I have taken a few Ha images but with the recent arrival of the OIII filter and the imminent arrival of the SII I have decided to have a go at NB composite. Everything Ive read says somehting like "take your NB images and process them as normal prior to combining" . This is a 4 hr (16 x 15 min subs) image - a lot I know but it was last weekend, the moon was bright and I only had a Ha filter - so I thought I may as well keep capturing data - calibrated just with darks and flats, and stacked then auto levelled and first light sharpen filter in PS (the spikes are just to see if I could do them - they'll go) - no curves - no gamma corrections etc etc.
To me this looks like what I would think a Ha image of this object should look like. Seperate to a stand alone Ha image, do any of the nb composite gurus out there have any tips regarding processing prior to false colour combination.
PS - I wasnt allowed to go to the observatory and play this weekend to capture OIII due to some Sunday function with my wife and the kids (Youre not my mother was my arguement - that went down like a lead balloon) I ducked really quick. I didnt push it because I really want that Meade 10" ACF in a cuple of weeks.
Joking because I think I would rather see my grandkids than stare at PHD for hours. Also any of you Qld guys that know us should know that I would be singing soprano if I tried to say that seriously.
Anyway any info re use of this image (and / or adjustments required) for use in a NB composite would be rgeatly appreciated. The learning curve gets steeper.
Black Diamond ED120, Orion Starshoot Pro 2, binned 2x2, EQ6 and processed as above. Taken at Leyburn Observatory
Obviously the jpg compression looks very little like the fits but hopefully you get the gist of my question.
dont know about narrow band processing, but it looks like some really good data through H alpha-so why not go the whole hog -curves and so on and finish processing as you would normally anyway?
Looks like a nice lot of data here to work with. It would be a waste to leave it as it is without doing a full work up in PShop. Nothing to loose by giving it a try.
Should make a nice Lum layer for LRGB or a good base to work on for NB composite.
I have been pondering the degree I want to push it. If I use it as Lum layer in a nb composite Hagar (dont forget I am a newbie to this multi channel stuff this is more of a question than a statement) as I understand it this is where I do all the processing (ie sharpening, levels etc) and I use a seperate set of nb images of shorther duration to make up the RGB channels from SII, Ha,OIII just for the colour.
Or do i do a just a SII, Ha,OIII with this as the Ha channel - I will try all of these obviously, once I get the the SII, OII this weekend - cloud gods permitting.
I did a OSC of the object a few weeks ago when planning the shot and like an idiot posted a version where I was showing Mrs Bolts_Tweed what gamma variation did - I have been laying low licking my wounds since then - THEN the day after I took the Ha Tom Davis posted his image of 6188 - dont know if it gave me inspiration or frustration - joking - astrophotography is like golf - I just want to lower my handicap - not beat Tiger Woods.
Thanks for the advice and info guys and I will post the (correct) results hopefully next week.
cant say i know much about narrowband, but i think youve got the right drive and are willing to experiment so you will sort it all out, learning all the tricks ... or at least most , takes years, guys like strongman mike, martin pugh, jase and others all started out small and probably would not want to be seen next to their early work. so keep at it
Quote:
I wasnt allowed to go to the observatory and play this weekend to capture OIII due to some Sunday function with my wife and the kids (Youre not my mother was my arguement - that went down like a lead balloon)
I like how it looks so far... Really interesting target! I must have a look for it myself.
If this is going to be your lum, I would sharpen it up a bit, process it to perfection... then take the original stacked file, process that a fair bit gentler, and dont worry about having it super sharp, as this will be a colour layer, it doesn't need to be excessively sharp or detailed, it just needs to be strong data... In an LRGB, HaRGB, HaSHO etc image, the luminance layer is where the image gets all its detail, sharpness and structure, think of the RGB layers as simply a set of crayons with which you will colour in your drawing preformed with the fine liner (luminance)
I have not done full NB coloured images, however I've used Ha data as a Luminance for my a few of my images, and thats how I've always understood the process. Luminance needs to be quality quality data ( like what you have here..) colour images dont have to be anywhere near the same quality. A lot of people will take their luminance at bin 1x1 for best resolution, and take their colour subs at bin 2x2 or even 3x3 as the resolution doesnt matter too much. You're just looking for colour.
I used to bin 2x2 using my QHY8 for Ha imaging as you get a lot more sensitivity, and the resulting image is monochrome. What I later found out was that doing this halves your resolution for starters, and your resulting image will be noisier.. The bayer matrix on your camera, like with the QHY8 assigns some pixels as Red, green and blue. Even when you bin 2x2, these pixels are still only sensitive to their assigned colour even if the resulting image is monochrome. You will achieve a cleaner image if you image narrowband bin 1x1, and then once you've stacked your colour images, for Ha, take out only the red channel.. When you do this, take a look at the green and blue channels, they will be nothing but noise. Thats because the Ha filter is stopping green and blue light hitting from getting to the sesnor. so by just using the red channel, you're removing a lot of the noise, and getting an overall better Ha image to use as a luminance for your final image.
Thx for the info Alex - answers a few questions - I was thinking along the lines you have explained wrt Lum. I took some SII, OIII last weekend but I have found a serious filter imbalance requiring far more data on the SII than the Ha (different manufacturers etc). Anyway that gives me heads up and encouragement to keep going with this.
Mark,
Using narrowband filters from different manufacturers wont make a difference. There's no channel cross-talk between SII, Ha or OIII filters. If there was, the filters wouldn't be narrowband, but broadband like conventional RGB filters. I know of people using a Custom Scientific 4.5nm Ha filter and use it with OIII and SII data acquired with 6nm filters. I'd use the SII data you've got. What I suspect you're trying to portray is that you haven't got enough SII data. This is a common problem with narrowband images as the SII filtered data typically requires much longer exposure than Ha and OIII. A weak SII channel is miserable to deal with, especially on SHO palettes, not too much of a problem with HOS. On occasion, I've blended Ha data into the SII to address this. Such an activity would upset the narrowband purist, but in the quest of producing a reasonable result, it works ok. Alex's advice is good. Many bin the SII and OIII data reducing the amount of time required to collect quality data. This however will depend on how you intend to process the image. With PS clipping masks, you process each channel to extract as much detail as possible, then combine. Its preferable to have unbinned data in this scenario. Binned works ok, but the features in each emission line will be weaker that the Ha lum. There is plenty of info in on processing narrowband images in other areas of these forums. A search will reveal them. Above all, don't be afraid of experimenting with the data you've got. In some cases, narrowband is easier to work with as there is no such thing as colour balance or at least not in the traditional sense of G2V calibration. Understanding the data's weakness will allow you to focus on what needs to be done to address it, be it acquiring more data for a specific channel or use alternate processing methods to extract the desired result. Keep at it.
Yoiu pretty well summed up what I was trying to say but it just goes to show that i shouldnt believe what I read or am told re filters. Coming from an experienced astrophotographer like yourself makes me believe I may have the image hidden amongst all this data. All up I have 12 hours of data on this so I might just hit it all again - its been a week since I slammed the laptop shut .
Thanks for the reply and I will keep playing with it all
Dont try to combine before PS, load them seperately into PS as colour channels. NB needs more (different) adjustment than RGB in each channel, which is easier in PS.
Take Ha at bin 1, as many subs and the longest exposure time as you can get, and use it as the Lum channel (where all the detail and sharpness is) and less OIII,SII subs at bin 2 to increase data with less exposure time. The OIII and SII subs can be really crappy without affecting the final result, just blur them to hell to get rid of noise, detail is not important (well, to get started anyway).
I ducked really quick. I didnt push it because I really want that Meade 10" ACF in a cuple of weeks. 
this is more of a question than a statement) as I understand it this is where I do all the processing (ie sharpening, levels etc) and I use a seperate set of nb images of shorther duration to make up the RGB channels from SII, Ha,OIII just for the colour.
