Terrible noise. Is it solvable?

[Xatrac]

Hi everybody here!

This is my first post here. I am pretty new at using PixInsight, and find it to be extremely powerful, but the learning curve is quite steep. I already have had some good successes with astrophotos taken with my equipment.
Anyhow, I have met a problem (and quite commonly, by the way). The attached picture (M2) has a terrible noise. The first processing it has, is simply DBE. The picture has been taken by a OneShotColour camera (I know that it might not be the best choice, but it is what I have). The picture is the integration of 9 frames (60sec).
The possibilities here are:
1.   Throw away the camera and look for a new one (for obvious reasons, this is not my first option).
2.   Take longer exposures, so as to increase the signal. This works better (having tried 150sec and 250sec), but the noise keeps appearing (less, because of less stretching needed when applying Histogram Transf). Anyhow, I would prefer (if possible) not to go for much longer exposures, because then the fainter objects would be out of reach, and I think that M2 is not underexposed in this picture.
3.   Try to solve this with PI.
Taking the third option, I have not been able to eliminate satisfactorily this noise. After DBE, I have applied ACDNR, and again DBE. None of the combinations worked enough (even those producing excessive blurring).
Can anybody help a novice?
Josep Drudis
C11 Edge HD
Orion Starshoot Pro

[Ioannis Ioannou]

Welcome to PI world,

Good imaging practices cannot be substituted by software. Meaning that you should keep the acquisition procedure (lights/darks/flat/bias at least, maybe dithering etc), especially if your camera is noise prone, is not cooled and it is OSC. The noise your image has is a combination of other things. For example calibration with flats has not been performed or was not successful, so it is very difficult to put ACDR to eliminate the noise in the background. BTW, DBE should be performed when the image is linear (i.e. before initial stretch) and ACDR after. MMT can be used for noise reduction when the image is still linear. I suggest watching carefully Harry's tutorial videos.

[Eddy Timmermans]

I agree with Ioannis.
When I made my first images, I took 90 frames of 90 seconds in one long session.
The result was a build up of a diagonal noise very similar to the noise in your picture.
So I started dithering and the noise went away for the most part. Then I realized that vignetting became a problem.
So I took every possible calibration shot, darks, flats, dark flats and bias.
And I took 280 frames of 90 seconds over 2 nights.
In the meanwhile I used drift alignment to align the mount as good as I could.
The result was a nice smooth background and much more room for processing. But PS didn't quite do the trick so I bought PI.

Just to show, you really need to do a lot of work yourself before beginning to use PI, but the rewards are great.

And once you know what to do and how to do it, it doesn't really take that much time preparing to take your pictures.

So I think more integration time (more pictures) more darks, flats and bias frames, would really help improve the S/N ratio of your picture and make processing in PI much more efficient.

Eddy

[Xatrac]

Many thanks to both!

The photo was calibrated only with darks. I will take flats and bias and recalibrate them.

Also the suggestion of dithering will be taken into account. Thanks again. I will come back with the results once done.

Josep

[Cheyenne]

Flats need to be taken at the same "time" as the image lights.  The reason being is that you need to accurately match the orientation of the camera in respect to the telescope and you need to match the focus.

My process is to take my images, and prior to readjusting focus, or orientation is to take the flats using a light-box over the telescope.

[martin_magnan]

Josep

Dithering will certainly help here. But take many darks (40) and many many many bias (200)

I have come to the conclusion that this pattern is caused by two things

1- flexure if you guide or drift if you dont
2- Noise "injected in calibrated images with noisy masters (dark and or biass) made with not enough individual frames.

Martin

[GJL]

Hello everybody here,

I´m new, this is my first post.

My question to Cheyenne, you wrote  "My process is to take my images, and prior to readjusting focus, or orientation is to take the flats"

How important is it for flats to have an accurate focus or is it not so important?

Thank you, clear skies from Germany   Gerhard

[Cheyenne]

One of the ideas behind flats, is that they are supposed to represent an image of the "flaws" within the optical path of your system.

For example, lets say you have a Newtonian telescope with a big blotch of dust on the diagonal mirror.  You take your images (lights), which will include the image of the blemish on the mirror.  You then take an image (the flat) using a flat background (e.g. a light box, or an image of the day time sky without any clouds, etc.) which will also include the image of the blemish.

You then mathematically remove the flat from your lights, which should remove the blemish from your image.  In order to do this properly, the flat have the same alignment as the light.  By changing your focus or by changing the orientation of the camera in relationship to the optics, you will change the alignment of the blemishes.

Hopefully that makes sense.

[Ioannis Ioannou]

What am I missing here? under what conditions refocusing changes so drastically the imaging train so to need different  flats for each position? maybe mirror flop/shift on a sct?

[Cheyenne]

When taking a flat, you want the lightpath to be exactly the same as when you took (or will take) the image.  Refocusing changes the lightpath.

If you refocus, then you have changed the focus of whatever is in the lightpath.  Try this little experiment with either a SCT or a Newtonian (or some other telescope that has an obstruction in the lightpath).  With an eyepiece, you can visually see the change if you focus way in then way out (this is an extreme example to demonstrate what happens), at somepoint you will be able to actually "see" the obstruction (such as the secondary mirror, or dirt on the lens, etc).

Found a thread on cloudynights that discusses this whole issue http://www.cloudynights.com/ubbthreads/showflat.php/Cat/0/Number/4632384/page/10/view/collapsed/sb/5/o/all/fpart/1

[Ioannis Ioannou]

Oh, com'on, . Read CN thread's results:

So far I would say that focus is very important for Newtonians, somewhat important for refractors and not important for SCTs...

1) Dust donuts are not significantly affected by focus.
2) An out of focus flat is much better than doing no flat at all.
3) Focus will affect your flats, so try not to change focus between your imaging and your flats. But don't worry too much about it!

and even for Newt scopes it was

focuser racked all the way in

. Who said that you will have to change focuser's position SO much before flats?

This is completely different of taking flat for EVERY focus change. In my case, most of the times, the difference in temperatures during session forces me to refocus every hour or every filter, whatever first. The total difference from start to the end of session is very small. Big enough to appear at the stars when zoomed, but not enough to change flats in any visible way. Taking flats in this case would be such a waste of valuable time for lights !! I leave the focuser to the last good position and I take flats with a lightbox while the sun rises.

[Cheyenne]

The racked in the whole way was to visually illustrate an exaggeration of what happens. 

Even in the CN thread, look at the statement:

3) Focus will affect your flats, so try not to change focus between your imaging and your flats.  But don't worry too much about it.

Do I take a flat between every minor focus change -- no.  I usually set up in the evening get focused, do my imaging with the minor focus adjustments for temp, and while I'm packing up, I'll do my flats.  However I do make sure that I don't "move" the camera in terms of rotation. 

If I were to really be working on a "project" and was really concerned about the quality of the image, I would spend the couple extra minutes taking flats after any adjustments (including changing filters in the filter wheel - there is always a possibility of a shift with the filter).  This may be overly paranoid, but I think the resulting image reflect the extra effort.

Getting back to the original point of the thread.

It's important to not alter the lightpath between taking flats and images.  Try to keep the focus the same, and don't "adjust" the orientation of the camera in relationship to the scope.  Of the two I would say the orientation is the most important aspect.