Auto Dev in StarTools

[Mikey0368]

My set up is relatively new and I'm dabbling with some things while I impatiently wait for the clouds to clear and the skies to get a bit darker. I took 10 x 120s images of M31 last weekend and I've stacked them and tried to run them through StarTools. I've attached a couple of screenshots to show what the FTS file looks like and then how it looks once I hit the AutoDev button.

When I compare the AutoDev outcome that I get to a YouTube tutorial it's nowhere near where I'd like it to be.  Is it simply because I only used lights and there are only 8 images that have been stacked?

 

[almcl]

Autodev at the outset is intended to show up problems within the image. 

Crop (if necessary) followed by bin (again if required) then wipe, using which ever preset is appropriate, then a further autodev or dev and then the rest of the modules as required in the order shown in the video here: https://www.youtube.com/watch?v=IDY3ov2dQe4

Your data looks (although it's hard to tell from just the screen shots) as if you might well have some stuff there once the colour cast is removed.  

 

[happy-kat]

Yes autodev is great to show how much to crop then if my data is thin I'll cancel them go crop then bin 50% then wipe then manual develop then colour as a first pass of what I've got.

There are preferred settings to follow for getting the best from DSS for then using in startools, I don't have the link to hand but basically don't select align rgb boxes.

[Mikey0368]

Thanks folks. It was clear last night so I now have (what should be) better data to work with. I have a .FTS file that is 24 x 120s stacked images and I'm hoping there's something worthwhile hidden away in there!

I'll try it as per that YouTube video and see how it goes........

[Mikey0368]

Well here's a screenshot of what I got from last night's data and it's fair to say that it aint good. It looks like someone has taken a sander to it! Any pointers on where I'm going wrong would be appreciated. The images were taken using........

SW Esprit 100ED, SW HEQ5 Mount, Flattener (which is brand new and the first time I've used it), Modified Canon 600D, 24x120s and ISO800.

I also have a decent .FTS file of M42 from last February so I'll run that through StarTools as well and see how it comes out.

Edited August 18, 2019 by Mikey07

[prusling]

This is the workflow I use for DSLR; for other cameras I adjust the Wipe params (e.g. Aggressiveness 83%, Drop-off Point 50%).  I don't think your result is caused by 'Canon banding' but if so, there are ways of fixing this.

STARTOOLS EXAMPLE WORKFLOW

Command	Action	Params	Notes (for DSLR imaging)
Open	Open the stacked image from DSS (ideally from fits format)		Use left option on opening - DSS actually does some white balancing which can't be disabled
Bin	Scales down pixels and noise and reduces subsequent processing time	50% (default)	50% converts each 2x2 pixels to 1 pixel
AutoDev	'Digitally develops' the image - stretches it to show faint signal and issues to be addressed	Drag RoI around target	Will probably look horrible but this shows you the worst of the image that the processing will fix
Crop	Discards unwanted border around target (and stacking artefacts)	Drag area to be left	Best to eliminate worst of gradients, vignetting, etc around the edges
Wipe	Removes light pollution, gradients, vignetting and colour casts	Drop off point 0%	Say Yes to Temporary AutoDev as well and maybe a few pixels of dark anomaly filter
Dev	'Digitally develops' the image - stretches and applies curves	Re-do global stretch	Use "Home in" or manual % to stretch until best compromise achieved (detail vs. noise)
(Decon)	Compensates for atmospheric distortion (e.g. tightens stars)	2 pixels	Optional - set Mask manually first using Clear, Auto, Stars, Do, Shrinkx2, Invert, Keep
(Sharp)	Sharpens the image using wavelets		Optional
(Contrast)	Increases local contrast		Clear mask from Decon beforehand (Mask, Clear, Inverse, Keep)
(HDR)	Optimises dynamic range - replaces manual curve manipulation	Choose algorithm	Optional - apply algorithm to suit the target - e.g. Tame, Reveal, Optimize, etc
(Life)	Brings back life into an image by remodelling uniform light diffraction	Moderate	Optional - not sure how natural this action is - may border on artistic licence!
Colour	Adjusts the colours of the image (e.g. compensates for modded DSLR)	Fill mask if not already	Can adjust to suit - presets for scientifically accurate or artistic (latter can seem more natural!)
Filter	Modifies features in the image on the basis of their colour	e.g. Ha	Can boost nebulae
Track	Switch off tracking in top menu to reduce noise and finish image	10 pixels?	Initial example screen makes it look as though everything will end up fuzzy, but it is selective
Save	Save resulting image to storage	Need to put type e.g. TIF	Not supported in trial version
Key commands are in bold
N.B. Can re-do the AutoDev (or manual Digital Dev) any time before tracking is switched off and it reapplies the other changes made

Edited August 18, 2019 by prusling
Corrections

[Mikey0368]

Thanks for that. I'll try that on my M42 image before trying M31 again.

[Mikey0368]

Here's what it did with the images. M42 was taken last February using my 8" reflector and M31 was last night with the SW 100ED Triplet. Clearly one is much better than the other!

The M31 image is much improved but there's still very little detail.

 

Edited August 18, 2019 by Mikey07

[happy-kat]

When using wipe do use the mask tool on any dubious areas. See if there are any to remove from your image. Did you take flats?

[bottletopburly]

The key to processing is good Data , take time to get darks and flats  get the latter two and you’ll start to see a difference and more data the better .

[prusling]

The better-calibrated the image (darks, flats, etc) the more you can stretch it to obtain fine detail; my crib sheet for calibration frames is below (any comments very welcome).  Darks are pretty well essential with a DSLR, especially at this time of the year; also it is best to turn off LiveView when imaging as that adds to the noise.  I also cover the optical viewfinder to avoid stray light entering.

I judge the exposure time and ISO according to the darkness of the skies and the camera.  I have a Canon 550d (precursor to your 600d) and use 1-minute subs with ISO800 in dark skies and the same exposure with ISO400 in light pollution at home.  This means that you can afford to discard poor images and use a sigma clipping algorithm when stacking to eliminate outlier pixels giving a cleaner image.

	Lights	Darks	Flats	Dark Flats	Bias/Offset
When		Before/after or can use a dark library	Ideally every session; can be next day	Optional if using Bias	Occasional
Number	The more the better	20+	25	20+	50++ (> Darks)
Format	RAW	RAW	RAW	RAW	RAW
Optical Setup		Capped	Exactly the same as lights	Capped	Capped
Temperature		As close to lights as poss	any	As close to flats as poss	any
ISO	400 (LP) or 800 (dark sky)	Same as lights	Same as lights	Same as lights	Same as lights
Exposure time	Can vary with tracking / guiding, I normally use 1 min when tracking but not guiding	Same as lights	AV mode until 50-70% histogram then set manually	Same as flats	Shortest e.g. 1/4000 sec
Lens F Stop	Widest / 1 click or more back	any	As used for lights	any	n/a
Lens Focus	LiveView x 10 on a bright star / distant target	any	As used for lights	any	n/a
Stacking algorithm	Kappa-Sigma if >15 frames otherwise Average	Median	Median	Median	Median

[Mikey0368]

Thanks everyone. That's this weekend's project sorted if skies are clear!

[jager945]

On 19/08/2019 at 01:50, Mikey07 said:

Well here's a screenshot of what I got from last night's data and it's fair to say that it aint good. It looks like someone has taken a sander to it! Any pointers on where I'm going wrong would be appreciated. The images were taken using........

SW Esprit 100ED, SW HEQ5 Mount, Flattener (which is brand new and the first time I've used it), Modified Canon 600D, 24x120s and ISO800.

I also have a decent .FTS file of M42 from last February so I'll run that through StarTools as well and see how it comes out.

AutoDev serves a dual purpose; first it is used to show any defects in your image, then it is used to perform your final global stretch after you have mitigated the issues you found earlier.

In the above image, we can see gradients, dust donuts and (rather severe) walking noise. Flats are the #1 way to improve your image (and processing enjoyment! 😀), with dithering between frames a close 2nd (this solves the walking noise issue). They are really not optional and they don't really cost anything, except some time.

Also don't forget you can (should) click and drag a Region of Interest for AutoDev to optimize for (by default the RoI is the entire image). In the case of the image above, a slice of M31, including its core would constitute a good dynamic range sample we'd be interested in.

There are also cases where AutoDev simply cannot lock onto celestial detail. These cases usually involve heavy noise and/or lots of "empty" background. The latter can be solved with an RoI. The former can be solved by increasing the "Ignore fine detail <" parameter, which makes AutoDev "blind" to small noise grain. If you cannot fix or work around the calibration/acquisition issues in your dataset and cannot obtain a good result with AutoDev, you can always resort to a manual Develop.

As bottletopburly already highlighted though, post-processing becomes infinitely easier and repeatable when you have "good" data. I'm putting "good" between quotes here, because it just means "properly calibrated". It doesn't have to be deep, gradient free or noise free, as long as it has been dithered and properly calibrated with - at least - flats.

Hope this helps!

[happy-kat]

There are 4 dust bunnies I'd have masked out when doing autodev and the aircraft line.

[Mikey0368]

I'm pretty sure I've got good data from last night's crack at the Witch's Broom. 90% of 186 x 30s Lights plus 40 each Bias, Flats and Darks all stacked in DSS.

AutoDev is much better but the finished article isn't! Will keep plugging away at it.

 

[Mikey0368]

This is much better, more through trial and error than anything else!