I finally got a break in the clouds for little while last night. These images are all captured at secondary focus (f/10) of the C11 with the DMK 41AU02.AS.
Best 100 frames out of 500 stacked in Registax 5, "Dennis" wavelets applied. Some may have had a curves adjustment to lighten them a little, and maybe a little extra Smart Sharpen. I had exposure set to produce a mosaic of the whole visible moon at the time but closing clouds put an end to that. The 2 frame mosaic was assembled in PS3.
I did attempt some hires images as well, but struggled with focus in the poor seeing conditions (and being out of practice I forgot all about using a UV/IR filter)...
Looking at them on my desktop monitor it looks like I've over sharpened them - the detail is good, but I've got that grainy, gritty look to the smooth surfaces. I seem to get this processing on the lappy. It must be a lack of dynamic range or something on the lappy monitor. Maybe I'll have to buy a second screen to put in the obs so I can do a better job in future...
Quote:
Originally Posted by iceman
They're nice Al, I like the last one especially.
When did you get a C11 and DMK41?!? I missed that one!
Yes, the C11 arrived just before Lostock. Combined with all the other stuff, that's why its been so cloudy up here... not enough rain, too much cloud.
Still got a Lunt B1200 coming so I can use the SM40 on the ED80.
Al.
Last edited by sheeny; 24-11-2009 at 07:07 AM.
Reason: answer the right question
That’s a very nice portfolio of lunar images, some really nice detail in there not to mention the sweeping views with that monster chip of the DMK41!!!
In terms of the (slightly) grainy look, do you think that is due to over sharpening or perhaps just the relatively low number of frames you managed to stack from the sets of 500? I think that I once noticed a gritty look on my images with a low number of frames in the stack and this tended to lessen the more (good) frames I stacked.
I have noticed that really paying attention to the histogram in IC Capture paid dividends in choosing the optimum settings to record the best tonal range, even though its only an 8 bit camera and this gave me smoother looking images.
Nonetheless, a great start Al. I look forward to seeing more from that potent combination of the C11 and DMK41 – including yourself of course!
In terms of the (slightly) grainy look, do you think that is due to over sharpening or perhaps just the relatively low number of frames you managed to stack from the sets of 500? I think that I once noticed a gritty look on my images with a low number of frames in the stack and this tended to lessen the more (good) frames I stacked.
I have noticed that really paying attention to the histogram in IC Capture paid dividends in choosing the optimum settings to record the best tonal range, even though its only an 8 bit camera and this gave me smoother looking images.
Hmmm... I've never really thought of 100 frames as a "low number of frames" to be honest. I understand the principle of the more you stack the better the S/N ratio so the smoother the image, and I usually watch the quality curve in the stack graph to be my guide.
I have stacked as few as 10 frames at times when the quality curve drops sharply at the left hand side. Generally I've experienced that during poor seeing when only a handful of frames are sharp. I would, of course, prefer to get at least 30 frames on the basis of 30 being the lower limit of any statistical validity for sample size (drawing on my uni stats). I've done the experiment before to confirm to myself that stacking 10 good frames is better than stacking 50 or 100 mediocre ones...
It has been a long time since I've stacked 200-300 frames!
If I get a quality curve that's high and level from the left, it means one of two things: either the data is very, very good (and seeing was excellent) or the data is crap (poor focus, very poor seeing, etc) and will result in GIGO or at best a mediocre image after much effort (and usually downscaling).
I will have a play with this data and let you know how I go though, Dennis.
With respect to the histogram, I usually use brightness = 0, gain = 380 (I believe this setting optimises the dynamic range of the sensor. Merlin66 put me onto that one) and then adjust the exposure to have zeroes appearing about the 220 mark to avoid clipping the highlights. Do you differ significantly from this, Dennis?
For the Moon, at the prime focus of the Mewlon 180 F12 (2160mm) or C9.25 F10 (2350mm) I think I have the gain set to around 300-400 but this can vary if say, a bright region like Aristarchus or Tycho is in the FOV.
I usually have Gamma set hard left (can’t remember the value!).
And yes, I try to keep the whites at around 220/240 as often, seeing can make this number shoot up to 255 for sporadic moments.
From memory, provided the seeing is good (6 to 7/10) I tend to find that stacking 200 to 500 frames does give me more flexibility with the final stacked image in terms of trying to get nice tonal transitions and sufficient sharpening to make the detail pop, if it there in the first place. In the past I have really struggled with stacks of less that 200.
On nights of seeing 5/10 or less, I tend to end up with quite soft images that really aren’t worth the space on my hard disc!
Sadly, I don’t have your background in stats so I am more driven by the variables of just eyeballing the final results on an uncalibrated monitor, leaving scope for a lot of subjectivity and variation!
I'll do an experiment again and stack more frames and compare. I think my approach has been aimed at maximising the sharp detail, but I have to admit I have suffered from that "gritty" appearance of the apparently smooth surfaces. In some cases that may be real detail that is just exaggerated by sharpening into an artifact, but in others it is probably a poor S/N due to the number of frames I've stacked.
The other option that I want to explore is to apply a mask to the sharpened image to let an unsharpened layer below show through away from the edges. I have read a tutorial on how to do that ages ago, but I can't find it ATM. It involved using the edge filter to create a mask... I can't remember how to get the edge filtered image to the mask though.
This is the 4th image reworked and the original beside it so you can do a blink comparison.
I've also included a cropped screen dump showing the layers and channels dialogs in PS...
The procedure is:
1. Duplicate the image layer twice.
2. On one of the duplicates, apply the Edge Filter,then play around with Brightness, Contrast and Levels to get black where you want to sharpen and white where you don't, then invert this layer.
3. Sharpen the second dupicate and add any adjustment layers on top of it.
4. Add a layer mask to the second duplicate (the sharpened image).
5. Turn off all the layers above the inverted image we want to make into a mask, and make sure this layer is selected, then copy it.
6. Turn the sharpened layer back on, select it and go to channels. Turn on the Background copy mask channel and select it, then paste the mask into it.
7. go back to layers, and turn off the inverted mask source layer and turn the rest on.
Viola!
Al.
Last edited by sheeny; 24-11-2009 at 08:28 PM.
Reason: Uploaded wrong image
Actually, doing the blink comparison online I think I could push the mask blacker where I want the detail... fine tuning!
Al.
PS. I've changed the files in the previous post now... a bit better job. When you do the blink comparison the noise on the smooth areas should appear less now.
Al.
Last edited by sheeny; 24-11-2009 at 08:31 PM.
Reason: Changed files in previous post
Nice work mate – very informative! I blinked the 2 and I preferred the 1st image as it appears less gritty but still delivers all the detail I can see on my monitor.
Some good investigation and presentation work you’ve done there Al.
Nice work mate – very informative! I blinked the 2 and I preferred the 1st image as it appears less gritty but still delivers all the detail I can see on my monitor.
Some good investigation and presentation work you’ve done there Al.
Cheers
Dennis
Thanks Dennis. The original tutorial I saw on this, they were using this technique on macros and close up shots of flowers with impressive results.
. The 2 frame mosaic was assembled in PS3.
Now I have some reference pictures to go by. Gotta try to get close to this with my rig, when those clouds go away. 
I understand the principle of the more you stack the better the S/N ratio so the smoother the image, and I usually watch the quality curve in the stack graph to be my guide.
).
) and then adjust the exposure to have zeroes appearing about the 220 mark to avoid clipping the highlights. Do you differ significantly from this, Dennis?
