Jump to content

Banner.jpg.b83b14cd4142fe10848741bb2a14c66b.jpg

Simple explanation of lights, darks, flats and bias...?


Recommended Posts

Hi to everyone :-)

Can someone please explain to a noobie (and therefore in relatively simple terms) exactly what is meant when referring to lights, darks, flats and bias when using Registax and similar programs? I am camera literate but the techniques and terminology of astrophotography are all new to me!

Many thanks in advance :-)

John

Link to comment
Share on other sites

  • Replies 29
  • Created
  • Last Reply

I have been using DSS for a while but still have no idea as to what the difference is between dark frames and dark flat frames i dont think the FAQ in DSS helps much either it talks about both in the same para.

Alan

Link to comment
Share on other sites

Hi John

Lights are the main image files you take for your content. Darks are the same camera settings but with the lens cap on, so that the software can look for noise and hot pixels that appear mostly in every frame less the required image content. I think the other one I take are bias frames which again with the lens cap on you take at the fastest shutter speed. Then the software can rule out the other unwanted bits caused by the longer exposure to what should be the very basis of just exposing.

You will use these in software like DSS not Registax that I know of. Registax tends to be used for Lunar and Planetary work formed from videos. Not the rules but what most people do.

Link to comment
Share on other sites

Lights - the "pictures" you take of your chosen target.

Darks - same but with the lens cap on! (ie the same exposure and temperature as your lights) I take mine on cloudy nights and keep a library of them.

Flats - take with an even light source (lightbox or electroluminescent panel or a t-shirt over the end of the scope and do them in daylight!) over the end of the telescope immediately after (or before) you take your lights.  With a DSLR just set the camera to "AV" mode and it will set the exposure for you.  CCD cameras are different in the way you set them up.  The optical set-up and focus must be identical tou the set-up for your Lights - do not adjust anything!!

Bias - like darks but you use the shortest possible exposure time - you can keep and re-use them.

If you have an interest in Astrophotography then "MAking Every Photon Count" by Steve Richards (Steppenwolf on SGL) is an excellent book to guide you through all the processes. (usual disclaimer).

Edit:  Too slow!  I keep having to nip outside an run off my next set of pictures!!

Link to comment
Share on other sites

Ah, thank you for the clear explanations - I have heard these terms mentioned in passing many times so it's nice to be clearer about what they mean. I also now know one difference between Registax and DSS (the latter I don't have installed - yet!)

Just a quickie regarding the darks - My camera (and I'm sure some others too) has the option of automatically doing this by recording a 'black' image at the same settings then removing the noisy hotspots from the exposure proper. I use this as a matter of course for my low-light and night-time landscape photography, and have been doing so for my first attempts at astrophotography. However, this doubles the exposure time as the camera automatically goes into noise reduction mode immediately following the exposure, and to me if this can be switched off and processed in post (by adding 'darks' from a library) then I would have twice the time available to exposure 'lights'. Is this what I should be doing?

Many thanks for the help so far - John

Link to comment
Share on other sites

I try to take as many lights as I can until the clouds roll in, then 20 darks and the bias frames I'm told don't have to be shot at the same time were I've only recently started doing that.

If you are shooting more than object on the night, both with the same exposure settings etc. then you can just use the same darks for both sets. Again as I tried shooting three subjects the other night all at 20 second exposures at 300mm f5.6 on the camera I just did the one set of darks to use for all.

DSS use for non lunar and planetary objects. Doesn't matter if it is a wide shot of a constellation or wider, a star cluster or nebule, the whole lot can all be done with DSS.

Make sure you shoot in RAW mode too, as you will need it to ease out so much hidden detail. With this you will be surprised what you can get detail wise with short exposures as astronomy is concerned by just using the camera and normal lens.

Link to comment
Share on other sites

I have been using DSS for a while but still have no idea as to what the difference is between dark frames and dark flat frames i dont think the FAQ in DSS helps much either it talks about both in the same para.

Alan

Flat frames are taken with the scope and the imaging gear facing a uniform flat source of white light at the same focus point and camera orientation as the main light subs, this flat frames depending on the light source would have an exposure time of lets say X seconds,, if this exposure time is long enough to introduce noise in the flat frmes then these need to be removed as well as a bias subtraction. Dark frame subtraction is only applied to the light subs so a dark flat subtraction needs to be applied to remove the noise from the flat frames, it goes without saying that the exposure needs to be the same as the flat frames but with the scope covered. If dark flat frames are taken then I believe a bias subtraction is not performed as the bias is already in the dark flat frames and once subtracted from the normal flats then both dark noise and bias will be removed . The darks will then take care of the light subs as dark frames will too have both the noise and bias in them. I hope this helps a bit.

Regards,

A.G

Link to comment
Share on other sites

Flat frames are taken with the scope and the imaging gear facing a uniform flat source of white light at the same focus point and camera orientation as the main light subs, this flat frames depending on the light source would have an exposure time of lets say X seconds,, if this exposure time is long enough to introduce noise in the flat frmes then these need to be removed as well as a bias subtraction. Dark frame subtraction is only applied to the light subs so a dark flat subtraction needs to be applied to remove the noise from the flat frames, it goes without saying that the exposure needs to be the same as the flat frames but with the scope covered. If dark flat frames are taken then I believe a bias subtraction is not performed as the bias is already in the dark flat frames and once subtracted from the normal flats then both dark noise and bias will be removed . The darks will then take care of the light subs as dark frames will too have both the noise and bias in them. I hope this helps a bit.

Regards,

A.G

Thanks for that.

I follow most of it but not sure how you would create a dark flat frame or is  it  something that is derived from the others.

Alan

Link to comment
Share on other sites

It is a dark frame with the same exposure parameters as your flat frames - just as a "ordinary" dark frame has the same parameters as your Light frames

Ah that makes sense so the dark flats would be taken just after the flats once you had worked out what the settings where (i use  AV mode for flats).

Thanks

Alan

Link to comment
Share on other sites

Ah that makes sense so the dark flats would be taken just after the flats once you had worked out what the settings where (i use  AV mode for flats).

Thanks

Alan

Hi Alan,

You won't need them unless the exposure for the normal flats is something like 5~10 seconds.

A.G

Link to comment
Share on other sites

Brilliant stuff - thanks everyone!

I've spent a bit of time wandering around the forum and there was something else I forgot to ask as well (not sure if it warrants a separate topic though...)

A number of astrophotographers mention mixed exposure 'lights' such as so many 15 sec, so many 30 sec and so many 60 sec exposures etc, combined to create an image - possibly in Registax? I was wondering about this, and I was prompted to ask because I've seen some great DSO images taken with my model camera using stacked HDR exposures (which combines up to +/- 3 stops exposure into a single image). I have various lenses up to 600mm and a spare dovetail so I can mount the camera directly onto the EQ mount without resorting to using the 'scope tube.

Any thoughts would be welcome :-)

Link to comment
Share on other sites

The reason some folk take various length subs is to capture images with a high dynamic range (very bright bits and very dark bits on the same target !!).  M42 the Orion nebula is a good example.  If you take "long" exposures the central core of the nebula will be badly overexposed as it is very bright but you will capture all the outer faint stuff.  If you then take some "short" exposures you will get the core but not get the outer faint bits.  So you stack each set of images separately and then process each of them.  Finally you overlay one image over the other to get a good balanced image of the whole nebula without any pure white over-exposed bits. 

Link to comment
Share on other sites

more photograpghy way of explaining. 

Lights are the images you take of your Subject. 

Darks are an exposure of a dark image, this captures the cameras generated heat and noise.

in daytime photography we dont worry much about subtracting dark images from our taken light frames. 

astrophotography the more dark removal the better an image is with better SNR ( signal to noise ratio).

Flats are akin to a gray/white card custom white balance setting. here you capture a Flat all over image.

this helps remove veingeting (sp) and any dust bunnys you may have picked up. taking flats

you can pop a white t-shirt over your telescope tube and image the flat white you get.a good few of each dark and flat frames

subtracted from your stack of light frames, you have a nearly noiseless image,

the noise is the dead/hot pixels and the onboard amp glow ( graining at high iso we get when pushing iso hard in overcast light)

i tried flats and bias it just made things more complicated for me. so i run with around 15-50% of 

my taken images, 100 shots i would run off another 30-50 darks ( taking my darks at the end of my session) here the camera is producing maximum noise

and heat.  and i cheat to, i have the program make a master dark frame up.

it takes all your 50 dark frames, works its magic and makes them into one single master dark., import it with your light frames and less filesize

and images used. 

Link to comment
Share on other sites

The only dark you need for your flats is a master bias. (A stack of, say, 50 bias frames averaged into one new one.) A master bias will be statistically indistinguishable from a dedicated dark taken with identical settings to your flats. You do need to dark subtract flats or they will over correct, but a bias will do the job perfectly and save you time.

Flats are very important and not always dead easy to get right, but you do need them. Working without them is often absolute murder.

Very few (hardly any) objects need multiple exposure lengths. I think I've used them two or three times, the obvious one being M42 as stated above. Usually your long exposure data contains the signal you need but a routine stretch may just saturate it in post processing. You can make a soft stretch of the same data and blend it with the hard stretch in which the saturation has occurred.

Olly

http://ollypenrice.smugmug.com/Other/Best-of-Les-Granges/22435624_WLMPTM#!i=2266922474&k=Sc3kgzc

Edit. Actually one caveat to the 'no need for short exposures' point. Sometimes I do use my 'RGB only' on bright parts of an image, usually stars, because the colour filters pass only about a third of the light so can be seen as short exposures. They are, in effect, shorter exposures than my luminance. But, still, multiple exposure lengths are rarely needed.

Link to comment
Share on other sites

Yes, but why would you? It uses the bias for the flats anyway.

NigelM

Does it? Interesting. I'm now totally confused with these dark flats.

I got totally different gradient results when using my master bias as bias or as a dark flat.

Link to comment
Share on other sites

Don't want to nick the thread but I tried using a monitor as a light source for flats.

The flats turned out blue with AV mode, had to expose for near 10secs for white........thoughts please thanks.

It does not matter what colour the flats are as only the Lum is used for correction ( flats are not used for colour correction, that is the job of the software ), however none of the 3 wave lenghts should depart from linearity, the RGB channels must remain linear.

A.G

Link to comment
Share on other sites

I got totally different gradient results when using my master bias as bias or as a dark flat.

As far as I know, DSS master flats and master darks are stored with the bias subtracted.  There is a flow chart at http://deepskystacker.free.fr/english/theory.htm#CalibrationProcess which shows how it works.

I guess if you give it a master bias as a dark-flat  you end up with an extra bias subtraction in there somewhere, which might explain why you get different gradients.

You shouldn't really need a dark-flat unless you are taking very long flats (or have a very bad camera!).

NigelM

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • 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.