Another Flaming Star Etc, DSLR + UHC filter

[Tommohawk]

Having seen StargeezerTim’s very nice Flaming Star nebula with DSLR and UHC filter, I thought I’d have a go too. Turned out to be a bit trickier than I thought.

First off, Tim has a 200mm lens where mine’s 300mm. So the framing had to be fiddled with - in fact the framing turned out to be a big headache cos I couldn’t see anything on Liveview with the UHC filter in place.

Tim – if you see this I’d love t know how you dealt with this! Maybe I could have framed it without the filter, then moved to a star bright enough to focus on with the filter, and then go back to the nebula and hope the pointing accuracy was about right?

Anyhow my first efforts were poorly framed and didn’t include the spider, fly and tadpoles but the data looked good – nicely focused. The second go had the framing sorted, but didn’t look so well focused. To complicate things, I noticed the UHC filter loses some of the blue so the reflection elements were poorly represented. I hoped to fix this by doing some shorter exposures without the filter. Unfortunately, by the time I was done with the filtered images, I’d gone past the meridian and when I flipped for the blue data, by the time I’d refocused, the framing wasn’t quite the same. The kit was icing up and so was I so I called it a day. I also had yet to do darks cos for this I used ISO 1600, where I normally use 800.

Whilst I was bemoaning the poor focus of the main image set, I noticed that the red channel seemed really crisp  – I think the filter must focus the RGB  colours differently. So I created a false luminance layer from the red channel which seemed to make a huge difference.  Not sure what others think of this idea? It probably further compounded the lack of blue data – maybe I’ll redo that at some point. As an experiment I added in the blue from the unfilltered images and it definitely helped - just need to get the whole frame done!

So, Canon 550D with Tamron 300, 17 exposures of 6 minutes ISO 1600, F4. Darks, flats,  and offset. Deepskystacker, and PS.

Hope you like it. As always, please feel free to comment or criticise.

[Adam J]

Hi Tom, very nice.

I have found that when imaging with a UHC you often get a nicer color balance by re-assigning the green pixels to blue and then processing it as a bi-color image. That way you dont end up with green stars and you get a more natural color.

[Tommohawk]

1 hour ago, Adam J said:

Hi Tom, very nice.

I have found that when imaging with a UHC you often get a nicer color balance by re-assigning the green pixels to blue and then processing it as a bi-color image. That way you dont end up with green stars and you get a more natural color.

Thanks for the feedback. Well that opens a world of possibilities I'd never even considered!... I sort of thought that Bicolor images were the domain of CCD imagers. I'm away for a few days, but I'll definitely give that a go when I have a mo.

Daft question alert.. when doing Bicolor (say R and G) do you just leave the B channel dark?

[Adam J]

14 hours ago, Tommohawk said:

Thanks for the feedback. Well that opens a world of possibilities I'd never even considered!... I sort of thought that Bicolor images were the domain of CCD imagers. I'm away for a few days, but I'll definitely give that a go when I have a mo.

Daft question alert.. when doing Bicolor (say R and G) do you just leave the B channel dark?

Depends, look at it this way.

In a DSLR:

Red (1 x pixel) = H-a + SII

Green (2 x pixels)  = OIII

Blue (1 x pixel) = H-beta + OIII

Blue is also the most important pixel for reflection nebula.

So in terms of the blue....it contains a strong OIII signal, not quite as strong as the OIII signal found in the green channel but close. If I was using a narrow band OIII filter I would be using both the green and blue channels for OIII but a 8.5nm narrowband filter cuts the h-beta out from the blue channel leaving only OIII.  So the problem with a UHC is that blue also contains hydrogen information in the form of H-beta that will mix with the H-a contained in the red channel to make hydrogen areas appear slightly purple and meaning that you don't get a pure separation between the Oxygen and hydrogen channels so you wont quite get that full bi-color effect.

So I tended to remove the blue channel completely for this reason, you are still left with two pixels worth of OIII. The green channel would eventually become the new blue channel. I would then take the red and the newly created blue channel and process them separately before re-combining them and making an artificial green channel. Finally depending on the target I might perform a Gaussian blur on the resultant color image and then re-use the red channel as luminance. 

The only problem with this is that you will lose any reflection nebula components within the image...although those will be weak within a UHC image anyway unless very bright (depends on how narrow your UHC is). So at that point I would examine the original blue channel for hints of the reflection nebula component and then if present use a mask and blend mode lighten to integrate them back into the final "Bi-color image". Alternatively you can also add some RGB exposures using a wider filter like a CLS and blend those with the "Bi-color image".  

The other way is just to add the Green and Blue pixels together and then assign both to the Blue channel. You will gain a little OIII signal but depending on the target the end effect may not be so striking as you are failing to fully separate the hydrogen signal from the Blue channel. Probably easier to process though.

Its worth noting that all the above works much much better if you have a modified camera, which I see you do. 

Best thing is that you can just reprocess this image using this technique so nothing has been lost. I would be interested to see the result if you chose to do so. 

[Tommohawk]

Adam -  thanks for taking the time to give such a detailed response.

Although I had considered the impact of the UHC filter on the blue data, I hadnt thought about how Ha, Hb, and OIII are poorly represented via the RGB matrix

I can defintely see that the unfiltered (misaligned) subs contain some relevant blue data so it would be good to use this somehow. I take your point about the possibility of combining G and B data, though TBH I have no idea how to do this.

It strikes me that in the long term it would be better to just go to mono and then filter LRGB or NB as required.

I'll have a go with your suggestions, and post the results, though unfortunately wont be able to do this for a few days as am away ATM.

[Adam J]

No problem, you dont actually combine the channels that was misleading. You actually remove the red channel from the raw file using a program like iris to leave the green and blue channels. You do that to all the images prior to stacking. Then you do the opposite and remove the green and blue from the red then stack. That will leave you with two images, one containing the green and blue (or just the green) and the other with just the red.

[StargeezerTim]

On 10/01/2017 at 18:19, Tommohawk said:

Having seen StargeezerTim’s very nice Flaming Star nebula with DSLR and UHC filter, I thought I’d have a go too. Turned out to be a bit trickier than I thought.

First off, Tim has a 200mm lens where mine’s 300mm. So the framing had to be fiddled with - in fact the framing turned out to be a big headache cos I couldn’t see anything on Liveview with the UHC filter in place.

Tim – if you see this I’d love t know how you dealt with this! Maybe I could have framed it without the filter, then moved to a star bright enough to focus on with the filter, and then go back to the nebula and hope the pointing accuracy was about right?

 

Hi Tom, Just seen this,

Nice image and composition, What I usually do is focus on the brightest star available, then go to the target, take a short one with high iso... then adjust by going to a bright star at 10X in live view and move it around with the synscan cursors. Take another and move again etc if needed. As long as you can see a star in liveview it should be poss to frame OK. If not, then I would frame without filter, save as a user object no 1 or whatever, go to a bright star, focus then return to the same user object. Hope this helps, Tim. 

[Tommohawk]

2 hours ago, StargeezerTim said:

Hi Tom, Just seen this,

Nice image and composition, What I usually do is focus on the brightest star available, then go to the target, take a short one with high iso... then adjust by going to a bright star at 10X in live view and move it around with the synscan cursors. Take another and move again etc if needed. As long as you can see a star in liveview it should be poss to frame OK. If not, then I would frame without filter, save as a user object no 1 or whatever, go to a bright star, focus then return to the same user object. Hope this helps, Tim. 

Hi Tim, thanks for chipping in. I couldnt see any stars on liveview, so just had to move it take a sub at ISO 6400 and see what happened. Even with high ISO it needed a 20sec sub with the filter in. Because the camera is rotated to an odd angle its difficult to know which way to nudge it.

With Liveview I dont think I tried the magnified screen - maybe that might help to make the star brighter, albeit more difficult to locate. I'll try that next time. Or maybe framing without the filter, save position, then insert filter and refocus and then revert to saved position. In practice if I'm going to do unfiltered images for the blue data I might as well do that first anyhow.

I'm thinking so many of the practical and processing issues would be easier with a dedicated CCD/CMOS camera that I'll go that way when funds allow! the higher QE is bound to help with this.

[StargeezerTim]

1 hour ago, Tommohawk said:

 

I'm thinking so many of the practical and processing issues would be easier with a dedicated CCD/CMOS camera that I'll go that way when funds allow! the higher QE is bound to help with this.

For me... i want to avoid using a computer in imaging so can't see me going down a ccd route. I don't think a uhc filter should cause too many problems with colour. I use the auto colour balance feature in dss and don't  get particularly get green stars. I decided to invest in PS and try and improve my processing skills as far as i can... a neverending journey!

If i was a bit younger i might try ccd but my addled brain barely copes as it is. 

P.s. In ps, a touch of hlvg on a star layer would help with the stars, as would 'selective colour".

[Tommohawk]

On 12/01/2017 at 18:27, StargeezerTim said:

For me... i want to avoid using a computer in imaging so can't see me going down a ccd route. I don't think a uhc filter should cause too many problems with colour. I use the auto colour balance feature in dss and don't  get particularly get green stars. I decided to invest in PS and try and improve my processing skills as far as i can... a neverending journey!

If i was a bit younger i might try ccd but my addled brain barely copes as it is. 

P.s. In ps, a touch of hlvg on a star layer would help with the stars, as would 'selective colour".

I quite understand wanting to avoid using a computer when imaging - apart from anything else it keeps it all more portable. But how are you guiding?

I do use autocolour mostly because my colour vision is poor and it seem to help get the balance right. But sometimes it seems to affect the overall bright/contrast/curves so I cant always use it.

Also I'm not sure what HIVG is?

I chance on the selective colour option, but not sure how best to use it - what do you find it useful for?

[PeterCPC]

HLVG - Hasta La Vista Green. Photoshop/Paintshop free plug in.

Peter

[Tommohawk]

Hi Peter - thanks - never heard of it, but sounds useful especially for me... I cant see any green in it at all!!

Usually I resort to the low tech solution of calling my wife in to have a look - but I posted this whilst away so it didnt get the usual colour check. Now I understand the comments suggesting use of selective colour.

I'll down load and sort the color and re-submit the end result. Many thanks!

[StargeezerTim]

2 hours ago, Tommohawk said:

I quite understand wanting to avoid using a computer when imaging - apart from anything else it keeps it all more portable. But how are you guiding?

I use a Lacerta MGEN, a standalone autoguider...