ngc6302 Ha now with colour added + super-res stack
Used some fairly low quality RGB data to add colour to an earlier posted Ha image. Blending Ha with the noisy low res R was not an option, so the Ha was used directly as the luminance - this skews the colours a bit (particularly the star field), but the addition of colour of any sort seems to have brought out some structural features that were not as obvious in the Ha, so it seems to be worth having. And it looks more dramatic - that must have been some explosion.
Thanks for looking. Any ideas on how to better combine hi res Ha + low res RGB gratefully accepted. Regards ray
200f4 RCC1
SX H694 at 1.17 arcsec/pix
NEQ6
heavy crop - scaled 1.3x
Edit: added a manual sort+super-res stack of the Ha data - gained a tiny bit more detail.
That's really good Ray... you've caught some high res details in there. What exposure lengths did you use??
Hi Lee.
Sorry, forgot to include exposures:
132x1m 6nmHa 1x1 (good steady seeing)
34:71:41 x20s RGB 1x1 (mediocre seeing - grabbed between clouds with nearby moon)
regards ray
That looks awesome Ray Looks like you got a fair amount of detail in there as well.
Erik
thanks Erik. the Ha seeing was good (got below 2 arcsecs FWHM on occasion) and that allowed some of the finer detail to come through - which was fun to see on the screen.
I'd have to say that's the most detailed 6302 I've seen, quite like the subtle colour.
Cheers,
Justin.
thanks, that's very generous Justin! I would like to be able to get better colour, but that will have to wait on these skies clearing a bit and the moon moving on.
Quote:
Originally Posted by Lee
Ray - how do you get by (and get great results) with such short subs? especially with a 6nm Ha filter?
Hi Lee
There are two major advantages to the ICX694:
1. it has 65%QE at Ha, which makes it more than twice as sensitive as many other CCDs - you can use under half the sub length (and total exposure time) for the same detected signal,
2. it has about half the read noise cf many other CCDs, so you can take 4 times as many subs for the same contribution of read noise to the SNR (read noise adds in quadrature).
ie, a single sub of 10 minutes with some alternative CCDs is roughly equivalent to 4 subs of 1 minute with the 694, if the aperture and pixel scale is the same for both.
In practice, on a bright nebula, I use pretty much the same settings as any other filter when imaging in Ha. Of course, short subs means that the tracking of an EQ6 is just about good enough (not perfect, but not too bad over 1 minute) and subs from periods of short term degraded seeing and wind gusting can be deleted to improve resolution. In windy conditions, have been able to image effectively with really short subs - the colour for the posted image was from 20 sec subs.
The icx694 really is quite a revolutionary CCD from a systems perspective.
There are two major advantages to the ICX694:
1. it has 65%QE at Ha, which makes it more than twice as sensitive as many other CCDs - you can use under half the sub length (and total exposure time) for the same detected signal,
2. it has about half the read noise cf many other CCDs, so you can take 4 times as many subs for the same contribution of read noise to the SNR (read noise adds in quadrature).
Certainly more than twice as sensitive as my ICX274...
I think I'll try doing more-shorter subs rather than fewer-longer subs and look at the difference....
Antibiotics ain't gunna help either...it's a Super Bug
Mike
thanks Mike
Quote:
Originally Posted by Lee
Certainly more than twice as sensitive as my ICX274...
I think I'll try doing more-shorter subs rather than fewer-longer subs and look at the difference....
if any help Lee, ran your system through a system performance model - that suggests that subs down to ~1minute should be quite OK for broadband under dark sky, but you really start to lose SNR on average targets if you drop much below about about 3-5 minutes for Ha. On dimmer targets, the longer the better. Your chip has really low read noise, which helps a lot with shorter subs. If you do any tests, be interested to know what you find.
Stunning resolution, Ray. Best I've seen of this target for a long time, clearly the short exposure work well for you. Similar principle to 'lucky imaging', though exposure times are radically shorter compared to 60s subs. Not sure of the approach you are using to scale low resolution RGB data to the high resolution luminance but it is working for you so why change it. Great result. The image has an eye catching 3D feel to it. More please...
thanks Ross - it was great to have nice seeing for the Ha data.
Quote:
Originally Posted by jase
Stunning resolution, Ray. Best I've seen of this target for a long time, clearly the short exposure work well for you. Similar principle to 'lucky imaging', though exposure times are radically shorter compared to 60s subs. Not sure of the approach you are using to scale low resolution RGB data to the high resolution luminance but it is working for you so why change it. Great result. The image has an eye catching 3D feel to it. More please...
Thanks Jase. I guess this is at the "mildly fortunate" end of lucky imaging - 1 minute subs certainly allow the odd messy bit of seeing to be excluded and the camera is quite happy working with short subs. Used dynamic align in Pixinsight for combining Ha and RGB.
Quote:
Originally Posted by marc4darkskies
Wow, 1 Min Ha Subs?!! 20 sec RGB subs?!! Very impressive image Ray - great resolution!
Cheers, Marcus
Thank you Marcus.
This is brighter than the average target, so fairly short subs were used just to keep the core region from saturating (only to find that it looked best overall if was slightly saturated in processing). However, as a general rule, the icx694 allows quite short subs without too much penalty if desired - modelling suggests that the optimum dark sky sub for broadband with this system is 2 minutes. Ha should be as long as possible, but it can still deliver good results with subs below 5 minutes on average targets. It is also great fun to use - apart from waiting for the PC to do a super-res stack on over a hundred 6mp images
Just for interest, manually sorted and super-res stacked the best Ha data - noisy but gained a slight bit extra detail, so should have tried it earlier.
Looks like you got a fair amount of detail in there as well. 
thanks Mike
