Leo Triplet with an ED80

[graemlourens]

Hi everybody!

Not much going on with this weather but was patient enough to wait for more color data (i always tend to rush on that RGB-data, and i shouldn't!) and was enjoying processing the leo triplet.

I didn't do anything fancy, pretty rudimentary processing (very little steps actually, LUM was very very strong) but i didn't feel it required much more magic.... maybe i'm wrong!

You'll find a few halos here and there, excuse that! Lets call them 'beauty spots' of a workhorse ED80...

Feedback warmly welcomed (also very critical!)

Kind regards, Graem

[ecuador]

Nice!

However the title of the thread should be "Leo triplet with a doublet".

[toxic]

very nice indeed Graem 

[PatrickGilliland]

Nice work - all the key details there despite the challenges 

[Ouroboros]

I think the detail and colours look quite good.  I had a go at this target a few weeks ago with the same telescope, reducer and my canon 450D. My image doesn't look nearly as nice as yours, which is why I haven't bothered posting it. 

  I'm interested to see what you got here with presumably your Atik 460EX.  I'm thinking of buying this camera, or more likely the Atik One 6.0. So good to see what I might expect.  You weren't tempted to buy the 490EX, which would give you the same size chip with slightly better resolution because of the smaller pixel size? 

[ollypenrice]

It's a good result. Did you try a hyper-stretch just to see what might be buried in the data? I always do this, just to see what might be there if I shot four times as much. I can see a few hints of the Hamburger tidal tail even in this well judged stretch.

Olly

[tony210]

Great image - great workhorse- Skywatcher ED80 living up to its reputation - Tony.

[graemlourens]

Thank you so much for all your replies!

12 hours ago, ecuador said:

Nice!

However the title of the thread should be "Leo triplet with a doublet".

Indeed

12 hours ago, toxic said:

very nice indeed Graem 

Thank you Chris!

12 hours ago, PatrickGilliland said:

Nice work - all the key details there despite the challenges 

Thank you Paddy. Always hurts me publishing LRGB pictures with these horrible stars, but i now just accept them and rather focus on the other elements of the image....

11 hours ago, Ouroboros said:

I think the detail and colours look quite good.  I had a go at this target a few weeks ago with the same telescope, reducer and my canon 450D. My image doesn't look nearly as nice as yours, which is why I haven't bothered posting it. 

  I'm interested to see what you got here with presumably your Atik 460EX.  I'm thinking of buying this camera, or more likely the Atik One 6.0. So good to see what I might expect.  You weren't tempted to buy the 490EX, which would give you the same size chip with slightly better resolution because of the smaller pixel size? 

Hi. I am very happy with the 460EX, and in my case i was even considering its smaller brother (with the same chip). The 490ex has a different chip type, chip and pixel size, so you'd have to research that (its been too long ago and i don't remember my research...). 

The fov i'm receiving with the ED80, Focal Reducer and the 460ex is pretty much perfect for so many targets. Combine it with the angle 77/257 degrees, and you can do so many targets without framing issues. There will be targets (for example M31, Soul & Heart nebula) that you will have to do mosaics, but based on my research to get a larger chip just for those targets, didn't match up with the extra cost, and i didn't want the smaller pixels than the 460ex has, as my seeing will always be the limit, not the resolution.

If you have any further questions, send me PM with your thread where you are discussing this, or we can continue via PM so we don't clog up the thread too much

11 hours ago, ollypenrice said:

It's a good result. Did you try a hyper-stretch just to see what might be buried in the data? I always do this, just to see what might be there if I shot four times as much. I can see a few hints of the Hamburger tidal tail even in this well judged stretch.

Olly

Hi Olly, thank you for your feedback! I indeed did (always do, in PI with autostretch its a normal process to start analyzing everything) but the quality rapidly deteriorated and the merits of going nuclear (even with masking) did not convince me to do so. I would have had to shoot more lum, but i'm already on 40 x 600s, so the benefits of adding more 10min-shots is marginal. I would have had to shoot 1000/1200s shots, but then i've got a whole different work to get my stars back, that are already now way too close to being overblown.

11 hours ago, tony210 said:

Great image - great workhorse- Skywatcher ED80 living up to its reputation - Tony.

I respect this scope highly.... and it doesn't get any special treatment! Just sits there in the observatory, waiting for me to remotely bring it to life

Kind regards, Graem

[ollypenrice]

18 minutes ago, graemlourens said:

Thank you so much for all your replies!

Indeed

Thank you Chris!

Thank you Paddy. Always hurts me publishing LRGB pictures with these horrible stars, but i now just accept them and rather focus on the other elements of the image....

Hi. I am very happy with the 460EX, and in my case i was even considering its smaller brother (with the same chip). The 490ex has a different chip type, chip and pixel size, so you'd have to research that (its been too long ago and i don't remember my research...). 

The fov i'm receiving with the ED80, Focal Reducer and the 460ex is pretty much perfect for so many targets. Combine it with the angle 77/257 degrees, and you can do so many targets without framing issues. There will be targets (for example M31, Soul & Heart nebula) that you will have to do mosaics, but based on my research to get a larger chip just for those targets, didn't match up with the extra cost, and i didn't want the smaller pixels than the 460ex has, as my seeing will always be the limit, not the resolution.

If you have any further questions, send me PM with your thread where you are discussing this, or we can continue via PM so we don't clog up the thread too much

Hi Olly, thank you for your feedback! I indeed did (always do, in PI with autostretch its a normal process to start analyzing everything) but the quality rapidly deteriorated and the merits of going nuclear (even with masking) did not convince me to do so. I would have had to shoot more lum, but i'm already on 40 x 600s, so the benefits of adding more 10min-shots is marginal. I would have had to shoot 1000/1200s shots, but then i've got a whole different work to get my stars back, that are already now way too close to being overblown.

I respect this scope highly.... and it doesn't get any special treatment! Just sits there in the observatory, waiting for me to remotely bring it to life

Kind regards, Graem

Why worry about the stars? If you went for longer subs and used Layers in Ps you could very easily combine a hyperstretched long-sub version with a softer one with better stars. In my own version my stars are from the RGB-only layer. The galaxies, tidal distortions and faint fuzzies come from a very deep set of long L subs added to RGB. 

Olly (Layer-o-holic.)

 

[graemlourens]

9 hours ago, ollypenrice said:

Why worry about the stars? If you went for longer subs and used Layers in Ps you could very easily combine a hyperstretched long-sub version with a softer one with better stars. In my own version my stars are from the RGB-only layer. The galaxies, tidal distortions and faint fuzzies come from a very deep set of long L subs added to RGB. 

Olly (Layer-o-holic.)

 

I somehow have never mastered replacing stars, but in fact i have never been trying to just enhance the pure RGB image, as my RGB ist mostly binned data (also in this case, the RGB data is 2x2).
Am i correct in the assumption that your way of doing it, would only work if the RGB data is also 1x1? I just tried taking my very weakly stretched RGB and with layers just added the galaxies back in, it works perfectly and its pretty simple, BUT my stars are obviously pretty wonkers, as lots of them were 4 pixel stars to begin with (at a resolution of over 3.5" / px)

I do like the idea though... so i would then rather shoot fewer but longer 1x1 RGB's, and even longer L's (and not using the stars from L at all).

(p.s. i love layers too! nowadays i heavily use PS very early on in the process, PI is for managing the project, as well as alignment & calibration)

Kind regards, Graem

[Barry-Wilson]

Lovely work Graem.  This trio is ever appealing to me - there seems to be a harmony between the shapes and their relative positions to my eye and you have captured that sense of 'harmony' for me.  Well done.

[ollypenrice]

2 hours ago, graemlourens said:

I somehow have never mastered replacing stars, but in fact i have never been trying to just enhance the pure RGB image, as my RGB ist mostly binned data (also in this case, the RGB data is 2x2).
Am i correct in the assumption that your way of doing it, would only work if the RGB data is also 1x1? I just tried taking my very weakly stretched RGB and with layers just added the galaxies back in, it works perfectly and its pretty simple, BUT my stars are obviously pretty wonkers, as lots of them were 4 pixel stars to begin with (at a resolution of over 3.5" / px)

I do like the idea though... so i would then rather shoot fewer but longer 1x1 RGB's, and even longer L's (and not using the stars from L at all).

(p.s. i love layers too! nowadays i heavily use PS very early on in the process, PI is for managing the project, as well as alignment & calibration)

Kind regards, Graem

Good points. One of the reasons I don't bin colour is that I do want to use my RGB-only stars in the final image so they are best unbinned, of course. However, you could easily do a 'short L-sub/RGB' as a starfield layer and blend it with a 'Long L-sub/RGB for the 'faint stuff' layer.

You have a great Triplet but (next year, maybe) you could easily take it deeper. Just thinking aloud! It is a hell of a target, as Barry says.

Olly

[graemlourens]

On 5/13/2017 at 20:14, Barry-Wilson said:

Lovely work Graem.  This trio is ever appealing to me - there seems to be a harmony between the shapes and their relative positions to my eye and you have captured that sense of 'harmony' for me.  Well done.

Thank you Barry for your kind words!

On 5/13/2017 at 20:52, ollypenrice said:

Good points. One of the reasons I don't bin colour is that I do want to use my RGB-only stars in the final image so they are best unbinned, of course. However, you could easily do a 'short L-sub/RGB' as a starfield layer and blend it with a 'Long L-sub/RGB for the 'faint stuff' layer.

You have a great Triplet but (next year, maybe) you could easily take it deeper. Just thinking aloud! It is a hell of a target, as Barry says.

Olly

I get it, and i have never thought of doing it this way round. This though will add required imaging time substantially if i'm calculating correctly. Probably double the time spent on RGB-data?

This approach (using RGB as base layer with stars & background) will probably only work best having single objects (like galaxies) so you're only enhancing the existing extracted RGB luminance with your deep LUM in that specific area. 
I guess it will be very target specific, if those 'shortish' RGB subs (for stars & background) will have enough color information for the target its self. If not one would have to shoot (as you said) another set of long RGB's that will only be used for colouring the target.

You have inspired me to do this approach on a very difficult and rarely (with my kind of equipment) imaged target NGC403, even if my dada there is binned. I'll post a separate thread with the findings & the image.

Kind regards, Graem

[swag72]

My first thought on this was that it appeared a little green..... so I downloaded the image shoved it through PI SCNR and the galaxy colours came out much better in my opinion..... Have a go, bannish the green and see what you think  

[ollypenrice]

34 minutes ago, graemlourens said:

Thank you Barry for your kind words!

I get it, and i have never thought of doing it this way round. This though will add required imaging time substantially if i'm calculating correctly. Probably double the time spent on RGB-data?

This approach (using RGB as base layer with stars & background) will probably only work best having single objects (like galaxies) so you're only enhancing the existing extracted RGB luminance with your deep LUM in that specific area. 
I guess it will be very target specific, if those 'shortish' RGB subs (for stars & background) will have enough color information for the target its self. If not one would have to shoot (as you said) another set of long RGB's that will only be used for colouring the target.

You have inspired me to do this approach on a very difficult and rarely (with my kind of equipment) imaged target NGC403, even if my dada there is binned. I'll post a separate thread with the findings & the image.

Kind regards, Graem

The 'Full Monty' version of this approach is to make two images. The first is a very hard-stretched LRGB pulling out the faint fuzzies but giving big stars as well. I get the background to a flat 23/23/23 in RGB. The second image is RGB only, processed just for good stars. Again, the background must be a flat 23/23/23, strictly identical to the first or it won't work. Counter intuitively, perhaps, I put the starfield RGB on top of the hard stretch as a Ps layer and set the opacity of the top layer to zero. Armed with a soft edged eraser of varying size I then carefully apply it over every galaxy, tidal tail or faint fuzzy that I can see. Nothing changes because the top layer is invisible, but when restored to full opacity you should have the best of both layers. I check carefully to ensure that all faint fuzzies are visible.

You're quite right that this is target specific. It's no good where you have stars embedded in nebulosity.

Olly

 

[graemlourens]

3 minutes ago, ollypenrice said:

The 'Full Monty' version of this approach is to make two images. The first is a very hard-stretched LRGB pulling out the faint fuzzies but giving big stars as well. I get the background to a flat 23/23/23 in RGB. The second image is RGB only, processed just for good stars. Again, the background must be a flat 23/23/23, strictly identical to the first or it won't work. Counter intuitively, perhaps, I put the starfield RGB on top of the hard stretch as a Ps layer and set the opacity of the top layer to zero. Armed with a soft edged eraser of varying size I then carefully apply it over every galaxy, tidal tail or faint fuzzy that I can see. Nothing changes because the top layer is invisible, but when restored to full opacity you should have the best of both layers. I check carefully to ensure that all faint fuzzies are visible.

You're quite right that this is target specific. It's no good where you have stars embedded in nebulosity.

Olly

 

Hi Olly. Thank you. It makes sense.

Just to clarify, why are you 'erasing' instead of switching the layers, putting a black layer mask on the top (hard stretched) one and 'whitening' the galaxies in the mask? (or is this probably boiling down to the same?)

Kind regards, Graem

[ollypenrice]

11 minutes ago, graemlourens said:

Hi Olly. Thank you. It makes sense.

Just to clarify, why are you 'erasing' instead of switching the layers, putting a black layer mask on the top (hard stretched) one and 'whitening' the galaxies in the mask? (or is this probably boiling down to the same?)

Kind regards, Graem

I'm notorious for not bothering with masks. I just use the feathered eraser to take off the top layer. Sorry to be such a yob!!

Olly

[graemlourens]

16 hours ago, swag72 said:

My first thought on this was that it appeared a little green..... so I downloaded the image shoved it through PI SCNR and the galaxy colours came out much better in my opinion..... Have a go, bannish the green and see what you think  

Sara, sorry i missed your comment yesterday! Will give it a try and see. I'm surprised if i had not already done that, but maybe i did skip this for some reason by accident!

Kind regards, Graem

[graemlourens]

On 5/15/2017 at 13:54, swag72 said:

My first thought on this was that it appeared a little green..... so I downloaded the image shoved it through PI SCNR and the galaxy colours came out much better in my opinion..... Have a go, bannish the green and see what you think  

Hi Sara.

I quickly checked this and indeed i had done a pass of SCNR (setting to 0.6). I re-did it on the final, and even with a full 1.0, the differences are for my (colorblind) eye nearly not noticeable, only if i 'pixel piep' like you usually say  i'm surprised you've got that initial feeling of the greenish touch! Maybe you have super human eyes

Any ways, i banished green forever now, but i'll spare you of posting the image, as its not really a visible change in reduced jpegs.

Kind regards, Graem

 

[swag72]

I can happily post my image if you like - I found it a considerable difference...

 

[graemlourens]

1 minute ago, swag72 said:

I can happily post my image if you like - I found it a considerable difference...

 

I checked on a different screen, and i see what you mean, the cores both, do profit from full scnr vs only a partial removing (my work screen does not show this as well sadly)

My workflow always contained SCNR, but never at full 1.0, but if i compare original to 0.6, or original to 1, my eyes can honestly not spot the difference. So i guess my color blindness includes green as well now!!

Kind regards, Graem

 

[swag72]

That's a really noticeable difference to me Graem........

[DaveS]

Agree with Sara

Even on the bog-standard monitor I'm using ATM the difference is very clear.

[graemlourens]

38 minutes ago, swag72 said:

That's a really noticeable difference to me Graem........

 

24 minutes ago, DaveS said:

Agree with Sara

Even on the bog-standard monitor I'm using ATM the difference is very clear.

I give up! as i said, i had already corrected it, as i trust your judgement and i know my eyes are not the best in this area....

Kind regards, Graem

[graemlourens]

And here it is no green!