Here is a test RGB image of NGC6888 just to see how the partially-collected data might look together. No Red collected yet, so this uses Green and Blue in their proper roles, with Hydrogen-Alpha as the Red channel. 160 minutes Ha (as Red) 120 minutes Green 180 minutes Blue Calibrated & stacked in CCDStack, post-processed and colour-combined in Photoshop.
With a little imaging time left in the evening, I started on NGC6888, the Crescent Nebula, as it is now well positioned in my limited hole in the sky.
The moon was quite bright – between 1st quarter and full – and washing out the sky a little. However, NGC6888 is an emission nebula, so produces a fair amount of H-Alpha light, and this is much less affected by moonlight.
The first night I managed 70 minutes in 10-minute subs. The next night, 90 minutes more in 15-minute subs. So, for the start of this project, this is 160 minutes of H-Alpha in 10-minute and 15-minute subs.
More data on this target will take several weeks, I imagine.
It was a pleasant evening last night – very clear, and unseasonably warm (+17 degrees at midnight, after frost the night before).
I was able to capture an hour of Red data for M13 under good steady skies. I thought it might be interesting to share the colour channels data for such projects as they are collected, in case others might like to use them to experiment with processing, colour combination, etc. (I found few such examples to play with when doing my own searching some years ago.)
I re-did the blue channel in 3-minute subs, getting 27 of them before the target sank below my hedge. This larger number of subs reduced the noise in the channel quite a lot (although it is still a bit noisy).
Also reprocessed more carefully, especially by taking the individual colour channels into Photoshop and removing the skyglow gradient from each (using Gradient Xterminator).
This is starting to look good; I’m pleased with the direction. If there are more chances this season, I’ll add more subs in all the channels.
I have gathered at least some data now for 3 colour channels and luminance, and did a first combination tonight. It needs more colour data to improve the signal-to-noise ratio, especially in the Blue channel which was taken through light cloud and is very noisy. However, it’s still quite pretty and I’m pleased with the result so far.
This would be 90 minutes of Luminance, and 60 each of Red, Green, and Blue binned 2×2. QSI583 camera on an AT8RC, autoguided with a Lodestar, all on a Losmandy G11.
I’ve started a multi-night project for M51. July 28 was the first chance to catch Luminance data, 1.5 hours so far (in 5-minute subs). Sky-glow is fairly bad in this part of the sky (removed manually in the attached shot), and will require some careful processing as I get more data.
A satellite trail wasn’t completely removed via the median-combine. Once I get more exposure time, I’ll probably just discard the frame with the satellite.
Update: Problem in Colour Darks
I’ve had a couple of evenings chance now to gather colour data, and have captured R, G, and B, all binned 2×2, on separate nights. I was quite disappointed with the results – all are showing nasty black blotches. Looking for the cause of this, I believe my 2×2 calibration dark frames are faulty (I keep a library of darks at different reference temperatures and different exposure times, and think my 3-minute, -15 degree darks are bad).
It’s supposed to be cloudy for a few days now, so this will be a good chance to take new darks, then test those against the acquired data to see if that was the problem.
Here the Ha is mixed in with Luminance, increasing the intensity of the gas envelope in the nebula. The colour balance came out a little different, but I rather like it so I’m leaving it as is.
Backing up and re-focusing on some basics has paid off. I’m quite pleased with this first result on M27. This is 4 hours of data over about 2 weeks, 1 hour each of LRGB, with RGB binned 2×2. QSI 583 camera on an AT8RC, autoguided with a Lodestar, on a G11.
What I did differently to make this result better than previous ones:
- Better autoguider and more attention to guider settings
- One colour per session, taking time and not rushing
- Focused for each filter
- Flats for each filter, immediately after the data gathering, so filters are certain to be in exactly the same spot
- Normalized & stacked with CCDStack, basic processing in Maxim, more processing in Photoshop.
The problem is, this is quite addictive. Now I want more data – several more hours of each, and I can see never being satisfied. More data to follow when the clouds leave.