The core of the Milky Way Galaxy with Rho Ophiuchi captured from home (SQM=18.8, Bortle 7) on 18MAY2020 just before Moonrise, with the bottom of the frame just above our roof.
63 stacked (untracked) exposures of 13s @f/2 ISO800 (13.6 minutes total exposure time), Nikon D600, Sigma Art 35mm f/1.4, Marumi Filter. No Darks & No Flats used. ACR lens corrections & then image stacked in Sequator & cropped.
I'd like to get a longer exposure for more Rho and MW colour next time without the scene getting too bright. I may have to dust off the mounts, but I'm happy with untracked (and its portability) none the less.
Nice tight shot there JA but its very dark for the exposure time. It does stretch and that will bring up a few gradients as well but those are correctable with a few tools - Pixinsight dynamic background extraction, Russell Croman's Gradient Xterminator or the Photoshop "apply image" trick.
Nice tight shot there JA but its very dark for the exposure time.
Thanks Greg. Yes I agree a little dark, but not sure I could do much better untracked since I was at 13 seconds @f/2 iso800 and I loose a stop or so with the Marumi Filter so it's as if it was around 6 seconds @f/2 iso800 in terms of light gather, which is very light on for subexposure given the other colourful Milkway images we see here (30 to 90 seconds) , I suppose depending on howmuch one wants to stretch and/or push the camera sensitivity .
Quote:
Originally Posted by gregbradley
It does stretch and that will bring up a few gradients as well but those are correctable with a few tools - Pixinsight dynamic background extraction, Russell Croman's Gradient Xterminator or the Photoshop "apply image" trick.
There's more in this one.
Greg.
Yes good point, I did stretch somewhat and try stretching further, but as you rightly point out, the gradients needed to be contended with. I was using Sequator for the stacking and its light pollution removal option and I was not using flats and the resulting gradients / artifacts from stacking uncorrected vignetted images were significant. I then used lens correction in ACR in lieu of flats which improved the gradients, but they were still objectionable particularly with a brighter stretched image. I don't do Pixinsight, but I did try various Photoshop tricks. I may have to bite the bullet and use real flats to see if there is any improvement, but still given that I was about 50 metres from Nepean Highway with all those lovely High Pressure Sodium Lamps (SQM=18.8), I'm sort of OK with the result . Proper flats, tracking and much longer subs may work if the light pollution doesn't blow things out too much.
I did a apply images process that removes gradients and it got rid of a lot of the gradients. I can post that if you want and the process.
Your stars seem quite round so you could probably go a little longer and sacrifice a bit of star roundness for more exposure.
Greg.
Hi Greg,
Yes I wouldn't mind seeing it, but just to advise the image I posted is a slight crop (leaving out the messy bits) of the full frame image directly from the output the Sequator after its Level 3 Uneven Light pollution reduction, plus or minus some minor level/contrast adjustment. Because I didn't use true flats, and because it is somewhat buggy (according to various sources) it creates all sorts of strange edge gradients. Take a look at the sample below which shows from left to right: No Light Pollution Reduction followed by Level 1, 2, 3, 4 and 5 Light Pollution Reduction using the Sequator Tool- I suppose their form of background elimination. I cropped out the too hard basket stuff in the Level 3 image and called it good enough, until I could get Flats. The artifacts or gradients in the the images (especially the 4th, 5th and 6th image) are not evident in the subs or final stack and somehow created by the process (as also mentioned by various sources on the net).
I did try some Ps "Apply images" work, with various settings and masks, but gave it up once I discovered that the real solution may be in using true flats and possibly using DSS instead. I'm not 100% sure of that but i will try that and also compare it with DSS. To top it all off I discovered that in the later half the focus had shifted and the final stacked image was a little imprecise, although at the image size maybe I can get away with it somewhat.
Best bet to solve everything and not have to dig so deep- get out and take some longer exposures somewhere darker than home.
Duplicate image in Photoshop
dust and scratches filter set to 500 to blur it out.
Usually here I use the healing tool to remove any blurred lighter outline of the object but here there was none.
unselect the duplicate layer open apply image under tools and select subtract. Select the background copy and you can set the offset to various values to suit - 25 worked here.
This technique basically creates a false flat and subtracts it.
Not sure where you are going with this....rather than heroic efforts to salvage average data....maybe get better data?
Vixen Polarie's are compact, cheap, and work well. (there are other options, but a good quality tracker will help immensely)
Sure imaging from the burbs is not much fun, but if you do, it is all about maximising signal and minimising noise. Seems the first and critical bit, is being ignored here: let the signal build without being ruined.
Stick with it, but it will be hard to do much better leaving things as is.
Not sure where you are going with this....rather than heroic efforts to salvage average data....maybe get better data?
Yes...True that Peter Originally I had not started this with a view towards creating an image, but rather as a study to visually demonstrate the effect on noise of stacking an increasing number of dithered images, as I had done in another thread.... This sort of thing ....
Unfortunately a tiny shift (error?) in focus after the tenth image in the 64 image run, meant I couldn't really do that to my "scientific" satisfaction, as normally I use 20 times or so enlargements of a selected image area, so instead I decided to make something arty, albeit somewhat dark. (In the meantime I used some other images for an unfortunately too short, similarly targeted 45 image run, which I may post in another thread when completed.)
Quote:
Originally Posted by Peter Ward
Vixen Polarie's are compact, cheap, and work well. (there are other options, but a good quality tracker will help immensely)
Sure imaging from the burbs is not much fun, but if you do, it is all about maximising signal and minimising noise. Seems the first and critical bit, is being ignored here: let the signal build without being ruined.
I have tracking mounts, but I chose not to use them so as to (lazily) take advantage of the inherent "auto" dithering afforded by using a fixed tripod as I wanted to look at the noise as a function of the number of exposures in a dithered stack, (in particular the colour mottle from a DSLR) but still did not want excessive star trailing as I was using a 35mm lens on full-frame and I went for a 13 second exposure at my favourite ISO800, fairly wide open at f/2. That's really as much "build" as i could afford on a fixed tripod, given my aim.
Quote:
Originally Posted by Peter Ward
Stick with it, but it will be hard to do much better leaving things as is.
Thanks Peter. I may get it better with some real flats, rather than lens corrections, given the funky gradients/artifacts from Sequator's Light Pollution reduction algorithm, which I have to say is amazing from certain perspectives. (I haven't yet graduated to Pixinsight, etc...)
. Proper flats, tracking and much longer subs may work if the light pollution doesn't blow things out too much.
