Mysterious histograms

[chiaroscuro]

Forgive me for reposting some content here rather than the beginners astro forum, but I'm desperate for some advice.
I'm a beginner in astrophotography, using a Canon 70D with a 120mm refractor. My issue is that after taking 30 min of data of M42 (60 x ISO1600 30sec subs), the subs look fine but the RGB histograms lying quite to the left, but after stacking in DSS with darks and bias files, the stacked image is terrible with the dotted line crossing diagonally across the histogram (I'm not sure what it signifies) ie., not a normal sigmoid shape like a wave front.
I've attached screenshots of the camera's RGB histogram of a single unprocessed sub from the camera, and the stacked files with its histogram. The last screen shot is a single sub in detail.

Any pointers would be gratefully accepted.

Luke

[lazjen]

Referring to your stacked image (#3) with the RGB levels, if I'm using DSS, what I would do from there would be:

* Checked the "Linked settings" box
* Grab one of the middle triangles and pull it to the left until the solid curve is near the left - don't go too far otherwise you'll starting clipping.
* Switch to the luminance tab and then adjust each of the settings there until the current diagonal line becomes a curve to mostly cover the solid curve.
* Switch to saturation tab and adjust if necessary.

You can get a result from this to export if you want.

A better idea is to take the stacked result you've got and process it further in another tool, like PS.

[chiaroscuro]

Hi Chris,

Thanks for the reply. Here is the image after adjusting the RGB, saturation and luminance in DSS. When I tried in Startools and Nebulosity, I couldn't really improve the images contrast or colour balance (second screenshot)
I'm suspecting that maybe I just need to take longer exposures than 30 seconds at ISO 1600.


Cheers
Luke

[lazjen]

I think you need to do more adjustments in the luminance tab - the diagonal line should be more a curve to get better results. Really do some experimenting in this tab to see the differences.

You may need to also check your settings in DSS to ensure it's correct for your camera. Also ensure you're using the latest version (3.3.4 at least).

[ZeroID]

Nah, 60 @ISO1600\30 secs should give you a usable image. I've done it with far less. It's a weird histogram for sure. I'd shift all the raw files to a new folder, dump the txt files and restack. I find that sometimes I need to close and restart DSS to give it a clean slate or weird things happen. Also use the correct version of DSS, or try the latest beta if there is one.

[multiweb]

I'm not familiar with DSS but, from looking at your raw histogram, the data looks fine. I suspect the black point is clipped when stretched, so try to back it down. As a general rule, don't go by what the picture looks like in the preview window, but keep your eyes on the histogram and the distribution of all the samples, making sure you leave enough room on each sides of it for further processing.

[chiaroscuro]

Thanks Chris. I've moved the luminance and saturation tabs all over the place, and really can't seem to improve the image there either. I am using DSS 3.2.2 via wineskin, and I'm not sure if that may be causing the problem, but other Mac users seem to be OK running it through wineskin.

Is there a setting that you use to indicate what camera your using? I can't find that under any of the menus.:question:

Hi Brent, tried doing that too. I'll look on the web for a newer version of DSS. I've just upgraded my memory to 16GB, so I'm hoping that may improve things - at least I'll get my terrible stacking results a lot quicker!



Hi Marc,
What do you mean by the black point being clipped?

Thanks for your input. As you can see, its all a bit of a mystery to me, and I'm really just tweaking the buttons trying to improve the image without really understanding what I'm doing. At least its forcing me to do a lot of reading and learning about DSLR's, histograms, colour spaces etc..

Thanks
Luke

[multiweb]

It means the left side of your histogram is clipped. So there is a sharp vertical line on the left truncating the bell shape of your histogram. You're losing all the dark areas of your picture.



I'd recommend saving the file as a 16bit FIT or TIFF file without stretching it in DSS. Then open it in Photoshop if you have it and start stretching it progressively in increments. I've PMd you. If you're still confused email me the TIFF file and I can send you back a PSD file with all the layers so you can use that as a template for your future photos.

[multiweb]

Ok, your histogram looks alright. Nothing is clipped. One of the problem you're facing is that you have various subs with different orientations so you're going to have to pick an area to crop around the nebula.

[chiaroscuro]

OK, that makes sense about the clipping. I did try to do as you say using Startools, and then Nebulosity, without much luck. Here is a dropbox link to the files - rustigmed (Russell) has offered to try processing it too. Got your PM and will have a look at that link - thanks again.

Link: https://www.dropbox.com/s/srxpsfu4xn5oy5v/Autosave.fts?dl=0

Cheers
Luke


PS: To the mods, sorry to have two threads running concurrently on this topic but the other one went quiet for some time (the other one is in Beginners astrophotography, so if you want to merge them, go for it).

[chiaroscuro]

Hi Marc,

Is it quite important to have the camera orientated the same way with every sub taken? It's obvious that there is stacking artefact in mine, but I thought that would be taken care of in the post-processing.

Cheers
Luke

[Camelopardalis]

As close as possible, to minimise the area you need to crop :) I'm not very good at it but usually manage to get it within a few degrees, which leaves me with "widescreen" ratio images :lol:

Btw, what focal ratio is your scope? That could have quite an impact on how bright the image is looking, relative to what you might be seeing (and expecting) from other peoples' posts. Some folk are using f/4 and f/5 scopes which would be noticeably brighter. Just trying to set your expectations ;)

[chiaroscuro]

Sorry to be too-ing and fro-ing between the two threads. the focal ratio is 7.5, aperture 120mm. I was considering getting a focal reducer, but I'm still looking at the implications of that.

[Camelopardalis]

No worries, I know how frustrating it can feel, and with the weather the past few months there have been precious few opportunities to be imaging!

So here's the thing with the focal ratio...the images should be brighter the faster the scope is. But it's a fine line. The faster scope you get, the shorter the focal length is, the wider the FOV becomes, and the resolution decreases as a result. There's no perfect scope. Maybe there is, I just haven't found one at a price I can afford ;)

But faster is quicker...if you can't go faster, go longer...i.e. increase your exposure time. Simples!

[multiweb]

Well, if you have a lot of subs with different orientation you can get away with it but the overlap will always have a better SNR than the rest and it's harder to process out. So ideally, yes, you should keep the same orientation from one imaging session to the other.

[Camelopardalis]

Btw, the implications of the focal reducer will usually not just be focal length reduction...that actually may only be by a factor of 0.8x or 0.85x, which doesn't sound a lot. But often - just not always - focal reducers are combined with field flatteners, which flatten the focal plane out so that stars out towards the edges of the image don't look out of focus. Note, I'm not saying yours do, I'm just saying it's quite common for this to happen with fast scopes and/or large sensor chips, and a flattener is how it can be remedied.