|
Post by southernskies on Sept 4, 2019 2:08:16 GMT
… Continued from Part 1
To put this method into practice, on the evening of 2nd September 2019 when the 3 day old Moon was sinking into the western sky I selected the Lagoon Nebula, Messier 8 as a suitable target. I first took three control images, each resulting from a live stack of 4 exposures – see images 1 to 3 below. The exposure lengths were 30 seconds, 1 minute, and 1 minute 30 seconds. Their dark point histogram settings were noted at 67, 108 and 137 respectively. The colour curve peak readings were at some small distance to the right of those settings. The 30 second exposure reveals the Hourglass Nebula. The I minute exposure swamps the Hourglass but shows some detail in the immediately surrounding part of the nebula. The 1 minute 30 second exposure reveals more detail in the outer parts of the nebula but swamps the inner part.
The author did not indicate whether the stack should start with the shorter exposure or the longer exposure. So I began by stacking 4 images of 30 seconds and then adding further images of 1 minute until 10 of those had been added. Each time that a new 1 minute exposure was added I took a snapshot of the result and saved it to file. Examining the results showed that the addition of 5 x 1 minute exposures gave the best compromise between retaining the Hourglass and displaying detail in the brighter parts surrounding it - see image 4 below. This proved that the concept works although the result did look a bit muddy.
Reversing the process I started with 3 exposures of 1 minute and then added 5 more exposures of 30 seconds – see image 5. This showed a much crisper result and I concluded that it was better to start a stack using the longer exposure.
Continuing along that line I then used 3 x 1 minute 30 seconds exposures to start a new stack and then added a further 4 x 1 minute exposures. The result was image 6.
The last stack started with 3 x 1 minute 30 seconds exposures followed by 4 x 1 minute exposures. It was ended by adding a further 9 x 30 second exposures so as to reveal the Hourglass reasonably satisfactorily – see image 7.
After the EAA session I applied a conventional HDR adjustment to the control images using Photoshop to obtain three new images for comparison with the EAA results. The first was a blend of the two shortest exposures – see image 8. The second was a blend of the two largest exposures – see image 9. And the third was a blend all three control images – see image 10.
Comparisons
The two shortest exposures (30 seconds and 1 minute) :
Image 5 EAA HDR
Image 8 Photoshop HDR
The two longest exposures (1 minute and 1 minute 30 seconds) :
Image 6 EAA HDR
Image 9 Photoshop HDR
All three exposures (30 seconds, 1 minute and 1 minute 30 seconds) :
Image 7 EAA HDR
Image 10 Photoshop HDR
Conclusions
The above comparisons show that the EAA HDR results compare quite well with their equivalent Photoshop HDR counterparts. They are perhaps not quite as crisp and bright but the differences are really minor. Thus there is no doubt that the method does work. Suitable viewing targets using this method should include the brighter diffuse nebulae, most of the globular clusters as well as the brighter galaxies such as Centaurus A and Messier 83. But the preparation time needed to identify the requisite histogram dark point settings and the colour curve peak readings as well as the extra time needed to allow for the different length exposures cannot be ignored. If you are doing EAA on your own they will probably not be a disincentive and this method could prove a welcome addition to your EAA toolbox. But it is probably a different matter if you are displaying EAA for other viewers such as at a public outreach event. Then the extra time involved may prove too long to sustain the viewers' interest.
|
|
|
Post by Dragon Man on Sept 4, 2019 13:52:12 GMT
Well written and explained.
I had a bit of a play with the HDR in RisingSky but because I am a proponent of 'fast' Video Astronomy I got fed up waiting for the image to build and found the whole process quite convoluted. And being a public Observatory the viewers like the instant gratification as well. Even when I do stack a live view I never go over 3 stacks, otherwise if defeats the 'immediate' view people expect to see. The general public (and myself) don't like waiting.
BUT, back when I used Miloslick to capture from my Mallincam cameras I often used its HDR feature which is quick and seamless. It automatically does all the steps need: amount of frames, differing exposures lengths, stacks, all with one click.
But it only works with the Analogue cameras.
Here's a video I made doing HDR in Miloslick:
|
|
|
Post by howie1 on Sept 5, 2019 0:07:45 GMT
Great stuff southern. Yes indeed, that technique is how long exposure astrophotographers do it too, in order to bring out the detail otherwise swamped by bright stuff. You may also one day like to try to use the same technique to create false-colour images ala those beaut colours in Hubble images. I'll explain ... very similar starting point to the HDR technique. Actually, you can also use the technique to create false-false-colour narrowband shots too ... without actually using the filters!!! I first read about this creative use of HDR without having to use HA, OIII and SII filters in posts by Sara Wagener. I shall put a link down below for her stuff. So first Hubble false colour images can actually be a whole gannut of combinations like this ... www.bintel.com.au/narrowband-preview-tool/?v=6cc98ba2045f In that link you would have seen you can assign the HA OIII and SII to various combinations of the Red Green and Blue channels in your astrophotography software of choice, to achieve different effects/colours. And (tada!) you also assign your low gain/medium gain/high gain images to various combinations of RGB channels to achieve different color effects! With or without filters! So the "fake" way of doing this without filters(!!!), is to do your HDR process of short, medium, long exposures (well however long you wish to consider long) and / or different gains at the one exposure time to achieve the same effect only with the same short exposure. But save the three shots/integrations image with saturation turned way down so they are all monochromatic. You then run the 3 mono frames through your AP software of choice assigned to whichever RGB channel combination you wish ... ala that site in the link above showing the different combinations. Here's Sara Wageners link ... www.swagastro.com/artistic-one-channel-processing.htmlCheers
|
|
|
Post by southernskies on Sept 5, 2019 5:41:36 GMT
Thanks Ken and Howie1.
As regards creating false colour images, I have a little experience from past use of a cooled mono CCD camera and narrowband filters. For example, this image of NGC 6188 using the Hubble palette.
So I was intrigued to see how the three control images I took would perform as simulated narrowband images. After converting each one to mono I tried using the Bintel online narrowband RGB combination preview tool as suggested. As required, I allocated the control images to the three channel slots provided and clicked the “Upload 3 images” button. This seemed to trigger some temporary “I'm busy” action in the web page's tab heading but without thereafter producing any result. So I gave that up and used Photoshop to manually allocate the images to the three separate RGB colour channels of a new image (in various permutations). In each case the result was essentially a single colour image, the colour being determined by whichever channel the longest exposure was allocated to. So no success there – see below.
I did read Sara Wager's article in which she referred to what she calls monochrome color mapping but it did not seem to explain how she goes about it. Nor could I find any such reference on her website. Not to worry - it was fun looking.
Cheers
|
|
|
Post by howie1 on Sept 13, 2019 14:55:56 GMT
Here you go Southern ... couldn't sleep so I made a couple of video's for you.
First up ... false colour images from OSC without filters
Next ... Stacking using PeekThrough to reduce noise ... without AstroToaster, SharpCap etc.
Cheers
|
|
|
Post by southernskies on Sept 20, 2019 6:02:52 GMT
My apologies for not responding sooner Howie1 but a storm two weeks ago knocked out my internet connection. Those are two great videos. You have gone to a lot of trouble and they are very informative. I will follow up on the ideas presented and I’m sure they will be of much interest and benefit to other readers of this forum.
Cheers
|
|
|
Post by howie1 on Sept 23, 2019 20:18:55 GMT
Thanks Southern. I must remember that the ShareX softwares HD video capture always appear MUCH darker than what I actually see on my laptop. The resulting image in the "poor mans stacking" video looks very dim in the video. Sorry about that, as it doesn't really show how nice the end image looks. Still, the idea/process using peekthrough is shown good enough in the vid to get the idea.
Cheers
|
|