Please help with M33, 12 hours OSC and NO COLOR?

[oymd]

Good evening everyone.

After spending the past few months focusing on Nebulae, I had a go at M33.

Considering its a Broadband target, with some areas of Halpha, I decided to shoot it WITHOUT any filters, following advice I received on Cloudy Nights.

I am over 12 hours into imaging, and I am not happy with the result.

The image is very MONOCHROME, with barely any color.

The filters I have at my disposal are the Optolong L-eNhance, IDAS LPS D2 and the IDAS LPS V4.

My skies are Bortle 8-9, and I was advised to aim for 20 hours, but I do not see that any more data will bring any color into this project?

I was advised to use ELLIPTICAL GALAXY option in PixInsight's PCC, as well as ARCSINSTRETCH to bring out some color, which it did, but very very faint.

The scope I am using is not the best for the target, agreed, but surely after more than 12 hours I should be doing better?

Scope: WO Star 71 5 element Apo, ZWO ASI294MC Pro and NO FILTER. 120s subs, 400 subs so far.

Here is a screenshot showing my Master Light having had Dynamic crop, DBE, PCC, EX Denoise, Arcsinstretch and histogram stretch.

When I then proceed to Starnet, Curves and Saturation, the colors are barely there, and I somehow need to force color into the image.

You can see that in the histogram I have NOT MOVED the middle slider LEFT, as doing so BLOWS OUT THE WHOLE GALAXY!

Would using the IDAS LPs D2 help at all? or the L-eNhance?

Please help!!

I've been imaging M33 over 5 nights, and I feel beat.

 

[Scooot]

Somethings not right but there’s colour in it, lots of colours around the stars, but a lot of green. I’d do PCC on the stack first and have a look at it before you do any other processing. At that stage the correct colours should show quite well. If it still looks like it does now I’d wonder whether something was amiss during calibration & integration, maybe the wrong debayer setting? 

Edited November 11, 2021 by Scooot

[oymd]

I haven't changed my calibration routine and stacking in ages.

WBPP 2.2. ALL STOCK, except choosong GALXY in Weighting, and RGGB for all the Lights.

50 Fs, DFs, & Ds

The above was just before I applied SCNR.

Here's what it looks like after SCNR.

And here's the final image. Something's not right? Andromeda, that I shot last month, with the same exact setup was much easier and cleaner, with some great colors?

 

[ONIKKINEN]

1 hour ago, oymd said:

I was advised to use ELLIPTICAL GALAXY option in PixInsight's PCC, as well as ARCSINSTRETCH to bring out some color, which it did, but very very faint.

I don't have pixinsight, but this sounds bizarre. Elliptical galaxies being round featureless mostly single coloured blobs are about as far away from M33 as possible when it comes to galaxies.

Maybe try just "normal" photometric colour calibration (if there is one in PI?). SCNR also shouldn't be necessary after a good photometric colour calibration so something is wrong.

[Scooot]

 

“Here is a screenshot showing my Master Light having had Dynamic crop, DBE, PCC, EX Denoise, Arcsinstretch and histogram stretch.”

I always do Photometric Colour Calibration first before DBE, which is why  I suggested you try that in case DBE or any of the other processes had upset your colour balance.

[Xiga]

You might have more joy posting the stacked file to see if others have more luck. You also have to remember that, shooting Broadband in Bortle 8-9 is like playing AP on INSANE mode! Believe it or not, but 12 hrs of OSC is probably nowhere near enough under these conditions. You might need 30-40 hrs, a sobering thought if ever there was one. 

Saying that though, your screenshots do show quite a lot of colour i think. I use APP, and the order i usually do things in is: Background Neutralisation, then Gradient Reduction (DBE for you), followed by PCC (i sometimes use Siril instead of APP for this). Then remove the green, and finally do Arcsinh stretching to preserve the colour. 

HTH. 

[Scooot]

It’d be interesting to know what your background flux is? There’s a simple script in Pixinsight that measures it. It’s called Calculate Sky Limited Exposure. 

Edited November 11, 2021 by Scooot

[oymd]

10 hours ago, ONIKKINEN said:

I don't have pixinsight, but this sounds bizarre. Elliptical galaxies being round featureless mostly single coloured blobs are about as far away from M33 as possible when it comes to galaxies.

Maybe try just "normal" photometric colour calibration (if there is one in PI?). SCNR also shouldn't be necessary after a good photometric colour calibration so something is wrong.

 

9 hours ago, Scooot said:

“Here is a screenshot showing my Master Light having had Dynamic crop, DBE, PCC, EX Denoise, Arcsinstretch and histogram stretch.”

I always do Photometric Colour Calibration first before DBE, which is why  I suggested you try that in case DBE or any of the other processes had upset your colour balance. 

 

9 hours ago, Xiga said:

You might have more joy posting the stacked file to see if others have more luck. You also have to remember that, shooting Broadband in Bortle 8-9 is like playing AP on INSANE mode! Believe it or not, but 12 hrs of OSC is probably nowhere near enough under these conditions. You might need 30-40 hrs, a sobering thought if ever there was one. 

Saying that though, your screenshots do show quite a lot of colour i think. I use APP, and the order i usually do things in is: Background Neutralisation, then Gradient Reduction (DBE for you), followed by PCC (i sometimes use Siril instead of APP for this). Then remove the green, and finally do Arcsinh stretching to preserve the colour. 

HTH. 

 

5 hours ago, Scooot said:

It’d be interesting to know what your background flux is? There’s a simple script in Pixinsight that measures it. It’s called Calculate Sky Limited Exposure. 

Thank you all for your replies.

Here is my RAW data: 372 120s Lights (discarded quite a few), Darks, Fs & DFs.

https://www.dropbox.com/sh/8pazqynoadupch9/AACewrQUy3ZAZSG_JcdNtAvGa?dl=0

And here is my version of WBPP:

https://www.dropbox.com/sh/vs2tet3hfm98xol/AAApMMnk1Yj-XUZJ6FteKvPGa?dl=0

[MartinB]

This is a very interesting thread.  We have become accustomed to brightly coloured galaxies.  These images have been heavily colour processed typically emphasising the blue in the outer arms and the gold in the core.  The true colour, as we see it from earth is far more muted.  

Earlier this year FLO released  M81/82 data from it's IKI observatory for a processing competition  https://stargazerslounge.com/topic/373269-iko-m33-lrgbha-processing-competition/  Within this thread @vlaiv raised the issue of colour presentation.  Should we go for authentic or an aesthetically pleasing process.

I have done 2 very crude colour processes of you png, one with a relatively light touch and one which I would class as a little over saturated.  As you can see, you have colour!!

 

 

 

[Mr Spock]

The unsaturated images look far more natural to me. Gaudy galaxies aren't very appealing!

[Budgie1]

It doesn't help with the colour issue but; when you say you didn't use any filters, does that include a UV/IR cut filter?

The ASI294MC doesn't have a UV/IR cut sensor glass, so if I'm not using any other filters I always have a UV/IR cut in the image train. It stops any star bloating, if nothing else.  

[vlaiv]

Not having IR/UV cut filter will make colors washed out / unsaturated (even if colors are properly handled).

This is because QE in IR part of the spectrum tends to be similar for all three components of bayer matrix - R, G and B, so each pixels gets added same value in RGB triplet - like adding some amount of white (1,1,1) - that de-saturates colors.

Add on top of that non linear stretch that also tends to de saturates colors - and you'll end up feeling you don't have enough color in the image.

[oymd]

58 minutes ago, Budgie1 said:

It doesn't help with the colour issue but; when you say you didn't use any filters, does that include a UV/IR cut filter?

The ASI294MC doesn't have a UV/IR cut sensor glass, so if I'm not using any other filters I always have a UV/IR cut in the image train. It stops any star bloating, if nothing else.  

 

52 minutes ago, vlaiv said:

Not having IR/UV cut filter will make colors washed out / unsaturated (even if colors are properly handled).

This is because QE in IR part of the spectrum tends to be similar for all three components of bayer matrix - R, G and B, so each pixels gets added same value in RGB triplet - like adding some amount of white (1,1,1) - that de-saturates colors.

Add on top of that non linear stretch that also tends to de saturates colors - and you'll end up feeling you don't have enough color in the image.

Oh, I never thought of that. 
 

Yes, I think the 294MC Pro just has AR glass, but no IR/UV coating. 
 

I do not have any UV/IR filters. The filters I have are L-eNhance, IDAS D2 and V4

Would the D2 LP filter be a good option?

Alternatively, does Baader do a 2” UV/IR filter?

[oymd]

1 hour ago, Budgie1 said:

It doesn't help with the colour issue but; when you say you didn't use any filters, does that include a UV/IR cut filter?

The ASI294MC doesn't have a UV/IR cut sensor glass, so if I'm not using any other filters I always have a UV/IR cut in the image train. It stops any star bloating, if nothing else.  

 

57 minutes ago, vlaiv said:

Not having IR/UV cut filter will make colors washed out / unsaturated (even if colors are properly handled).

This is because QE in IR part of the spectrum tends to be similar for all three components of bayer matrix - R, G and B, so each pixels gets added same value in RGB triplet - like adding some amount of white (1,1,1) - that de-saturates colors.

Add on top of that non linear stretch that also tends to de saturates colors - and you'll end up feeling you don't have enough color in the image.

Is this what you had in mind?

https://www.firstlightoptics.com/baader-filters/baader-uvir-cut-filter.html

It says discontinued? No stock?

[Budgie1]

13 minutes ago, oymd said:

I do not have any UV/IR filters. The filters I have are L-eNhance, IDAS D2 and V4

Would the D2 LP filter be a good option?

Alternatively, does Baader do a 2” UV/IR filter?

If you use one of those then that will cover it.

12 minutes ago, oymd said:

 

Is this what you had in mind?

https://www.firstlightoptics.com/baader-filters/baader-uvir-cut-filter.html

It says discontinued? No stock?

I use a ZWO version: https://www.firstlightoptics.com/zwo-filters/zwo-1-25inch-iruv-cut-filter.html

[vlaiv]

13 minutes ago, oymd said:

 

Is this what you had in mind?

https://www.firstlightoptics.com/baader-filters/baader-uvir-cut-filter.html

It says discontinued? No stock?

Probably because they have new version:

https://www.firstlightoptics.com/uv-ir-filters/baader-cmos-optimised-uvir-cut-and-l-filter.html

But, why do you need 2" version?

1.25" is sufficient for 23mm sensor diagonal.

Also, no need to break the bank - maybe ZWO 1.25" will suffice:

https://www.firstlightoptics.com/uv-ir-filters/zwo-1-25inch-iruv-cut-filter.html

[iantaylor2uk]

I use a ZWO 2" UV/IR filter with my 071 MC Pro camera (you can get the filter here: https://www.firstlightoptics.com/uv-ir-filters/zwo-1-25inch-iruv-cut-filter.html)

I recently imaged M33, although for only 2 hours, using high gain (240), 2 min subs with the sensor cooled to -5 C, and the photo can be seen on my blog: 

https://chesterastroblog.wordpress.com/2021/11/08/first-images-from-takahashi-tsa-102/

This was using a 102 mm f/8 refractor (with a 1x flattener).

Some of the surrounding stars definitely had colour in them.

[Phillyo]

I always use the Optolong L-Pro filter when imaging RGB with my ASI533MC Pro, not tried M33 as it's a little small for my set up however a good technique I've found for bringing out colour is to extract a luminance layer from the image (after ColourCal, BackgroundNeutralisation, DBE and stretch). Then, recombine the Lum into the image using the LRGB Combination tool. Deselecting the RGB and put the Lum layer in the L, then drop the Saturation slider down to around 200 (adjust to taste) and apply it to the RGB image. You can tick the Noise Reduction too.

Then use the curves to add a bit more saturation to taste then use SCNR to reduce the green values.

Might be worth a try?

Phil.

[oymd]

6 hours ago, vlaiv said:

Probably because they have new version:

https://www.firstlightoptics.com/uv-ir-filters/baader-cmos-optimised-uvir-cut-and-l-filter.html

But, why do you need 2" version?

1.25" is sufficient for 23mm sensor diagonal.

Also, no need to break the bank - maybe ZWO 1.25" will suffice:

https://www.firstlightoptics.com/uv-ir-filters/zwo-1-25inch-iruv-cut-filter.html

I’ve decided to only buy 2” filters. 

I already have a 2” RGB set, 2” ha and o3. 
 

I decided to go 2” sometime ago, so as to not worry about any future mono camera purchase.