Messier 33 **PROCESSING CHALLENGE**

[johnrt]

After the success of Olly's thread where he opened up his NGC 1333 data for one and all, I've decided to let some of my data out in to the wild, to see what the good folks of SGL make of it.

Perhaps this could be a regular monthly thing where an imager releases data for a monthly processing competition?

So anyway, on to my data, there is a set of fully calibrated and aligned masters, Luminance, Red, Green, Blue and Ha. The data in total adds up to about 22 hours of imaging with my 6"RC and Atik 460ex. Everything has been aligned and prepared in PI and saved as a .tiff so you can choose your processing weapons of choice.

Here is my latest version of processing the data, I'm very interested to see what you good folks can come up with as M33 seems to be one of those objects with no actual "correct" way to present the data.

 

 

Links:

https://www.dropbox.com/sh/a010qbtlwtxcp9q/AABJG67fy960Ahn78n75mMRLa?dl=0&lst=&preview=M33_Lum.tif

https://www.dropbox.com/sh/a010qbtlwtxcp9q/AABJG67fy960Ahn78n75mMRLa?dl=0&lst=&preview=M33_Red.tif

https://www.dropbox.com/sh/a010qbtlwtxcp9q/AABJG67fy960Ahn78n75mMRLa?dl=0&lst=&preview=M33_Green.tif

https://www.dropbox.com/sh/a010qbtlwtxcp9q/AABJG67fy960Ahn78n75mMRLa?dl=0&lst=&preview=M33_Blue.tif

https://www.dropbox.com/sh/a010qbtlwtxcp9q/AABJG67fy960Ahn78n75mMRLa?dl=0&lst=&preview=M33_Ha.tif

 

When you click on the link, you should see a download button in the top right, and then select direct download.

 

 

 

[wimvb]

That's a great initiative. The risk is of course, that this forum gets flooded with "challenge me!" posts. But a once a month challenge would be great. Do we need someone to moderate this?

I'll have a go at your image. (But at the moment I've put it in queue. )

[alan4908]

Since I really need the practice and I'm currently acquiring my own M33 data I thought I'd have a go. I used CCDstack for the deconvolution, Pixinsight for the DBE and colour calibration and PS for the rest. The image is a LRGB with a H red blend.  

Alan

[johnrt]

7 hours ago, wimvb said:

That's a great initiative. The risk is of course, that this forum gets flooded with "challenge me!" posts. But a once a month challenge would be great. Do we need someone to moderate this?

I'll have a go at your image. (But at the moment I've put it in queue. )

I think the majority of imagers are quite precious about sharing their data, so it might be a struggle to g et volunteers rather than get flooded? We'll see how generous everyone is

 

2 hours ago, alan4908 said:

Since I really need the practice and I'm currently acquiring my own M33 data I thought I'd have a go. I used CCDstack for the deconvolution, Pixinsight for the DBE and colour calibration and PS for the rest. The image is a LRGB with a H red blend.  

Alan

Very interesting Alan, you have a very different colour scheme to me, and I can see the benefit of deconvolution the the inner dust lanes of the galaxy. My version had no DBE applied.

[Nova2000]

Hi that's a great image of the triangulum galaxy. I'm a new guy here. Are these images free to download ? If it is I would like to try them and learn some processing in photoshop. And are those bright red areas nebulosity? 

[alan4908]

6 hours ago, johnrt said:

Very interesting Alan, you have a very different colour scheme to me, and I can see the benefit of deconvolution the the inner dust lanes of the galaxy. My version had no DBE applied.

 Hi John

Yes, I agree - it is interesting how different the results can be from the same starting point. On the colour calibration point, FYI I used Pixinsight CC with the option of the white reference point being the entire galaxy, this was applied after DBE.

Alan 

[Knight of Clear Skies]

3 hours ago, Nova2000 said:

And are those bright red areas nebulosity? 

Yes, they are emission nebulae. The big one at top left is NGC 604, which is something like 40 times as large as the Orion Nebula and 6,000 times as bright!

[Davey-T]

Good idea John, not that I'll have time to join in but it is giving me a kick up the rear to have a look at my subs of M33, not intended as a project as I only took them as it was handily placed when setting up the new mount.

Dave

[Mike Hawtin]

Well John here is my attempt, I've blended the Ha into the red and luminance channels, if it was my image I would spend more time on the Ha/red/lum proportions to remove the bit of background Ha continuum remaining. 

Still, very nice data and it was interesting to have a play with another imager's data.

Thanks.

 

Mike

[johnrt]

13 minutes ago, Mike Hawtin said:

Well John here is my attempt, I've blended the Ha into the red and luminance channels, if it was my image I would spend more time on the Ha/red/lum proportions to remove the bit of background Ha continuum remaining. 

Still, very nice data and it was interesting to have a play with another imager's data.

Thanks.

 

Mike

Great stuff Mike, I think you have managed to pull a little more out of the Ha than I did in my version.

Any more takers? Get your processing hats on!

[johnfosteruk]

I hope you don't mind me having a go John. Fingers crossed I'll be in possession of the means to do DSO soon, so I thought I'd dip my toe into processing with your data. 

Photoshop to stretch and blend. Red channel & HA blended and luminosity on top. some saturation and colour balance adjustments, guassian blur and sharpening and here's the result.

Only a quickie so I hope it does your data at least a little bit of justice.

[gnomus]

Sorry for taking so long to respond, John.  I was at the IAS and then ended up dealing with the aftermath of that 

Here is my attempt:

[johnrt]

Thank you John & Gnomus for taking part, again 2 different takes on the same data! I'm really enjoying seeing these differing outcomes, I have some parental visiting at the moment so haven't had much time to respond, but keep 'em coming!

[Starlight 1]

As someone do not had any paint programmes this challenge will be very difficult this is what I come up with. Use nebulosity 4  Used RGB L  no ha . Applied  digital stretch, one auto contrast, auto balance, greyscale for the noise. Two minutes of process, but find it interesting to see how it deals with data. Was not planning on destroying your lovely data but  very interesting outcome.

PS the images is on another computer so I edit post in a minute or 2.

[Space Cowboy]

Thanks John for sharing your data! As a deep sky rookie having access to quality like this is invaluable to see how my processing is coming along. I've only done an RGB as I can't yet figure how to add the Ha & L channels.

This was done with curves in CS2 and channels combined in image analyser :

 

[wimvb]

And here's my attempt, PixInsighted.

As I'v never done NB and RGB, I followed Kayron Merciecas (lightvortexastronomy.com) tutorial on NB RGB combination.

TGV noise reduction in the linear stage.

Then added saturation steps. HDRMultiscaleTransform gave a too harsh Luminance image, so I used LocalHistogramEqualization to boost details in the dust.

Some more colour saturation steps using a luminance mask.

[Joe12345]

Mine:

[gorann]

John,

this was a great idea for all of us suffering from a cloudy sky. Here is my effort in progress so far. I have not done much deconvolution and sharpening so far, but I may give that a try if the clouds hang in. After posting it I realize I have not even cropped away the bad edges, but...

[johnrt]

More interesting versions, no 2 are still alike, and all so different from my original! This is why astro imaging is not a "documentary" process!

Joe, you appear to have some artefacts in your processing, perhaps overdone it with the deconvolution or something?

[gorann]

I wonder a bit about the yellow structures in the centre that some of us brought out and some not, like me. Did I lose them somewhere during processing? I only used Photoshop CS5 by the way.

[johnfosteruk]

1 minute ago, gorann said:

I wonder a bit about the yellow structures in the centre that some of us brought out and some not, like me. Did I lose them somewhere during processing? I only used Photoshop CS5 by the way.

I think you'll be able to pull that out with a little bit of selective colour under adjustments.

[gorann]

19 minutes ago, johnfosteruk said:

I think you'll be able to pull that out with a little bit of selective colour under adjustments.

That also came to my mind so I have tried selective colours and I can produce a yellowish tint on it all (rather ugly) but not specifically on those central dark dustlines.

[Scott]

@johnrt,

Great set of data and kind of you to give others the opportunity to work with quality data.

[gorann]

I question struck me while processing this image: can we really see stars in another galaxy with a 6" telescope (or rather take pictures of them)? Or what else are those bright spots in the arms of M33? If they are stars I assume they must be enormous.

[johnrt]

3 hours ago, gorann said:

I question struck me while processing this image: can we really see stars in another galaxy with a 6" telescope (or rather take pictures of them)? Or what else are those bright spots in the arms of M33? If they are stars I assume they must be enormous.

I'm unsure if these you have pointed out are clusters of new born stars or perhaps orbiting globular clusters, but they are clearly within M33 not our own Milky Way by the way they follow the curve of the spiral arm.

As to the question "Can you image a single star in another galaxy with just a 6" OTA?", the answer is a definite yes, there is a confirmed a luminous hypergiant visible in my image, the star is B324 and is just to the northern edge of IC 142, here is a crop of my image with the annotation added for the star B324.

 

 

M33 isn't the only galaxy in which you can image a single star either! Here is my image of Hubble's Cepheid Variable in M31 taken with the same 6" RC.

 

Mind boggling isn't it!!!