looking at general artifacts in the image - like background illumination and flat correction - you can improve signal to noise ratio in another way - you can bin image. I think I used bin factor of something like x10 on above image. Image will be tiny and massive amount of details will be lost - but we don't care for such details in the image to diagnose flat issues.

20 minutes ago, Marmo720 said:

There wasn't any change between lights/flats besides moving the scope to better position for the flats. Don't think there was any focus shift as I followed up the M31 Flats with M42 lights and they also have the same issue.

Ok, then focus position is not the problem here.

21 minutes ago, Marmo720 said:

I don't currently have any dew heaters and my original thought was maybe dew had covered centre part of the scope lens but I would think that would be obvious from lights and flats? Is this something could explain my images? 

No, I don't think it can be issue with flats. You can check this with your scope - add central obstruction, but in principle many people provide proof or that on daily basis -  scope with central obstruction acts as if frost is blocking light in central part of the lens (in fact being more severe - as it is 100% light blockage).  Central obstruction or in fact any obstruction before light is bent has no effect on field illumination (that is not strictly true - there are cases where level of obstruction depends on incidence angle - like with very long dew shield that will act as aperture stop at angles).

What can happen on the other hand would be dewing up corrector lens or dslr sensor (or frosting up) - any chance that happened? I mean it's a long shot since it is closed system once everything is screwed together so absolute humidity does not change, but it can happen if it was particularly humid and quite a bit warmer when you put your gear together and it cooled considerably during the night?

I highly doubt it is sensor (as it is not actively cooled), but could be corrector lens (field flattener)

We can test that theory - central blockage of corrector by dew/frost. I need to be careful about this one as it is easy to mix up things, so I'll write down everything to be corrected if I make mistake.

Flat correction = light / flat

In center we have over correction (look at that bright dust shadow) and in outer parts we have under correction - which means

center - higher value

edge - lower value

light / flat = higher value can happen if light has higher value, or flat has lower value

light / flat = lower value can happen if light has lower value or flat has higher value.

I don't think it is this either. Central blockage would cause over correction in the center but it would not cause under correction at the edges (it can't boost flat value or reduce light value).

37 minutes ago, Marmo720 said:

Forecast is clear skies after midnight on Christmas Eve so maybe I will get a gift from Santa and I can have clear images too  

I think there is something that you can try without a clear night as it just involves flats.

Try flat-flat calibration to rule out flat settings as culprit for the problem.

This would mean that you take bias subs (you can do just one or couple - number does not matter here except for reducing noise), at each ISO setting that you will test. Test with at least two different ISO settings.

For each ISO setting - shoot 2-3 different "type" of flats (where type here only means exposure length / strength of flat panel). Aim to have proper histogram each time (no clipping either to left or right) - as you did with above flats, but have different position on histogram each time - around 20-30% around 50-60% and proper one at around 80-85% for highest peak.

Each of these flats should calibrate others (once bias is removed) - meaning that you take any two flats - flat1 and flat2 and divide them, you should end up with perfectly uniform illuminated sub (no vignetting / no dust shadows to be seen).

If this happens and each flat calibrates others - you can be 100% sure that your flats are working as they should - then we need to look elsewhere for the source of the issue.

[Marmo720]

19 minutes ago, vlaiv said:

I think there is something that you can try without a clear night as it just involves flats.

Try flat-flat calibration to rule out flat settings as culprit for the problem.

This would mean that you take bias subs (you can do just one or couple - number does not matter here except for reducing noise), at each ISO setting that you will test. Test with at least two different ISO settings.

For each ISO setting - shoot 2-3 different "type" of flats (where type here only means exposure length / strength of flat panel). Aim to have proper histogram each time (no clipping either to left or right) - as you did with above flats, but have different position on histogram each time - around 20-30% around 50-60% and proper one at around 80-85% for highest peak.

Each of these flats should calibrate others (once bias is removed) - meaning that you take any two flats - flat1 and flat2 and divide them, you should end up with perfectly uniform illuminated sub (no vignetting / no dust shadows to be seen).

If this happens and each flat calibrates others - you can be 100% sure that your flats are working as they should - then we need to look elsewhere for the source of the issue.

Thanks for the advise. I will try this. Just to check:

- The idea is to stack flat1 and flat2 in DSS where flat1 is input as a light sub and flat2 is used as a flat sub? So the DSS process will subtract the two and result in uniform sub? Of course using Bias as well as suggested. 

- How do I view the histograms you shared? I have been using the histogram in APT and you said the split between the peaks is correct but the location of the entire histogram should be to the right about 3/4? I have noticed this is the case if I take brighter flats so just want to check that increasing the brightness of the lightbox so that the histogram in APT is about 3/4 of the way from the left is going to be correct? Or have I misunderstood? I read a few places the histogram in APT should be 1/3 from the left so just checking.

Many thanks again.

 

[vlaiv]

7 minutes ago, Marmo720 said:

Thanks for the advise. I will try this. Just to check:

- The idea is to stack flat1 and flat2 in DSS where flat1 is input as a light sub and flat2 is used as a flat sub? So the DSS process will subtract the two and result in uniform sub? Of course using Bias as well as suggested. 

- How do I view the histograms you shared? I have been using the histogram in APT and you said the split between the peaks is correct but the location of the entire histogram should be to the right about 3/4? I have noticed this is the case if I take brighter flats so just want to check that increasing the brightness of the lightbox so that the histogram in APT is about 3/4 of the way from the left is going to be correct? Or have I misunderstood? I read a few places the histogram in APT should be 1/3 from the left so just checking.

Many thanks again.

 

Maybe it would be best to first just take flats and then we can see how to best test them. I would personally use ImageJ to do all processing and testing. You can use that also - it is free software used for scientific image manipulation and it's written in Java, which means that it will work on almost any computer / operating system. Once you have your flats and bias, I'll walk you thru the procedure how to do various things with them - like check histogram, do statistical analysis (mean value, min/max and such sort of things) and how to do image math (calibrate them, stack them, etc) ....

You will need one more piece of software (also free) - FitsWork. That one is used to convert DSLR raw files into fits that can be then used in ImageJ.

As for lights and histogram position - it is down to two things:

1. linearity of sensor

Here you want to ensure that there is no clipping left / right (low or high values - all three peaks present and looking nice in histogram). There is also concern about linearity of sensor response - it can happen that sensitivity depends on signal level (or rather gathered signal so far). Which means that doubling of amount of light, or doubling exposure length does not produce twice as high recorded signal (ADU level). This is bad thing for calibration and needs to be addressed in certain way. I think that nowadays pretty much no sensors that we use in AP behave that way, but "most linear" region was around 2/3-3/4 back then when it happened. In fact saturation and anti blooming gates cause non linearity in modern sensors - but we consider them to be "clipping" to the right.

2. Minimizing noise impact.

Once you are nice linear region and that part is covered - well, you want to have the least noise in your flats as possible - that is why we take more than one flat as stacking them reduces their noise. This noise will end up being injected back in the image in a certain way so you want to minimize it. There is two ways of minimizing the noise:

1. stack more subs

2. make sure SNR of subs is high to begin with - this part relates to histogram position. SNR is signal to noise ratio or signal/noise. If we want it to be high - we want our signal to be strong. Stronger signal is right on histogram or higher value of pixels (so closer to 1 but not too close as it will star saturating - that is why 80-85% is often mentioned).

bottom line - you can use histogram at even 5% but make sure you have enough flat subs to stack to minimize noise coming from subs.

 

 

[Marmo720]

1 hour ago, vlaiv said:

Maybe it would be best to first just take flats and then we can see how to best test them. I would personally use ImageJ to do all processing and testing. You can use that also - it is free software used for scientific image manipulation and it's written in Java, which means that it will work on almost any computer / operating system. Once you have your flats and bias, I'll walk you thru the procedure how to do various things with them - like check histogram, do statistical analysis (mean value, min/max and such sort of things) and how to do image math (calibrate them, stack them, etc) ....

You will need one more piece of software (also free) - FitsWork. That one is used to convert DSLR raw files into fits that can be then used in ImageJ.

As for lights and histogram position - it is down to two things:

1. linearity of sensor

Here you want to ensure that there is no clipping left / right (low or high values - all three peaks present and looking nice in histogram). There is also concern about linearity of sensor response - it can happen that sensitivity depends on signal level (or rather gathered signal so far). Which means that doubling of amount of light, or doubling exposure length does not produce twice as high recorded signal (ADU level). This is bad thing for calibration and needs to be addressed in certain way. I think that nowadays pretty much no sensors that we use in AP behave that way, but "most linear" region was around 2/3-3/4 back then when it happened. In fact saturation and anti blooming gates cause non linearity in modern sensors - but we consider them to be "clipping" to the right.

2. Minimizing noise impact.

Once you are nice linear region and that part is covered - well, you want to have the least noise in your flats as possible - that is why we take more than one flat as stacking them reduces their noise. This noise will end up being injected back in the image in a certain way so you want to minimize it. There is two ways of minimizing the noise:

1. stack more subs

2. make sure SNR of subs is high to begin with - this part relates to histogram position. SNR is signal to noise ratio or signal/noise. If we want it to be high - we want our signal to be strong. Stronger signal is right on histogram or higher value of pixels (so closer to 1 but not too close as it will star saturating - that is why 80-85% is often mentioned).

bottom line - you can use histogram at even 5% but make sure you have enough flat subs to stack to minimize noise coming from subs.

 

 

Thank you - helpful advise and appreciate the support.

I will do the flats/bias and share so we can work through them.

Based on the last part then, is it ideal then for the histogram for lights and flats to be in similar position? Or at least not too far apart to result in over/under correction without the use of large number of flats? I guess going further, I am imaging from bortle 8 location so at some point I would need to maximise my sub duration/iso settings so will be interesting to know where the light histograms should be? Around 80% to maximise SNR without clipping? I have guiding so I could technically do 5 min subs but without LP filter, I am easily pushing the histogram too far to the right even at ISO800 and sub lengths of 3mins. 

For reference, the images I attached have the histogram for the lights at about 50-60% while the flats are around 30% although I am not so sure on the flats reading as they took 10-30 seconds to capture. Both taken at ISO800 and lights duration is 60 seconds.

 

[vlaiv]

1 hour ago, Marmo720 said:

Thank you - helpful advise and appreciate the support.

I will do the flats/bias and share so we can work through them.

Based on the last part then, is it ideal then for the histogram for lights and flats to be in similar position? Or at least not too far apart to result in over/under correction without the use of large number of flats? I guess going further, I am imaging from bortle 8 location so at some point I would need to maximise my sub duration/iso settings so will be interesting to know where the light histograms should be? Around 80% to maximise SNR without clipping? I have guiding so I could technically do 5 min subs but without LP filter, I am easily pushing the histogram too far to the right even at ISO800 and sub lengths of 3mins. 

For reference, the images I attached have the histogram for the lights at about 50-60% while the flats are around 30% although I am not so sure on the flats reading as they took 10-30 seconds to capture. Both taken at ISO800 and lights duration is 60 seconds.

 

While you can apply similar logic to lights as to flats - there are differences which in practice prevent you to do so.

For lights you also want as large signal as you can get in single sub within certain we could say constraints. Flats light is way too brighter than stuff that we image and it is really easy to achieve 75% or above of histogram. With lights it is much easier to go the other way around - gather as much subs as possible then it is to push histogram past 75% or so - as that would require multiple hours for single exposure - maybe one would not even manage to do it in a single night.

There is rule that will help you with understanding relationship between best SNR possible and sub duration and light pollution, but that requires you to know things about your camera - namely it's read noise and gain factor.

What is always true is following - fewer longer subs will be better than more shorter subs - totaling to same imaging time (it's better to have x12 5 minute subs than x60 1 minute subs - both total 1h or imaging time). This is true always, but how much difference there will be - depends on read noise of the camera and light pollution, or rather other noise sources. Once any of other noise sources "swamps" read noise (is higher in intensity about x4-5) there will be minimal difference in end result. This is why you can use shorter subs in higher light pollution - as there will be no difference (visible) to using longer subs (shorter up to a point - there is always that point of diminishing returns). You need to know specs of your camera (not something that is readily available for DSLRs unlike dedicated astro cameras) in order to calculate optimum exposure length for your conditions.

In absence of solid numbers, there is simple way to determine your exposure - expose for "reasonable" amount of time so that your guiding works well (there are no signs of star trailing and stars are round and tight), you gather enough subs for your imaging time - advanced stacking algorithms require some amount of data - they work better if you feed them 20, 30 or 50 subs as opposed to 2, 3 or 5 subs, and be careful about wasted data - if there is wind gust or cable snag or passing airplane or anything happens to make you throw away your data - it's better to throw away 1 minute of imaging than it is to throw away 20 minutes of imaging.

For this reason - pick exposure times in range of 2-3 minutes (1 minute is ok, 4 minutes is ok - as long as it is "reasonable" amount of time in above sense).