Time for a new Project.....on noes.......

[pyrasanth]

Well after after my little bit of success with M1 I've picked a new target...to bore you with.....This will be M82...only because I lack ambition but it's an all year round target at my latitude so I've got plenty of room to make a mess of the image......

I've started with just a couple of hours of subs in the image below. This is LRGB & HA data. The weather was not good for these subs. The stars are not round & the data is noisy & the stars lack saturation but it shows the potential. I just need a lot more data. The ratio of sub time is 10 minutes per RGB colored filter; 20 minutes L (2x 10 separate subs) & 20 minutes HA- so its 70 minutes per block of data. I reckon 20-25 blocks of data for this project would be good.

I'm putting into action all I learnt from you guys so you should see an improvement in the image in each step. Posted below is the result of last night representing 2 blocks of data but the red R has dropped a frame-deleted because a cloud came in.

As always let me know what you think I need to do as the images land & we can all learn together!

You may have noticed a parallel line on each of the bright stars. This is the SMFS control cable. This month I will order the split SMFS cable from Optec in the USA and this assembly will give the familiar four point star pattern which will look better as it splits the control cable into 4 twisted pair strands placed on the corrector at right angles.

Clear skies to you all.

 

 

[pyrasanth]

I took the same data & applied the lessons from my previous attempts. By protecting the background stars with a mask I was able to process the background noise out without changing any of the stars. This will be very useful for the future as it always intense image processing on the background without affecting the stars.

See image below with this in action- no change in the stars but the noise in the galaxy is improved- ignore the background as there is not enough data to make it clean yet- I await the weather to get better!

 

[laser_jock99]

Good detail in the galaxy.

[laudropb]

Very nice image. Lovely detail captured.

[MikeODay]

Lovely image - I really like the colours and the detail in the galaxy.

[pyrasanth]

Hi all- thanks for the encouragement but its really early days on this one- I only have 2 hours of data just 2 sets of Lx2 RGB x1 & HAx2- So when I have about another 10 sets it should be better!

I'm capturing sets of data in blocks of 10 minute subs so each set is:

L (2x 10 mins)

RGB (1x 10 mins each filter)

HA (1 x 20 mins)

Atik 460 Celestron C11 Edge at F7 using the ONAG & focuslock

 

[pyrasanth]

I continue to work with M82 & captured another 6 hours of data. The improved image is below. I still need loads more HA data as the HA layer is still dirty with noise- I may need to increase the exposure from 20 minutes to perhaps 40 and see if that helps.

 

 

[pyrasanth]

[iksose7]

As a general rule i wouldnt apply much if any noise reduction to the luminance in images, i.e the galaxy here. Thats where all your detail is and NR will do it no favors! Look forward to seeing how this image progresses like i did with your M1 image.

Callum

[pyrasanth]

39 minutes ago, iksose7 said:

As a general rule i wouldnt apply much if any noise reduction to the luminance in images, i.e the galaxy here. Thats where all your detail is and NR will do it no favors! Look forward to seeing how this image progresses like i did with your M1 image.

Callum

Thanks for the advice. I don't have enough data captured yet so until I have all of the data I'm into more rough processing mode to get as real as to how things are progressing. Certainly the stars need to be processed in there own layer as the HA integration kills the colour & gives the stars a blue cast as in the first image. The L data is fairly good so I hope I can build on that but it's going to take many hours more data as the problem is the HA which is going to need lots of long subs. The above image is over 11.5 hours & still needs loads more data! so I will just need to keep going. I'm going to focus a bit more on the HA for the next session so I can get a full 6 hours of data to add to the pot.

[pyrasanth]

I did some digging around on my hard drive & I found 10 hours of 2x2 binned HA data- this was like mana from heaven- I put this into the image & it is now getting better. Let me know what you think- I just need to improve the stars and as they are on a separate layer I can do a star catch session to finish off at the end and lighten the background.

 

 

[gorann]

Great looking image!

but, in what channel did you put the Ha data? It looks like you lost the red in the outflow, which you had in your previous versions and in most ones I have seen, like this one:

http://apod.nasa.gov/apod/ap130704.html

I would blend the Ha in with the red channel.

[pyrasanth]

1 minute ago, gorann said:

Great looking image!

but, in what channel did you put the Ha data? It looks like you lost the red in the outflow, which you had in your previous versions and in most ones I have seen, like this one:

http://apod.nasa.gov/apod/ap130704.html

I would blend the Ha in with the red channel.

Yes- there is an issue with the HA channel- I will look at this. I think the image messes up in the photoshop actions if you try & blend LRGB with HA instead of RGB with HA. Let me have a look & I will post a new image when I've examined the workflow again.

[gorann]

Just mix it into the red channel and see what happens (in PS delete all channels except the red one from a copy of your image, make that a gray scale image, then blend the Ha into it, and then paste it into the red channel of your LRGB image)

[pyrasanth]

I've retested & the blend fails if you have an L layer in the RGB image. The new image is below. Let me know what you think.

 

 

[pyrasanth]

Since the stars have been derived from the original subs they are not as round as I would have liked. This is because of the 20 minute subs & poor weather conditions so the guiding has drifted a little. What I will do is take a lot of short subs then produce a new star field at the correct plate scale and integrate this into the view. This is really easy to do as I now try & process all my images with of the main target with no stars.

[pyrasanth]

The image below was colour calibrated in PixInsight with a deconvolution & noise reduction. This has toned down the dramatic colours & enhanced the fine details in the nebula. Final task as I said earlier is to recreate the stars from a few fresh subs & get a load more HA data to expand the jets.

[iksose7]

You have made excellent progress from your first version! Those Ha jets are looking great. One little thing to pick at is a slight green hue around the edge of the galaxy. A quick hit from PIs SCNR should easily take care of that problem

Callum

[gorann]

Looks really good. Amazing what the deconvolution did to the detail. Good luck on the Ha hunting!

[gorann]

9 hours ago, pyrasanth said:

I've retested & the blend fails if you have an L layer in the RGB image. The new image is below. Let me know what you think.

 

 

That is probably because the L-layer contains no information about the new Ha data so it is not illuminated. You would somehow have to get that information into the L-layer (blend Ha in there too). In any case, it looks good to me.

[pyrasanth]

1 hour ago, gorann said:

Looks really good. Amazing what the deconvolution did to the detail. Good luck on the Ha hunting!

Thanks for the encouragement from everybody. I discovered something I did not know in PixInsight which will prove really useful. If I shoot the HA at a 2x2 bin (and most of the old data incorporated into this image is 2x2 with no noticeable image degradation) it will automatically scale to the same resolution, 1x1, as the other 1x1 subs  when registered. This I did not know & will save loads of time as I thought I would have to re scale the 2x2 binned data. I will increase the 2x2 HA exposure to 30 minutes (I had been using 20 mins at 1x1 recently) which should substantially increase the sensitivity & speed of data capture. I've certainly proved that with my seeing conditions I've been substantially over sampling. This is why I'm anxious to get the Atik 11000 pressed into service as the 1x1 or 2x2 bin on the 9 um pixels is a much better match at my focal length either 0.94 or 1.88 arc secs per pixel. I should be able to order the new full chip ONAG in the next few weeks- it's just very expensive as with the purchase of the new guide camera (which I have) your looking at a combined investment of nearly £2500.00.

I did a quick test on the image with SCNR and the green quickly goes. I will incorporate that in the final image when I redo the stars as they are bound to have some green around which will be undesired.

As for the Starlight Express Ultrastar camera its a great match for the ONAG at 2x2 bin with PHD2. Great sensitivity and a reasonably wide field which is all that is required.