Does it work out with bias and flat frames only?

[Guest Tuomo]

They say that dark frames might do more harm than good when imaging with DSLR. Problem is to match light temps and dark temps. Can I just stack 200 bias frames with good flats and leave it there?

[Uranium235]

If its a CCD or cooled CMOS, then yes - all you need is bias and flats (I do it all the time).

[ollypenrice]

Quite possibly, yes. But use your bias twice. Use them as a 'dark for flats' to make a master flat. Then use them as you would use a dark for for your light frames (that is your images of the target.) If you dither between sub expousres during capture by about 12 pixels this may well give you the best possible result.

Olly

[emyliano2000]

2 hours ago, ollypenrice said:

Quite possibly, yes. But use your bias twice. Use them as a 'dark for flats' to make a master flat. Then use them as you would use a dark for for your light frames (that is your images of the target.) If you dither between sub expousres during capture by about 12 pixels this may well give you the best possible result.

Olly

OK, I'm only using flats and bias too. I'm trying to get this right. I do dither at 12px between exposures. So far I've been using DSS and I only entered the flats and the bias. DO I make a master out of the bias and use it as a dark flat or use it as a dark for the lights?

You probably know that, in DSS you have 1-Lights, 2-Darks, 3-Flats, 4-Dark flats and 5-Bias. Do I use the Bias frames in 4 to make a Master Dark Flat and then use that Master Dark Flat in 2 and the Bias frames in 5? Can I use the Master Bias in 2? Is there any difference in creating the master when you enter the frames in 4 or 5?

What I understood is that you use the Bias in 5 and the master created out of them is used in 2? Did I understand right?

I could take dark to match the lights as APT is giving me the EXIF temperature but it's such a waste of  valuable time that can be used to take light.

I hope I made myself understood.

Thanks

Emil

[ollypenrice]

 

12 minutes ago, emyliano2000 said:

OK, I'm only using flats and bias too. I'm trying to get this right. I do dither at 12px between exposures. So far I've been using DSS and I only entered the flats and the bias. DO I make a master out of the bias and use it as a dark flat or use it as a dark for the lights?

You probably know that, in DSS you have 1-Lights, 2-Darks, 3-Flats, 4-Dark flats and 5-Bias. Do I use the Bias at 4 to make a Master Dark Flat and then use that Master Dark Flat in 2 and the Bias in 5? Can I use the Master Bias in 2? Is there any difference in creating the master when you enter the frames in 4 or 5?

I hope I made myself understood.

Thanks

Emil

The trouble with DSS (or my trouble with DSS) is that I don't know what it is going to do with files I put into it. I use Astro Art where I know how to control what will be done with each file I put in. I can only advise from first principles.

1) A master flat should be made using a master bias as a dark for flats.

2) A master bias can be used as a dark for lights provided the lights have been dithered.

You need a DSS expert to give a more complete answer.

Olly

[emyliano2000]

Please DSS experts, shine a light on my exploding brain

From what you just said I understood that I can use the Bias 3 times

Have a look at the Screenshot. Is that how I should be using the Bias?

Thanks

Emil

[Guest Tuomo]

6 hours ago, emyliano2000 said:

Please DSS experts, shine a light on my exploding brain

From what you just said I understood that I can use the Bias 3 times

Have a look at the Screenshot. Is that how I should be using the Bias?

Thanks

Emil

Gets confusing....But do you really need to use BOTH dark flats and bias frames? I think using both might cause more troubles...

[Knight of Clear Skies]

It's confusing. Just using dark bias in DSS seems to work, I accidently stacked an image with just light frames the other day and it came out much worse. (If I get the chance I'll double check this to confirm.)

Reading the DSS manual it says its only properly calibrated if it has lights, flats, dark bias and darks. In the absence of proper darks I've been taking a copy of my dark bias and using that as a master dark recently, this seems to work OK but I haven't done a proper side-by-side comparison.

[emyliano2000]

Ok, I've done a little test on the M97-M108 exposures to see if I can get any improvement. I created a master dark and a master dark flat out of the bias frames and used them all in DSS. So now I have the Lights, the master dark created out of bias frames, the flats, the master dark flat created also out of the bias frames and the bias frames. The improvement is quite great compared to the one where I only used light, flats and bias, with no darks and dark flats.

I'll post the result later.

Emil

[happy-kat]

Oooh I'll try this later

[alacant]

8 hours ago, Tuomo said:

BOTH dark flats and bias frames

Hi. If you follow the process of a light-bias-flat dss session you'll see that the first thing it does is create a master bias. Then it uses the master bias to create a master flat. It then calibrates the light frames. I check nothing in dark-anything. Summary: stack light, bias and flat only. Conclusion: far less noise than with the conventional light-dark-bias-flat.

Note that these are my experiences with a Canon 700d.  HTH.

[emyliano2000]

These are my results. 2 hour exposures @ 400 ISO for each of them.

Levels, curves and match color in PS and a little noise reduction in LR

 

[emyliano2000]

Maybe it's just me with my reduced knowledge in stacking and processing that sees an improvement. Anybody can give it a try and see what they think.

Emil

[spillage]

I have found this very interesting. I am at work but just ran through some data I have in DSS with just light, flat and bias. It seems to have come out fine.

I will have to take the results home and run through startools to see how it turns out.

Spill.

[emyliano2000]

I've heard it before, people using the master bias as a replacement for darks  or dark flats, don't remember which one, but never got into it. It is the first time I try it out. Maybe it's just me with my reduced knowledge in stacking and processing that sees an improvement. Anybody can give it a try and see what they think.

Emil

[happy-kat]

I normally use lights, flats and bias so I'll try this bias matter as my darks master etc and compare. I'll use the same files I processed last week to compare.

[spillage]

I have just done this with just lights, flats and bias. My monitor is not the best so may look awful on a descent screen.  I will have to create a master bias and use this for a dark file and see if there is any difference. I did run this through startools and a little tweak in lightroom.

 

[happy-kat]

This is my test and it did work for me.

This is a crop and autodevelop compare of the same source files one flats, bias and lights. the other flats, bias and lights, and then used the master bias created as a master dark flat and a dark (needed to copy and rename it to use it as the dark file).

Before the process from last week, same source files used same crop and same bin 50%.

Second process tonight using this methoad of utilising the bias master. whilst my processing on my bloated stars was better tonight the overall background is better on this new file I think on pixel peeping.

 

 

[ollypenrice]

4 hours ago, emyliano2000 said:

I've heard it before, people using the master bias as a replacement for darks  or dark flats, don't remember which one, but never got into it. It is the first time I try it out. Maybe it's just me with my reduced knowledge in stacking and processing that sees an improvement. Anybody can give it a try and see what they think.

Emil

Using a master bias as a dark for flats (or what DSS calls a flat dark) is a well proven method of calibrating flats. There is no significant difference between a true dark for flats (taken at the same exp time as the flats) and a master bias because there is next to no thermal noise building up in short flat exposures.

You can also use a master bias as a dark for lights but it will only remove the low level noise. You will also need a way to get rid of brighter noise. You can use a bad pixel map or you can dither during capture and use a sigma clip routine. Or both.

One thing is for sure: badly matched darks can make things worse.

Olly

[happy-kat]

My altaz mount will be fulfilling the dithering

[ollypenrice]

42 minutes ago, happy-kat said:

My altaz mount will be fulfilling the dithering

Quite seriously, so will a little polar misalignment. If it doesn't show in a single sub, does it matter?

Olly

[alacant]

22 hours ago, happy-kat said:

My altaz mount will be fulfilling the dithering

Hi. I think you'd have to pause between frames to get a big enough gap for the stacking  to play with; the ideal being that you move the camera only between exposures. Hope it works out. 

[Nigel G]

I have tried this and got a complete black image, 206 mb of black.

Clearly I have messed up somewhere.

I registered my lights flats and bias. I then took the master bias and renamed it and added it as a dark. I took the same master bias and added it as a dark flat.

So 70 lights 1 dark 30 flats 1 dark flat and 50 bias to stack ? Correct ?

Try again :/

Nige.

[alacant]

41 minutes ago, Nigel G said:

I registered my lights flats and bias.

Hi. As I understand -and do it- that's it. DSS makes the masters from there and uses them accordingly:

1. master bias from bias

2. master flat from flats and master bias

These are then used to calibrate your light frames during stacking. I think the confusion is the terminology used by different software. This post is using DSS vocabulary only. HTH.

[happy-kat]

You need to copy and rename the bias matter in order to use it twice