StarTools

I imaged the amateur challenging Iris Nebula.
I attached 2 processing attempts, one looking somewhat normal but lacking details.
The 2nd one is "interesting" in that it shows some aspects of the complexity of the dark gas in the nebula, but abundant noise.

In the "normal" one, I tried to specify a ROI, but it can be seen that Wipe removed signal.

In the "interesting" one, I tried to follow Mr. Jager's instruction to "lasso" the area with its dark lobes. But I didn't really know how to follow through. The outside areas are thus filled with way too much noise even though I then heavily cropped the image.

Can anyone suggest in simple terms a strategy that I should have used to protect the nebula detail while cleaning up the noise outside of it?

I realize I continue to lack skills, and masking isn't intuitive for me, but I would much like to improve.

arrowspace90 wrote:In the "normal" one, I tried to specify a ROI, but it can be seen that Wipe removed signal.

Hmmm... What leads you to believe Wipe removed signal? (it shouldn't, unless some very high Aggressiveness was used!)

Can anyone suggest in simple terms a strategy that I should have used to protect the nebula detail while cleaning up the noise outside of it?

I'm a little confused... What exactly do you mean by "cleaning up the noise"? Wipe does not clean up noise, nor are there any modules that clean up noise that require a mask to do so?
Selective processing is really rarely needed when processing in StarTools. The only exception is masking out stars during Deconvolution or Wavelet Sharpening. The other exception is Wipe - in extreme cases - when the data contains dark anomalies, or when Wipe threatens to remove large scale nebulosity that is not broken up at all by darker nebulosity or detail (a rarity, unless high Aggressiveness values are used).

It would be very helpful if you could upload the dataset, however at first glance it appears there is simply insufficient signal to bring out the darker parts of the complex. Your first rendition seems quite reasonable, though a slightly wider RoI could probably show a bit more of the Iris nebula itself.

It is likely you would also stand to gain some detail from using deconvolution.

Please feel free to upload the dataset so we can better help you!

Well I attempted to upload the zipped dataset, but it must exceed the limitations of the forum available bandwidth.

You could use something like dropbox for the fits file, there are others

Hi,

Thank you for sharing the data with me.
The issues you are having are caused by not (or incorrectly?) calibrating with flats. They are not optional and absolutely essential if you wish to tease out faint detail.
Just opening up the dataset and performing a first diagnostics AutoDev yields this; Attempting to remove the vignetting in Wipe makes clear how uneven the background really is; Instead of trying to address these issues in post, first order of business should be to get acquisition down pat and eliminate these issues during acquisition, stacking and calibration. Unfortunately, right now, any faint celestial signal is overwhelmed by uneven lighting artefacts and streakiness. Without a dataset you can trust in terms of signal, it becomes really hard and frustrating when trying to process it and corroborate faint detail with other renditions from other people.

Do keep us up to date with progress on the calibration frames!

Ok. Thanks.

I thought that might be the response. I have actually tried pretty hard to take flats. I went from sky flats to tee shirt/LED screen flats. I went from overexposed at 10 seconds each to something currently around .05 seconds. I have tried to refer to the historgram to find the highest point of the light curve and move it slightly left of center. This isn't very easy to do. I am considering taking several sets for each session and trying each of them in turn, not a fun prospect.
I have found that no 2 people take them the same way or much agree about how to get them right. It's a murky area for such an apparently important thing. The RASA I use is supposed to be good at achieving flat fields to begin with.

I got an image of this Nebula that I thought was ok "for me" after several attempts at processing.
Thanks again for responses.

Flat field is more about the flatness of the star shape across the entire lens surface.
Many lenses, or mirrors, distort the imaging shapes towards the edges and corners. FF are used for correcting a lens or CC for a mirror, or nothing if the device comes with a flat field

Flats are about the vignetting the lens or telescope creates over the imaging sensor.
Unless you are using a tiny sensor or a lens or telescope with a big imaging circle you'll get vignetting.
Flats tell the stacking software what the vignetting pattern is so that the stacked image ends up evenly lighted and without a bright centre and dark corners.

happy-kat wrote:Flat field is more about the flatness of the star shape across the entire lens surface.
Many lenses, or mirrors, distort the imaging shapes towards the edges and corners. FF are used for correcting a lens or CC for a mirror, or nothing if the device comes with a flat field

Flats are about the vignetting the lens or telescope creates over the imaging sensor.
Unless you are using a tiny sensor or a lens or telescope with a big imaging circle you'll get vignetting.
Flats tell the stacking software what the vignetting pattern is so that the stacked image ends up evenly lighted and without a bright centre and dark corners.

Thank you, so how do you go about taking yours? Do you have a reflector or a refractor. Obviously, they are different animals. If there is no exact way to do it, it's hit and miss. I guess I am missing. I have not ignored attempting flats. I work at them on every session but I'm guessing.

It is true that when talking about flats taking the opinions are very varied.

Each camera / telescope combination has its own flat model !. I match the peak of the exposure histogram of the flat with the peak of the exposure histogram of the object to be photographed that night. It works quite well for me, also Startools has in Wipe / Vignetting a very powerful tool that corrects vignetting very well if it is not too important.

Regarding your last proposal for treatment of the Iris nebula, I would say that the color has a greenish hue that you could correct and it would be even better. I use the Color / Max RGB option to balance the channels.

I have been paying more attention to flat acquisition. My software, ASIAIR can show a preview of what the scope sees along with its histogram. No doubt all of the software products do this.

I had planned to re-image the Iris last night to produce it with better flats. On a previous previous evening, I had my greatest success with the N. American Nebula. When I stacked the subs in APP, the raw image was almost a usable photo! Imagine that. I got this while using a LeNhance emission filter at .25 seconds after judging the histogram to put the top of the light curve in the middle.
Next evening, re-shot the difficult FireWorks galaxy. I say difficult to the extent that it is near the plane of the Milky Way and packed with stars. But to image it in the bright suburb, I switched to a LP filter. I was surprised that this apparently affected my flats. At .25 seconds, the peak of the curve was now way off to the right. I moved it back by reducing to .05 seconds. I guess I can understand this, but it caught me by surprise. I thought, "cool, just put it at .25 seconds every session".

Do filters influence the flats that much, or did I likely change something else as well? Anyhow, the clouds appeared at sundown, so no repeat Iris and I'm stuck here on the internet.
I also could try to just forgo the LP filter(but, Bortle 6 1/2)and use only the IR cut filter instead for other than emission nebulae. I read that some people leave their cameras attached between sessions so they don't retake flats. This would suggest I am wrong about what caused the requirement for a different exposure.