M42 Help

[JohnSadlerAstro]

So I might be able to get the same kind of thing with an EOS1000d, driven EQ5, and 130mm (focal length 900mm) scope?

[dannybgoode]

11 minutes ago, wornish said:

You can get great shots with unmodded DSLR cameras.

Here is a shot I took of M42 using a Olympus E-M5mk2.

Nice photo. Really nice. 

Thanks also to those on the heads up on the site I mentioned. Well worth heeding.

I'll at least have another read of his work with that in mind as it's good to dissect such things. 

[JohnSadlerAstro]

Hi,

Just a report on current progress: I'm running a re-stack of my data, and I found a couple of things: The scope seemed to have shifted halfway through, chopping off the Running Man, and secondly, only around 25% of the frames didn't show trailing. I'm stacking 13 frames, around 10 minutes in total.

John

[wornish]

52 minutes ago, JohnSadlerAstro said:

So I might be able to get the same kind of thing with an EOS1000d, driven EQ5, and 130mm (focal length 900mm) scope?

Certainly possible no reason why not.  Make sure you are polar aligned as accurately as possible otherwise you risk getting elongated stars at the longer exposures.

Remember the more shots the better for stacking helps reduce noise.  

[ollypenrice]

Looking at your picture I don't agree that there is a general lack of sharpness. The sharpness is quite reasonable. There are clean edges in places where the nebula has colour changes, for instance. The curled 'hook' of gas coming out of the hole is also showing crisply. Nor do you need exotic wavelengths for this one. M42 is a strange target in many ways, one of them being that Ha can blur some of the nicest features. My own favourite version is LRGB only.

There are things to watch in processing, though, and the first is not to black clip.

This image is healthy. Note the flat line to the left of the main histogram peak and the relatively light background sky.

 

The one below is not healthy at all. It is an amputee!  The black point slider (left hand side in the input levels) has been brought way too far to the right and 'black clipped' tha faint data. The sky is flat jet black. The histogram peak is missing its left hand side.

 

It may be very tempting to do this in order to clip out light pollution and colour gradients but, whatever you do, resist this temptation because nothing can be done with the image after you've clipped it.

The data is gone, and with it half the image. How should you deal with gradients (including vignetting)? 

1) Take flats. There is no way out of this, and take a set of bias to use as darks for flats (AKA 'flat darks') You don't need 'proper' darks for flats, a master bias will work just as well.

2) Rectify the gradient rather than clipping it. Pixinsight has Dynamic Background Extraction. For Ps you can buy Gradient Xterminator, a plug-in from Russ Croman. I don't know about Startools but I bet it has something.

3) Balance your colours by getting the top left of the histo peak aligned in each channel. You have to use the black point slider to move a histogram peak to the left so you must have plenty of flat line on the left available or you will black clip that colour.

There is a lot to learn in processing but I hope these pointers are of use.

Olly

 

[JohnSadlerAstro]

I wish my shots looked clear like yours and the one in Making Every Photon Count. I just always end up with this kind of thing. And there's just no neat background of an kind. I think I must be doing something fundamentally wrong somewhere.   I suppose lights have something to do with it, I'm suffering from awful gradient, how would you suggest I should take them? 

[Filroden]

If you want to share the image that comes straight out of your stacking software we could try and process it and give you a step by step guide of what actions we took.

[JohnSadlerAstro]

Certainly!

Its just stacking now.

[JohnSadlerAstro]

Hi,

Here it is: no changes applied in DSS (although I usually do apply as Making Every Photon Count advises to do so for GIMP users).

John

Stacked3.TIF

[happy-kat]

Re reading this thread again I think that is a stack of just 13 images, there is no mention of flats, darks or bias files. You have to collect the data to get the data to process. Don't lose heart keep going keep collecting.

[JohnSadlerAstro]

You're right, perhaps it will be better when I have collected more.  

[MattJenko]

If you still want to play or add extra data, then astrobin's public data pools are a nice source. Here is one example of some M42 publicly available data : http://www.astrobin.com/rawdata/publicdatapools/70/

[happy-kat]

Had a quick play in StarTools. the light gradient on the side was wiped fairly cleanly, not sure about the colour. I did pull more nebula out on a second go but the colour cast got really strong around the edges. When I opened the image in PSPro afterwards it was clipped when checking levels after a startools process, odd.

Edit: on checking the tiff file this morning it was already clipped before I had a play.

[ollypenrice]

OK, I've had a look at the data. As Happy Kat says, it does look as if it's clipped from the off...

But is it? Not sure. Given one simple log stretch it looks a lot better, with the three colour channels miles out of balance from the LP, as expected. This does not look clipped.

If we run it through DBE it looks better still.

With a little stretching and colour intensifying (not using saturation) it can give this.

So why did it looked clipped? I"m not absolutely sure but I think that the scope is simply not illuminating all of the chip sufficiently. The brightest image circle is offset to the right and around the vignetted edges the background sky is only about a third as bright as it is away from the edges. Flats would help this but this is an enormous fall-off in brightness. In our most vignetted rig we still have 75% signal at the corners.

I don't think the image has much more to give, though someone else might do better. Even so, there is a lot to like about it.

Olly

[JohnSadlerAstro]

Thanks! I will be gathering more data ASAP, and then perhaps I will be better. I think I was probably uderexposing, but better PA should sort that out.

John

[Filroden]

Hi John

Sorry it's taken me a while, but I've had a look at the data you posted. I found it quite hard to tackle the gradients. As Olly's noted above, there is a significant vignette which is offset in addition to other gradients. My skills with DBE were not up to the job of dealing with both. Nonetheless, I quickly processed this in PixInsight with a DBE to remove some background, BackgroundNeutralisation (which didn't do much as DBE had done a good job already) then ColorCalibration (again, with minimal effect). I did an initial stretch with MaskedStretch then the remainder with a HistogramTransformation and some final curves adjustments to increase contrast. I applied no noise reduction as I was just wanting to show there was good data hiding behind the gradient including the blue emission area around the Running Man.

So good luck with the gathering of more data and I can't wait to see more versions of this!

[JohnSadlerAstro]

Hi,

Wow! I almost got to that in my last process, but there's still quite a lot of work to do! I think I may go ahead and buy PixInsight, as working with GIMP 8-bit is probably causing a lot of my trouble: especially in the areas of colour and contrast.

£175 don't look to good though....when you're on £2 pocket money...!  I think I shall have to wait until July and my birthday--raiding savings accounts are frowned upon in my family.  

When I've had my hernia operation and my intestines are correctly Polar Aligned I shall have to get a paper round or something like that.  

John