[chupathingie]

Jesus. Just when I thought we had you beat 😉

You better follow up on this, that mount looks promising.

And damn, be careful with that primary!

/me drools, that’s some sweet aperture…

[chupathingie]

Why do I get the feeling that the iphone/dslr situation has an analogy with the 35mm/med. format/large format comparisons? The iphone looks to have the best quality phonecam currently on the market, so the limiting factors at this point seem to be lack of manual controls for DOF and resolution (which is limited less by megapixels and more by diffraction of a teeny lens)…

[chupathingie]

Sweet Jeebus. I’ll be in the corner kicking the cat.

Yaknow, Guinness Draught isn’t all that unpleasant on the sinuses….

Rav, do what I did…cheap watch with a second hand. Do a bunch of subs, and DO DARKS. I can’t stress that enough… no darks or too few darks really make for disappointing results. Insufficient dark data actually adds noise to your finals, and to make it worse it’s patterned noise; you’ll see streaks that march across your final and there’s not much you can do to remove them.

I’m working on my routine yet. I have to come to grips with making proper light flats and bias frames. Light flats counter vignetting, and when you stretch levels vignetting becomes very obvious. Bias frames knock down “read noise”, which is the noise generated when your camera pulls the data off the sensor. It’s small, but even small noise sources become significant when you start shifting the histogram around.

The digging I’ve been doing the past few days has yielded a lot of interesting info. Find info about a couple of things for your specific camera… look for the ISO that provides the greatest dynamic range and the ISO that provides the greatest S/N ratio and shoot in between the two. If your tracking is good, lean towards the best S/N and go for longer subs. You’ll get much cleaner finals with a surprising low ISO, but tracking is only good for so much and you may have to bump it up some.

I know guys who guide manually; they’ll spend all night with their eye to the guide scope and their hands on the slo-mo controls. It’s nuts. Auto guiders correct for tracking error, but they also add cost… this ain’t a cheap hobby if you want to make it convenient, but you get some bragging rights for doing this stuff the hard way.

[chupathingie]

If you read an article that actually has some details about what they’re doing, it’s rather more interesting than you might think at first. For one thing, it’s not really designed to give you 40-odd megapixels of output, but to take that data and downsample a low-noise 8 MP image.

http://www.dpreview.com/news/2012/02/27/Nokia-808-PureView-with-41MP-sensor

That actually has some merit. My boss dropped an email on me yesterday about this, and I explained that 41MP is WAY beyond the diffraction limit for a tiny lens…. I’d have to look into this a little more before I can say that there is an appreciable benefit to this approach…

[chupathingie]

yaknow, that could be very useful for HA imaging if custom drivers are available… the eff. FL on that crop would be insane mated to some good tele-glass. I don’t think vignetting would be much of an issue :lol:, but guiding would.

[chupathingie]

oooooohhhh! me likey!

Some things require a telescope, and that’s one of them… I looked towards m51 the other night, wishing…
Do you have a remote for your camera? If not, one of these (or similar) is well worth the cash. Set it up, lock the mirror and let it go… if you put 40+ subs on that target, you’d get some seriously impressive results. It would bring out the dust lanes and nebula nicely.

[chupathingie]

purty! I wasn’t out early enough to get any of Andromeda, I’m gonna have to wait on that one. Which is fine, it’s such a huge, pretty target that I want to refine some techniques and routines before I spend a night on it. And it will be a full night, I want a lot of lights for that one so I need to start when it’s low in the east. So I won’t get a chance to really do much with it until august or september.

[chupathingie]

I’m happy to help you grok the scripting if you want. Post here, PM or send e-mail if you want to explain what you’re looking to automate.

I’d like to script the following command:

composite -compose subtract imagexxxx.tif unchagingimage.tif resultyyyy.tif

so that it applies the command to all files in a directory, where xxxx is the numbered portion of the file name (ie IMG_2765.tif for instance) and yyyy is either incremented from 0000 or simply copied from the original file name (the latter is preferred, but not absolute). “unchangingimage.tif” is a static file, ie is the single file I’m trying to subtract from all of the other files in the directory. This is my dark frame subtraction step in my processing pipeline.

I think I can make the rest tick easily enough. My script-fu is weak, and something I have been wanting to come to grips with for quite a while now. If you know of a good reference for a beginner to learn the ropes from, I’ll be happy to kick the tires. I’m poring over linuxcommand.org currently as I find time, and it looks to be a good reference, but if there’s an easier way I’ll certainly give that a go.

[chupathingie]

wheeeee!

[chupathingie]

^that. You’re looking at over a grand for the astrotrac, polar scope and heads. But it makes it SOOOO easy…

[chupathingie]

well thanky! 🙂 Both of those are crops from the planned widefield, but I left my UV filter on in error and the bright belt stars left very visible flares reflected in several places. The original FOV covered just north of Orion’s belt to just south of his sword. I was happy to find that 22 subs actually cleaned things up to the point that a full-rez crop didn’t look ugly at all; so I wound up with 2 images from the set. Anyone wondering about the benefits of a larger number of lights need only compare a single exposure to the averaged final. The difference is immediately obvious. The single image shows obvious static, while the averaged stack is silky smooth. I bumped up the black point and ran the gamma up above 2 before the noise started to become objectionable and backed down to a trade-off point between wispies and noise.

[chupathingie]

Choc, you are my new hero! I’ll put together a chain of desired events and post back here later. Scripting is something I’ve needed to learn for a long time, and a little hand-holding is most appreciated.

Rav, I wasn’t sure whether to start a new thread just for astropics or not or post here. In the end, I figgered you already the appropriate thread title going with the “Astrophotography: A beginning” thread so I tacked a couple of images to the end of it. 🙂

[chupathingie]

Okie dokie… I think I’m ready to post a couple. These are 100% crops of a stack composed of 22 lights and 18 darks, 2 min exposures at 1600 ISO on a 5DII thru a 70-200mm f2.8L @200/2.8. I did a few things wrong; 1st was leaving the UV filter in place (don’t do this, you will wind up with prominent lens flares), 2nd was wasting over an hour of shooting time troubleshooting by leaving the long exposure noise reduction on (still don’t know how/why it was on, but there you have it) and 3rd was to shoot only a series of 2 minute subs on a bright nebula… next time I’ll make time to shoot 15/30/60/120 second subs to recover data for the blown-out core.

Processing done strictly in linux with free tools from the repos. The heavy lifting is actually done with Imagemagick, believe it or not. I downloaded Hugin’s pano tools to align the frames, and Krita for final level adjustments. No noise reduction applied besides dark subtraction and an average stack. Aside from the final levels adjustments, all processing was done via the command line. More subs would smooth things out a LOT in the fainter fringes, my levels are stretched to the ragged edge to pull out the darker nebulosities in the outer reaches.

[chupathingie]

I need to process yet, which is going to be a mostly manual affair until I grok writing scripts in Linux to automate things. I shot about a dozen darks and maybe 30 lights worth processing. I learned some things, of course: remove the UV filter before shooting (I have some ghosts) and make DAMN sure in-camera noise reduction is off BEFORE you waste an hour and a half wondering why your camera hangs up after each shot.

My single images look far, far better than my stacked images taken from within the LP zone in DFW, and I ran off my frames at f2.8 and ISO1600 so gathered 16x the data with each exposure. I can see wisps in my singles that never made it into the finals before. It was so dark out there (here’s where we parked) that once we were outside the jeep for 5 minutes it was very easy to see the parking lot etc by starlight. The Milky Way was bright.

[chupathingie]

Oh. My. God.

An Astrotrac is worth every penny. Polar alignment is a snap (after collimating the polar scope, which is an ordeal), and this thing just works. I had to battle wind for most of the night, so experiments on some 200mm widefields on the Coma Berenices galaxy cluster didn’t work so well, but….I can’t WAIT to get these Orion wide-fields processed! 🙂

[Author]

Posts