I was hoping to collect more data but some people of this forum (not pointing fingers at anyone in particular...) have been recently buying telescopes and good weather has ended
lol. How inconsiderate of them.
Quote:
Originally Posted by Slawomir
Now the questions is...should I get the scope out and collect more data for this image? It's a clear night and new moon...
I reckon you could smooth some of the fainter signal out with a little bit more data, but whether it's worth the effort is up to you; more data is always nice but there's a point you reach when you'd rather go image something else.
On the other hand, if there's nothing else capturing your fancy right now, then why not?
I decided to take advantage of clear skies tonight and collecting more Ha data for this nebula. Maybe tomorrow night we will be lucky again here in Brissie and I will add the last 2 or so hours to Ha.
Anyway, it looks like I managed to eliminate most of backlash - SGP has put the scope right on the target on the first attempt to within 3 pixels twice already- at the beginning of the session and after meridian flip. Currently guiding is as good as I ever got with AZ-EQ6.
That's an excellent image. I hane not seen this in NB before and it works a treat. The fact that there isn't much Oiii means that what there is is very subtle in the overall scheme of things and it comes across just right.
That's an excellent image. I hane not seen this in NB before and it works a treat. The fact that there isn't much Oiii means that what there is is very subtle in the overall scheme of things and it comes across just right.
Top job
Thank you David. Yes, there was very little O3 and thus it required some work to get reasonable signal out of it.
Quote:
Originally Posted by Rex
Beautiful image Slawomir, and a different take on a fairly common object. Love the warm colour palette. Outstanding!
I decided to take advantage of clear skies tonight and collecting more Ha data for this nebula. Maybe tomorrow night we will be lucky again here in Brissie and I will add the last 2 or so hours to Ha.
Anyway, it looks like I managed to eliminate most of backlash - SGP has put the scope right on the target on the first attempt to within 3 pixels twice already- at the beginning of the session and after meridian flip. Currently guiding is as good as I ever got with AZ-EQ6.
Awesome! Looking like another clear one tonight, at least up in my neck of the woods so you should be about to get those last couple of hours.
Added another 4.5 hours, bringing the total integration time to 23.5 hours.
Noise in the background went down by about 10%. I only applied very soft noise reduction, as I prefer noisier images that preserve detail over smooth and overcorrected.
I think I am happy with the current result and will probably leave it as is.
I like that even more now Slawomir. Colour is a bit richer now too. If you have not already, you should submit for APOD. This is a fairly unique looking image and might just grab their eye.
With regard to noise. I have found that no noise control is necessary after a certain amount of time in each filter. It depends on the object. Some objects require more data than others. Obviously the brighter the object the less data required to over whelm the noise. Although some objects such as dark nebula require deeper exposure even if they are close to a bright object. In your full res now the noise level is extremely low (like a very mild fuzz if anything) and not worth going any further in my opinion.
A bench mark image now of this object. Until now I would not have considered doing an NB image of this object. Excellent work.
I like that even more now Slawomir. Colour is a bit richer now too. If you have not already, you should submit for APOD. This is a fairly unique looking image and might just grab their eye.
With regard to noise. I have found that no noise control is necessary after a certain amount of time in each filter. It depends on the object. Some objects require more data than others. Obviously the brighter the object the less data required to over whelm the noise. Although some objects such as dark nebula require deeper exposure even if they are close to a bright object. In your full res now the noise level is extremely low (like a very mild fuzz if anything) and not worth going any further in my opinion.
A bench mark image now of this object. Until now I would not have considered doing an NB image of this object. Excellent work.
Thank you Paul very much for your feedback.
I tend to get more noise due to my location (heavy light pollution), so I usually try to apply a tiny bit of noise reduction, aiming for the background areas. I used to overdo noise reduction in the past...
Out of curiosity, and because I like simple graphs and basic maths, I integrated different number of calibrated Ha subs from this project and compared standard deviation for the same area for all of them. I picked a uniform patch of the image that represents background only. (see the attached graph).
Interestingly, equation that fits the trend line is pretty much spot on with what one would theoretically expect...St.Dev being inversely proportional to the square root of the number of subs.
The equation: Noise (st.dev.) = 50 / sqrt (n)
or generally: Noise (St.dev) = noise_in_one_sub / sqrt (n)
Might be a useful tool for roughly estimating noise at the beginning of a project and helping to decide/plan on the amount of integration...maybe.
Quote:
Originally Posted by codemonkey
Nailed it, S! That extra bit of data turned an awesome image into something, well, even more awesome. As Paul says, you've just set the benchmark.
Thank you Lee. It was your suggestion that prompted me to get a bit more data...
Last edited by Slawomir; 15-08-2015 at 03:44 PM.
Reason: Clarity
I like that even more now Slawomir. Colour is a bit richer now too. If you have not already, you should submit for APOD. This is a fairly unique looking image and might just grab their eye.
What Paul said :-)
And I'll add: excellent apparent depth
This is a fairly unique looking image and might just grab their eye.
Hi Paul,
Whilst I agree with the sentiment of your post, I have to go all English Nazi on you and point out that nothing is "fairly unique", it's either unique, or not. Having said that, aren't all of the images here unique?
That's an interesting post Slawomir. So the bulk of the noise drop is done by the first 0 or 5 subs and then its a much slower rate of improvement and by the time you've done 50 you are not going to see much better by doing more at increasingly more work and time.
I wonder what optimum is then between weather, setup speed, work, object location, equipment foibles.
8 to 20 hours on a faster system and 20-30 for a slower system? Depending on the object's brightness and whether its LRGB or narrowband.
What Paul said :-)
And I'll add: excellent apparent depth
Trev
Thank you Trevor!
Quote:
Originally Posted by rat156
Hi Paul,
Whilst I agree with the sentiment of your post, I have to go all English Nazi on you and point out that nothing is "fairly unique", it's either unique, or not. Having said that, aren't all of the images here unique?
Cheers
Stuart
Please do not pick on me...
Quote:
Originally Posted by Atmos
That is a very nicely looking Paw, like a cat has run over some hot coals
Thank you Colin, I am glad you like the colours.
Quote:
Originally Posted by multiweb
The high res is superb. Awesome details. Looks very fiery like an explosion. Top shot.
Thank you Marc. Took some time to get there (11 evening sessions)
Quote:
Originally Posted by gregbradley
That's an interesting post Slawomir. So the bulk of the noise drop is done by the first 0 or 5 subs and then its a much slower rate of improvement and by the time you've done 50 you are not going to see much better by doing more at increasingly more work and time.
I wonder what optimum is then between weather, setup speed, work, object location, equipment foibles.
8 to 20 hours on a faster system and 20-30 for a slower system? Depending on the object's brightness and whether its LRGB or narrowband.
Greg.
Agreed. There are many variables affecting the final image, so it might just serve as a rough guide. I think if one is after really faint detail, then perhaps every good sub would count. But it looks like for my setup and in my location and with my current skills, going over 25 hours with 3nm filters would be an overkill for majority of targets.
