Click said:Jon, I have a question for you.
What are those triangles on both sides of some of the stars?
Interesting, they seem to be concentric with the center of the image
Upvote
0
Deep Sky Astrophotography
- Thread starter jrista
- Start date
Interesting, they seem to be concentric with the center of the image
jrista said:Thanks.
So, there is noise. Lots of noise. It's only 2 hours of integration, and I probably need a minimum of 5 to really do it justice. It's just not easy to get that data, since there are only a handful of dark nights a month, and Orion is really racing towards the western horizon. I hope to gather another three hours at some point, which should help.
The 5D III, as I've tried saying so many times on these forums, is one craptastically noisy sensor! It is NOT, buy today's standards, a low noise sensor, at all. Which is a little sad, for a camera barely three years old. That's where the color noise comes from.
As for electronic noise overall, it's actually fairly low. I used ISO 1600 specifically to get read noise low. Its around 3.x e-. I was also imaging at around 3C (it's the heart of winter here, nights are 15-18F), so the dark current is very low. The reason the darker regions look noisy is they have been very significantly stretched. I had 21.3mg/sq" skies where I imaged this, which is getting pretty close to the darkest possible 22mg/sq" skies on earth. That was necessary to even get a reasonable amount of photons on those dark areas. Still, on a per-sub basis, the darker areas probably only had maybe 5-8 photons/pixel/minute tops!
So, yeah...there is noise. There is always noise, and when you do a ludicrous stretch like I did, that noise can present a bit of a problem. The only solution is to expose long enough to swamp read noise, and integrate more and more. I need three and a half more hours of integration for my minimum, and I would really prefer another 7 hours.
Regarding dark subtraction, you have to match the dark frame temps to the light frame temps. That can be a major PITA, so I stopped bothering and now use dithering instead. Along with Winsorized Sigma Clipping integration, that takes care of the hot and cold pixels, sat tracks, etc. I still use biases and flats, though...and flats actually tend to increase noise a bit as it removes LP and vignetting.
Thank you for the answers, well i agree with you, but as you said you are stretching the data beyond what anyone would do in other types of photography, maybe ask Canon for an astro camera? but yeah i want a better sensor as well, though currently i am very happy with the jump to FF
Upvote
0
dcm said:Click said:Jon, I have a question for you.
What are those triangles on both sides of some of the stars?
Interesting, they seem to be concentric with the center of the image
yeah these....stars?
Upvote
0
It's some kind of diffraction effect caused by the lens. Those are indeed stars. I'm not sure what causes it, but it's always been that way. I've noticed recently that my stars are not flat across the field, which has me worried that my lens has a decentered or tilted element. Frustrating, as if that's the case, it's probably going to cost me a fortune to get it fixed...and I don't know when or how it happened.
Anyway, here is an updated version:
http://jonrista.com/2014/12/28/orions-sword-wide-field-dust-and-reflections/
Anyway, here is an updated version:
http://jonrista.com/2014/12/28/orions-sword-wide-field-dust-and-reflections/
Upvote
0
meywd said:Thank you for the answers, well i agree with you, but as you said you are stretching the data beyond what anyone would do in other types of photography, maybe ask Canon for an astro camera? but yeah i want a better sensor as well, though currently i am very happy with the jump to FF
Eh, no point in asking Canon for an astro camera. They are the only DSLR manufacturer that has produced any, the 20Da and the 60Da. They don't filter out as much red, so they pass a lot more hydrogen alpha emissions (the primary reddish/pink emission of hydrogen gas nebula.) The problem with DSLRs is they have rather non-linear data. There is always some kind of in-camera processing. Canon used to be more linear than all the alternatives in the past, but they have been doing more processing in-camera lately, particularly with the 7D II (and DIGIC 6).
DSLRs make ok stop-gap astro imagers, but for anyone who is as serious as I am, the only real option is a CCD. CCDs have excellent data linearity and for the better manufacturers (i.e. QSI, FLI) they have perfect gaussian read noise (no banding, very few if any hot pixels), which is vastly superior for astro. The other benefit of a CCD is you can get them monochrome, and use various filters like LRGB (luminance + RGB) or narrow band (Ha, SII, OIII, NII, Hb, and a variety of other bands). The mono sensors have a far higher fill factor, no color noise to speak of, and are overall much more sensitive regardless of the filter used (simply because your using all of the sensors pixels for all bands.)
I'll probably be getting a QSI 683WSG-8 soonish here. It's an APS-C sized sensor, the KAF-8300, full mono, with an 8-position filter wheel (LRGB, Ha, SII, OIII and unfiltered), off axis guider port, and has a perfect gaussian read noise distribution. It's a very expensive camera though (with the filters and the various necessary adapters for use with my Canon lenses and standard telescope equipment)...about five grand. So I won't be buying any other cameras any time soon...least of all Canon, Canon sensors, even their newest ones, are just too darn noisy with poor noise characteristics.
Upvote
0
Thank you very much Jon and sorry to hear about the lens, The local astro club guys told me the same regarding CCDs, but yeah its expensive, i have no doubt you will put it to good use, btw i like the updated one more.
Upvote
0
Very nice. I am always amazed at the clarity of you images. I find it hard enough to get any clear nights with light pollution I can't imagine needing 2 or 3 times as many without . Orion is in the perfect location for me right now, but we only have had one clear night in the last month and that was Christmas eve.
Upvote
0
niteclicks said:
You need many times more clear nights imaging the same target WITH light pollution. You can get away with imaging for much less time WITHOUT light pollution.
Why? Because LP is additive to the object signal. If I'm imaging from my back yard, red zone light pollution, for every single object photon I get from a nebula, I get anywhere from 10 to 100 photons from light pollution (depends on transparency). To get 10 object photons, I have to collect 1000 light pollution photons. For 20 object photons, I've collected 2000 light pollution photons. Light pollution is something we ultimately subtract from our images...we correct a bit of it with flats, then we offset the rest. That leaves behind whatever object signal that was collected. If I offset by 2000, my object signal would still be 20. That is VERY low.
So, if my goal is to expose to about 1500-2000 e- at ISO 800, with a gain of say 0.6e-/ADU, then my image level in 16-bit integer would be something like 2000-3333 ADU. After correcting with flats and offsetting for the light pollution level, I end up with an object signal of 15-33. I could expose for longer, but because LP is compounding in the signal so much faster than the object, even if I doubled my exposure again, 4000 photons or 6666 ADU, my object signal is still only 66 ADU. Problem is, at ISO 800, 4000 e- is already getting close to the saturation point of an APS-C sensor, and is likely over half the saturation point of an FF sensor. That's too much exposure, your stars are guaranteed to be heavily clipped. Light pollution thus limits your object exposure depth by making your images "skyfog limited" very quickly.
Now, contrast this with a dark site. For every object photon gathered, I might gather 0.5 or 0.3 photons for skyfog (from light pollution, or at a truly dark site with 21.5mg/sq" or darker, from airglow). Then for every 2 or 3 object photons I gather, I get one skyfog photon. I can expose for much longer, or maybe just expose at a higher ISO for a similar amount of time (not generally recommended unless your imaging something REALLY dim, like the dust in my Orion Sword image.) For 100 object photons, I'd have 33-50 skyfog photons. Maybe even less (if airglow limited, you might expect about 0.1 skyfog photons for each object photon, meaning you can expose your object signal ten times stronger than skyfog.) Therefor, at a dark site, you are freed to expose for much longer, potentially as long as you want, before you become "skyfog limited"...which may be 10 to 1000 times longer or more than when imaging under light polluted skies.
If your goal is to expose to 2000e-, in a backyard with a ratio of 100:1 skyfog vs. a dark site with a ratio of 1:5 skyfog, you would have to expose 80 times longer in your backyard to gather the same number of object photons as you would at a dark site. Your gathering 20 photons per sub in the backyard, but 1600 photons per sub at the dark site. So, 1600/20, or 80x. If your goal is to expose 5 hours worth of total object integration at ~1600 photons per pixel per sub, your subs are 4 minutes (240 seconds) long, then your gathering 6-7 object photons per pixel per second at a dark site. You'll gather a total of 120,600 photons per pixel over five hours. Conversely, to get the same number of object photons from your back yard, you would be gathering 0.083 photons per pixel per second. To gather the same 120,600 photons...you would need to image for a whopping 403 hours under light polluted skies!!! Assuming you take the same 240 second subs, your still gathering 2000e-, however most of that signal is skyfog. So you would need over 6050 subs to get the same signal from your backyard as at a dark site.
Given that, on average, there are only 7 hours of total dark each night (a little more during winter, a little less during summer), it would take you 58 days to gather enough subs from a light polluted site. Assuming you started imaging an object as soon as it started rising high enough over horizon haze in the east, and continued until it was finally setting into the horizon haze in the west, you MIGHT have enough time in a single season to get enough subs to produce the same kind of signal from a light polluted yard as at a dark site. ;P (And, note, in reality, most places have a handful of clear nights a month to image, between cloudcover and the moon, so in a three-four month period of time you MIGHT get 15-20 clear nights tops...to get 6000+ subs on a single target, you would actually need YEARS to get all 57 full nights worth of imaging done.)
If your really interested in astrophotography, I highly recommend finding a close dark site. You might be surprised to find that, pointing either east or west, you have one closer than you think. There are a lot of light pollution maps on the web, but most are based on the bortle scale, which is a tool for visually gauging how dark your skies are, which involves how well you can see city light bubbles on the horizons. A "true or exceptionally dark" site on the bortle scale requires that LP bubbles NOT be visible on the horizons. That isn't necessarily required for imaging, you can often find skies with an SQM reading (sky quality) of >21 within 30-60 minutes from the middle of a downtown city area. Just image in the part of the sky that is dark, and in a few hours you might be able to gather enough "dark site" data that would be comparable to imaging for months from a back yard. You can use this site, which measures direct light levels from overhead, to find potential sites that might be dark enough to image from (anything blue or darker is good enough):
http://www.lightpollutionmap.info/#zoom=4&lat=4838950.03614&lon=-9735847.10785&layers=0BTFFFTT
The other option is to use a monochrome camera and narrow band (NB) filtration. Mono NB imaging blocks out 99% or more of the light, passing only very narrow bands of emission. You can image under LP, even when the moon is half or larger. The kicker here is you need a mono camera with a filter wheel and nice, very narrow band filters. A decent mono CCD is going to cost a couple grand, a filter wheel another grand, and the filters themselves could be anywhere from $400 to $1200 a piece. Expensive...but, if you don't have the ability to drive to a dark site all the time, it is the best alternative. Atik has some decent cameras and a nice filter wheel for pretty good prices, and Baader has some narrow band filters (little wider bands, but still narrow enough to be useful) for good prices. You might be able to get away with a mono CCD camera setup for about $3500 or so.
Upvote
0
Here is another one. My last image of 2014. Same equipment, same dark site, as the Orion's Sword image:
Managed to capture a number of galaxies in this one as well, maybe around a dozen or so. I highlighted the largest here:
Managed to capture a number of galaxies in this one as well, maybe around a dozen or so. I highlighted the largest here:
Upvote
0
Wow jrista, those shots are incredible. Unbelievable detail.
I've been making a few attempts at astrophotography lately. Unfortunately I live in a "white" zone for light pollution (right next to a large city). Here is a shot of Orion I took a few days ago. The cumulative exposure time was about 20 minutes, and was stacked in Deep Sky Stacker from forty 30sec exposures. No darks, flats, or bias frames. Taken with a 7D Mark II and EF 100-400L II.
To be honest I am very happy with the result, considering the light pollution in my area. Aside from shooting from darker locations, do you have any recommendations for improving the image, or dealing with light pollution in general? Will simply increasing the number of shots help?
I've been making a few attempts at astrophotography lately. Unfortunately I live in a "white" zone for light pollution (right next to a large city). Here is a shot of Orion I took a few days ago. The cumulative exposure time was about 20 minutes, and was stacked in Deep Sky Stacker from forty 30sec exposures. No darks, flats, or bias frames. Taken with a 7D Mark II and EF 100-400L II.
To be honest I am very happy with the result, considering the light pollution in my area. Aside from shooting from darker locations, do you have any recommendations for improving the image, or dealing with light pollution in general? Will simply increasing the number of shots help?
Attachments
Upvote
0
Hi Jon,
Just now looking back through this thread, nice work! I have one question for you. You have really nice red in the North American and Pelican nebulas and in some of the stars in the double cluster. I think I read that you're using Deep Sky Stacker to stack images, is that right? I took a few one minute subs of the double cluster a few nights ago with a 70D, and the individual frames show red stars but when I stack in DSS the red mostly disappears, and I can't seem to get it back when processing in Photoshop. I have had similar problems with the North American and Pelican nebulas. I know that the 70D cuts out a lot of the H alpha, but as I said I'm losing red in stars between the individual frames and the stacked result, so that can't be the explanation. And even with an unmodded 5D you're getting good red in the nebulas. What settings are you using in DSS (if you're still using DSS).
Thanks,
Glenn
Just now looking back through this thread, nice work! I have one question for you. You have really nice red in the North American and Pelican nebulas and in some of the stars in the double cluster. I think I read that you're using Deep Sky Stacker to stack images, is that right? I took a few one minute subs of the double cluster a few nights ago with a 70D, and the individual frames show red stars but when I stack in DSS the red mostly disappears, and I can't seem to get it back when processing in Photoshop. I have had similar problems with the North American and Pelican nebulas. I know that the 70D cuts out a lot of the H alpha, but as I said I'm losing red in stars between the individual frames and the stacked result, so that can't be the explanation. And even with an unmodded 5D you're getting good red in the nebulas. What settings are you using in DSS (if you're still using DSS).
Thanks,
Glenn
Upvote
0
jrista said:
I purchased the 583WSG with the 5 position filter wheel a couple of years ago. The QSI cameras are incredibly well made - there was a lot of nice design work that went in, unlike some other CCD cameras that feel like the mechanical design was an afterthought. I wish that the 8-filter wheel was available when I purchased mine, but I'm interested mostly in galaxies anyway, so the 5-filter wheel is fine for now. I purchased it when I had the funds but little time, so I've used it much less than I would have hoped. But I'm planning on building a ROR observatory in the spring and now have more time to dedicate to this hobby, so I hope to start getting my money's worth out of the camera at long last.
I did a little AP using the 20D before I got the QSI camera. Not having to combine the LRGB data was convenient, but there is really no comparison between the DSLR images and the CCD images.
And FWIW, I just tried the free PixInsight demo with my most recent image. It is really difficult to get my head around it, but the results are better than I could get with Nebulosity + Photoshop. But, I am a novice at it either way.
Dave
Upvote
0
jrista said:Here is another one. My last image of 2014. Same equipment, same dark site, as the Orion's Sword image:
Managed to capture a number of galaxies in this one as well, maybe around a dozen or so. I highlighted the largest here:
No artifacts in the stars this time. Lens okay?
Upvote
0
Flyingskiguy said:
Very nice! For heavily light polluted skies, without any LP filter, and without any calibration at all, that is VERY well done! Once you calibrate with bias, darks, and flats, you should be able to get a lot more detail out of that.
As for improving your images. You have two options. Find a dark site that's close enough to get to on a regular basis, or invest in an LP filter. If you are using an APS-C, you could get either the Astronomik CLS clip-in or the IDAS clip-in. If you are using an FF, then I do NOT recommend getting the Astronomik CLS-XL clip-in for Canon FF dslrs...the design of that filter is flawed, and it causes more problems than it's worth. You should get the IDAS LPS-V4 screw-on filter instead, which can be threaded 48mm for use with standard 2" T-adapters that attach DSLRs to telescopes, or if you are using one of Canon's great white lenses, you can get it in 52mm threaded for use with the drop in filter holder.
Personally, I say find a dark site. Use this map to find one:
http://www.lightpollutionmap.info/#zoom=4&lat=6156149.93155&lon=-11229480.6837&layers=0BTFFFTT
Upvote
0
gruhl28 said:
I think you would be surprised how much Ha you can get with modern DSLRs. All of my cameras are UNmodded...look at the Ha I'm picking up.
I get a lot. Personally, I like the amount I get...any more, an I think the Horsehead image would have been overpowered by Ha, which would have hidden the very faint blue reflection that's all over the place. (I think I have more blue reflection in my image than any other image of Horsehead I've ever seen...which I'm pretty happy about. 
) I used to use DSS to integrate. Now I use PixInsight. I've learned some things about Canon RAW files lately. I've NEVER liked how Lightroom demosaiced my CR2 files, not for years. It always results in pretty blotchy data, lots of red color noise, soft detail in the shadows, etc. DSS has always done the same. I've learned that LR/ACR, DSS, and a good number of other RAW editors use the AHD form of demosaicing, Adaptive Homogeneity-Directed demosaicing. I don't think AHD is ideal for Canon data.
I recently trialed Capture One 8, an alternative RAW editor to LR. I had a 60 day trial, and throughout that time, I definitely felt as though it was doing a better job with Canon CR2 files. It resulted in a finer grained noise, much lower color noise characteristic, less blotching, etc. Overall, I was pretty impressed with COne 8, and if anyone is looking to maximize their daytime photography quality, you should look into it. I let my trial lapse, and I've stuck with LR for now...simply because COne is a very different program, and has a tenth the camera compatibility, not to mention the fact that it does not work seamlessly with PS. I'm hoping that someday soon, Adobe may switch to the algorithm COne 8 uses.
Which brings me to PixInsight. PixInsight is an astrophotography editing tool, specially designed with tools to make the most out of ultra low SNR astro data. Unlike ACR/LR and most other RAW editors, it uses VNG (variable number of gradients) demosaicing. The VNG demosaicing of PI is very similar to the demosaicing of COne 8. I suspect they are both based on the same core VNG algorithm, with different tweaks to optimize results for their intended use case. Anyway, VNG results in VASTLY superior results from Canon CR2 files. I don't now why, but the noise is clean, usually an order of magnitude lower STDev, very low color noise, practically no color blotch. I use the PixInsight BatchPreprocessing script to calibrate and integrate my images now, as it just does a better job. Not just with the demosaicing either, it is far superior at star registration and rejecting out-of-sigma pixel data (hot pixels, aircraft/satellite/asteroid trails, etc.)
Anyway...PixInsight is not free. If you can afford the couple hundred bucks or so for it, I say go for it. You'll want it anyway if you really get into astrophotography...it's an essential tool, and WELL worth the money. If you use the BPP script, you shouldn't have too many problems with getting good results.
If you want to stick with DSS for now, then my advice is, register, calibrate, integrate...then just save off the data to a 16-bit TIFF file. Do NOT apply adjustments (checkbox in the save dialog), and do all of your processing in Photoshop. DSS does not render the integrated data well, and your just seeing it's poor rendering. Search the web for video tutorials on how to stretch your integrations with Photoshop, and go from there.
Regarding getting the deep red. A lot of my images from summer suffered from wicked-high dark current noise. A lot of the red in those images is false...blotchy color noise from Canon's heavily red-weighted color read noise. There is some Ha data in my images (less now during winter, with temps that hover around 0°C and ultra low dark current), but not a ton. What Ha data I have, I attribute to the power of PixInsight in helping me bring out. You might be able to bring out Ha data with Photoshop, but if you really want to get the most out of your images, PixInsight is a very worthwhile investment. It's a perpetual license, so once you buy it, you get updates until the next major version (which, at the current rate, is probably not going to be for another five to ten years at least.
)
Upvote
0
dcm said:
Oh, there are artifacts. I just stopped down to f/4.5, which creates the starburst diffraction effect, which hides the wedges. Look closely, though...and you can still see the wedge. My lens element is definitely tilted, which you can tell from the heavily elongated stars in the upper left (see the astrobin version.)
I like the starburst effect, so I usually stop down to f/4.5. I just forgot to with the Orion Nebula image.
Upvote
0
Tested out the sky-watcher star adventurer last night. Worked rather well with my 7d2 and 300mm f2.8l II. However needed a larger counterweight from an old Orion p.o.s. Mount...about the only good thing to come from that.
Seemed to track unguided pretty accurately and was able to take 30 sixty second subs of lovejoy at ISO 1600 without any drift errors. The moon had more of an impact than the mount.
Today I attempted to mount my 5d3 and 600mm f4 to it but with both counterweights it was still a bit top heavy. However I was able to lock the clutch and it tracked without stalling the motor. Took about 15 seconds to stabilize so I will likely have to use a manual cardboard shutter for the subs. I'm sure this combo exceeds the useful weight limit but it will be worth trying once the moon gets out of the way.
I have a trip planned next year to the Midwest and have been looking for something to take that will not require an extra suitcase. The skywatcher is small enough And takes about as much space as a couple of dslrs , excluding the tripod which has to go anyway.
I stacked the subs but having some issues with my laptop so the final will have to wait until that get resolved.
Here is one sub which I edited on my iPad.
Seemed to track unguided pretty accurately and was able to take 30 sixty second subs of lovejoy at ISO 1600 without any drift errors. The moon had more of an impact than the mount.
Today I attempted to mount my 5d3 and 600mm f4 to it but with both counterweights it was still a bit top heavy. However I was able to lock the clutch and it tracked without stalling the motor. Took about 15 seconds to stabilize so I will likely have to use a manual cardboard shutter for the subs. I'm sure this combo exceeds the useful weight limit but it will be worth trying once the moon gets out of the way.
I have a trip planned next year to the Midwest and have been looking for something to take that will not require an extra suitcase. The skywatcher is small enough And takes about as much space as a couple of dslrs , excluding the tripod which has to go anyway.
I stacked the subs but having some issues with my laptop so the final will have to wait until that get resolved.
Here is one sub which I edited on my iPad.
Attachments
Upvote
0