M106 22 mins of Lum

[Rodd]

Very rarely am I greatly surprised in a good way in this endeavor.  I finally got the opportunity to collect some Lum for my HaLRGB M106 data set.  In a month of rain I got lucky with an hour of clear, steady skies.  I was able to collect 11 2min Lum subs using the TOA 130 and ASI 1600.  I was not expecting much out of 11 2min subs.  I could not have been more off-base.  This dos not look mush different than my 10-12 hour HaRGB image as far as detail and signal.  Will 90-100 more 2min subs add much to this stack?  Maybe some depth (though the background fuzzies are quite well formed. Its not overly noisy, and despite being shot almost entirely during nautical darkness, does not seem to have a high median, and contained no gradients to speak of.

I am confused.  I am posting this because very rarely in this endeavor am I confused in a good way.   I know it won't last!

Edit:  Not intended for full resolution viewing (that's where reality rears its inexorable head) .

[tooth_dr]

Another fine example of an image from the ASI1600 with little data, great job Rodd.

[Rodd]

1 hour ago, tooth_dr said:

Another fine example of an image from the ASI1600 with little data, great job Rodd.

What happened to the " we need 100s of subs to overcome the 12 bit format?"  Lets wait and see what 100 2min lum subs looks like--if I ever get a few hours of clear sky!  Maybe we will see a big difference.  Or there will be a big difference in the final image.

Rodd

[tooth_dr]

Please update us if you get a chance. Would love to see the comparison. 

[wimvb]

23 minutes ago, Rodd said:

" we need 100s of subs to overcome the 12 bit format?"

?

I [may] have said so, on occasion. Just shows you're better off not listening to what I say. ?

But I maintain that no matter what technology you use, there's no shortcut in AP; the total integration time to get good results (whatever your yardstick for that is) doesn't differ much. But please prove me wrong.

Btw, bumblebees can't possibly fly, as any aerodynamics engineer can tell you. It's just that bumblebees don't have a degree in aerodynamics.

Great image.

[Dragon_Astro]

To me, viewing on my iPad anyway, the background sky has a mottled appearance.

Maybe you could add more subs using dithering? If you have phd2 in your setup?

It would help to smooth out the noise

[Rodd]

1 hour ago, wimvb said:

?

I [may] have said so, on occasion. Just shows you're better off not listening to what I say. ?

But I maintain that no matter what technology you use, there's no shortcut in AP; the total integration time to get good results (whatever your yardstick for that is) doesn't differ much. But please prove me wrong.

Btw, bumblebees can't possibly fly, as any aerodynamics engineer can tell you. It's just that bumblebees don't have a degree in aerodynamics.

Great image.

Well, 22 min or 11 2 min subs.  Either seems like a pretty short yardstick!

[Rodd]

18 minutes ago, Dragon_Astro said:

To me, viewing on my iPad anyway, the background sky has a mottled appearance.

Maybe you could add more subs using dithering? If you have phd2 in your setup?

It would help to smooth out the noise

No full resolution viewing!?

[MikeODay]

Very nice Rodd.

[Rodd]

2 hours ago, MikeODay said:

Very nice Rodd.

Thanks Mike--I am hoping that a 3-4 (5-6?) hour stack will be better. 

Rodd

[Filroden]

16 hours ago, wimvb said:

But I maintain that no matter what technology you use, there's no shortcut in AP; the total integration time to get good results (whatever your yardstick for that is) doesn't differ much.

I agree with it but only to a point. Newer technology will (hopefully) continue to drive up QE and drive down noise, so you can reduce total integration time when comparing technologies from different generations. Are the newer CMOS sensors a new generation that have reduced the need? Probably not yet. And even if they did, there are always benefits to getting more data (though with diminishing returns).

19 hours ago, Rodd said:

Very rarely am I greatly surprised in a good way in this endeavor.

Your image does not surprise me. I think the 1600 produces an "adequate" image very quickly due to its low read noise but still needs a lot of data to make it "exceptional". I'll be really interested in seeing the effects of more data. Might be nice to stack a range of subs to see the benefits at, say 10 subs, 20, 40, 80, etc.

It's also interesting to see you brightest star shows lensing effects. I have that same issue on my first gen 1600.

[Rodd]

20 minutes ago, Filroden said:

I agree with it but only to a point. Newer technology will (hopefully) continue to drive up QE and drive down noise, so you can reduce total integration time when comparing technologies from different generations. Are the newer CMOS sensors a new generation that have reduced the need? Probably not yet. And even if they did, there are always benefits to getting more data (though with diminishing returns).

Your image does not surprise me. I think the 1600 produces an "adequate" image very quickly due to its low read noise but still needs a lot of data to make it "exceptional". I'll be really interested in seeing the effects of more data. Might be nice to stack a range of subs to see the benefits at, say 10 subs, 20, 40, 80, etc.

It's also interesting to see you brightest star shows lensing effects. I have that same issue on my first gen 1600.

Upon retrospective analysis, surprise goes away very quickly.  But remember, I have been using the 1600 for a short period of time (short scope time--long ownership time due to weather!).  I have also integrated several short data sets (not lum though) and always feel that much, much more data is needed.  Never have I integrated just 11 subs wit this camera that resulted in a passable stack.  This may be due to many factors--conditions being foremost among them, but still.  Also, I realize that this "surprise" does not necessarily survive transfer--I don't expect you to be surprised.    

As far as exceptional images with little data, it is possible.  Barry Wilson proves that.  Once the image is analyzed from an aesthetics point of view, processing takes over as the main consideration.  My surprise mostly came directly after integration--the first view in the linear state, prior to me getting involved with how the image looks.  I have integrated stacks of 4 hours that left me feeling quite the opposite.  My surprise was due to 3 main reasons:

1) No sub was collected during astro dark

2) They are lum subs, not Ha (which I often start during nautical dark)--and, according to the forecast, the Moon was still up (low and behind the trees so I did not see it)

3) This is a 12 bit camera.  In order to achieve results similar to the 8300 as far as dynamic range, it is recommended that many subs be captured--total integration times may remain similar, but the number of subs needs to be much greater.  This is pretty standard  advice regarding the 1600.  There is a chart somewhere on this forum regarding dynamic range and number of subs.

Since there is no magic in this world (unfortunately), there are obviously reasons why 11 2min subs rendered the above image.  But those reasons really do not eliminate my surprise.  After all, I am still shocked that when I turn on my gear, it works.  And it is supposed to work!  

The image is not a great image, and is not intended to be considered as such.  I just thought it was nice for me, for a change,  to be able to post an intermediary stage that I was not complaining about and seeking help for some problem.  

Rodd