M27 with and without flats

[astrovirus]

Hi there,

Here is my first attempt of autoguided shooting and full processing with callibration files. Shot M27 with 20x 5 min subs, and callibrated with only 3 darks, 11 bias and 4x 20 flats (rotated 90° for every set of 20). Capturing, callibration and stacking with Nebulosity 2.3 and further processign with PS CS 3. There is still some gradient present which needs to be removed. Could this be from not fully effective flats, or is this remaining light polution? Processing after flats were applied was much easier in PS som the flats must have had some good effect.

Thanks for watching and any comments are welcome.

Greets Tim

[adamsp123]

I think you need a lot more darks (a small number will add noise) and more bias frames, why the 4x rotation for the flats, they should be taken with the set up used for the subs with nothing moved for them to work.

You have some nice data which I am sure you can get more from with improved technique. This is not a critism BTW because I have been there done it! in fact still a long way to go for me.

[George]

Most of the gradients looks like vignetting (the tunnel effect you can see) I'd take a lot more flats and create a new master flat without rotating them (never been a fan of this)

Were these flats taken with a lightbox or were they skyflats?

You need more darks too & a equal number of bias.

Have a read here

Not a bad image all the same

[ollypenrice]

Never heard of this 'flat rotating.' What's the thinking behind that? Surely you have to shoot through the setup as it was used for the lights?

I agree on more darks. Flats can introduce gradients because it is very hard to keep gradients out of them at the capture stage. Even using a panel or light box I only ever shoot them in the dark. Light gets in everywhere!

The gradient may be LP but we'd need to know the orientation of the shot. I assume LP (on which I'm not an expert) is worse on the horizon side than on the zenith side?

Does that square with your gradient? Russel Croman's Grandient Xterminator is very effective.

Anyway, a great start to guided imaging.

Olly

[George]

Olly

Ive heard of people with EL panels doing it for various reasons, to me it would not be a true flat as you would be losings its purpose ie recording defects in the imaging train (dust on various lenses etc)

[SteveP]

Hi Tim

A couple of good 'first' attempts - good focus, a well-centred object and good tracking.

You'll see some radial trailing on the stars which will be either a sign of a less-than-flat image or some chromatic aberration from the Newt.

As you know there is a gradient on the first image and some vignetting on the second. I would have expected both to have been removed by the use of flat fields. I suspect that the way you're producing flat fields has something to do with it. I'm unsure as to why you'd need to rotate the flat fields after each batch. It would be good to reprocess the images with say 20 flats all taken in the same orientation and median or SD combined.

Since you're using a DSLR, i'd also take a few more darks at a temperature similar to the lights

HTH

Steve

[yesyes]

When you say you rotated the flats, do you mean you rotated the lightbox before taking the image or you rotated the resulting digital image?

[ollypenrice]

Ah, now rotating the lightbox or panel makes perfect sense to me. If that has a gradient then it will be compensated for. However, what I thnk has happened to me on occasion while shooting flats is that light gets in between the box and the lens and creates a gradient, in which case rotating the box won't help. That's why I shoot them in the dark or semi-dark.

Olly

[astrovirus]

Never heard of this 'flat rotating.' What's the thinking behind that? Surely you have to shoot through the setup as it was used for the lights?

Thanks everyone for the comments and helpfull suggestions so far. I'll try and explain my workflow regarding the flats for this image.

With rotating I mean I rotated the lightbox after every 20 flatfield frames taken. With this I hope to remove any uneven illumination in the lightbox which may be present, during stacking for the masterflatfield. In theory this should provide an evenly lid masterflat with all the dust motes in the same place in ever single flat frame as the rest of the imaging train remains intact.

Furthermore, I agree with everyone regarding the number of darkframes, they should be way higher, however when taking them I had the idea something was wrong as the showed up on the monitor looking really light whereas I expected quite black frames with noise in them. However this may be the result of doing darkframes for the first time with nebulosity which captures en displays RAW frames in black/white as the actual masterdark looks OK when converted to color, so next time I'll just ignore the way the look during capture and shoot a lot more. For this time, as I was thinking somethong was not right I just stopped as it was 4:30 AM and I was really longing for my bed at the time .

Bias frames was the same story, looked really light whereas I was expecting really dark frames.

As with the remaining gradient, I'll try and see if I can establish my orientation of the frame and interpret that with regard to the major light sources in the area.

And again thanks for the comment so far,

Greets Tim

[astrovirus]

One final shot with the remaining gradient removed.

[ollypenrice]

Aha, nice!

What remains is to look at colour balance, which is strongly biased towads the green.

In Ps begin by looking in Levels at each colour channel separately. As you scroll through R,G and B the top left hand side of the peak should be in the same place for each colour. Cut back on the left (without clipping any data) if you can. If you don't have enough room on the left of the peak to do this then drop the green channel a little in Curves.

Olly

[astrovirus]

Ah, color balance, that's my main problem as I am very badly color blind. Have to do this completely by histogram as my perception of colors on screen is not as the rest of you will see it. Thanks for the line out how to do it, Olly, I'll give it a try. I had already noticed that in my stacks the green histogram i much stronger then blue and red. What I normally do in DSS is to bring them all together, and then export to CS3, however I also noticed on the pixel info in CS3 that there still remains (too) much green.

[Astroscot2]

An excellent final result, nice sharp focusing and good guiding