Sunday, September 20, 2015


More Short Sub Imaging

Since I had pretty good results with the little ZWO ASI120MC on the deep sky the other night, I decided to give the minimalist camera a second outing, applying what I’d learned. If you weren’t here last time, what I’ve been doing is experimenting with a current craze in astro-imaging, taking many short sub(frames) with a CCD camera or a DSLR and stacking them into a final image. Keeping your exposures to 30-seconds or considerably less means you will likely not have to guide and will still obtain a respectable image when you are done.

The VX and the Edge 800, Ms. Emma Peel were still set up in the backyard, so all I had to do with remove the Desert Storm Cover, wake the mount from its hibernation, and soldier on. If you didn’t read last week’s entry, the  camera, the ASO120MC, is a one-shot color job with a small 1/3rd inch CMOS imaging chip. I had high hopes for the second round since I’d had such surprisingly good luck the first time.

After checking focus on M13, I decided I might as well do another sequence of everybody’s  favorite (everybody north of the equator, anyhow) globular star cluster. As before, I used FireCapture as my image acquisition/camera control software, and also as before I saved my results (or, more properly had FireCapture automatically save the frames) as .bmps that were converted to color on the fly before they hit my screen.

When I was ready to begin, I set my exposure. Initially, I tried 5-seconds, which seems to be what a lot of the short-subbers are using. I Had FireCapture take an image and subtract a dark frame, and had a careful look when it was done. Hmmm...not so hot...

The globular was there, and stacking one-hundred or more 5-second frames would probably have yielded a recognizable final product, but I thought the image was just too thin and too close to the background’s brightness level in my light polluted backyard. Yes, stacking many frames reduces noise and allows you to more easily process the resulting image, boosting its levels, etc., than you could with just a single frame, but stacking images does not increase detail. The detail must be there in a single exposure. If it’s not, stacking will not create it.

M13 with XTi...
So, I compromised. 15-seconds seemed a good exposure length to me. Short enough to keep the  stars round without auto-guiding, but long enough to deliver detail aplenty in brighter objects. I did go up to 25-seconds for my dimmest targets, Stephan’s Quintet and the Crescent Nebula, NGC 6888, however, since in a 15-second exposure both were just barely, barely visible.

I do believe 5-seconds could work from a dark site. But my backyard? Just no. My Zenith Limiting Magnitude is usually around 5.0, and it’s a struggle to pull some objects out of the background glow. Also, since, most equatorially aligned scopes can do 15 – 20-seconds without guiding, I don’t see much point to 5-seconds unless you are trying to image with an alt-az mount. The ground truth is that longer subs are just better all things being equal.

How many of those 15-second frames did I accumulate for each object? 50 was my average number. Didn’t take long to do a sequence of that length, and it also gave me enough of a stack to keep the background of my finished images looking reasonably smooth. Remember, this is an un-cooled camera, and my nighttime temperatures are still in the mid-70s.

I kept going a good long time, doing more objects than I thought I would. It was quite pleasant outside. Yes, it was still in the mid-70s before midnight, but it was also relatively dry and I was quite comfortable on the deck at the laptop. I did pop inside a couple of times to look at the Arrow episode the CW was showing, but unlike on my previous run, the bugs didn’t chase me indoors.

When I was done, I’d re-shot M13 and M92 and also done M2, M27, NGC 6888, the Blinking Planetary, the Cat’s Eye Planetary, Delphinus’ little NGC 6934, and Stephan’s Quintet. A good variety of objects to help me assess what the little camera could really do. When I was finished, I shut down the scope and covered her with the Desert Storm again, since I fancied I might, amazingly, get a third night of imaging or even a fourth judging by the weather forecasts I was hearing. That kind of streak has been awfully rare over the last year.

Next morning, I fired up Nebulosity 3 and stacked and processed my frames. Not only does Nebulosity make processing images easy, it includes what I believe to be the best and easiest to use deep sky stacking routine in the business. After my subframes were stacked, I gave each image the Digital Development treatment in Neb, tweaked levels a little, sometimes did a little sharpening, and, finally, exported the fits images to Photoshop and did a small amount of tweaking. It was all relatively easy.

M71 with the ZWO...
With the exception of the Crescent Nebula, which was just too thin, I had plenty of signal in all the images, and was able to deal with background noise easily enough in most of the shots. I was quite pleased. No, NGC 6888 and Stephan’s Quintet were most assuredly nothing to write home about, but they were there, and, as I reminded myself, this was from a light polluted site, my backyard, which is only a couple of miles from Airport Boulevard with its endless strip malls and car lots. It was thrilling to get anything that dim with a cheap camera on an unguided telescope.

I’ve been praising the ZWO, and that praise is well deserved. I don’t see how you could do better for the miniscule price. What should also receive praise, however, is the VX mount. I didn’t throw out a single frame—well, one or two satellite trail plagued ones. Even at 30-seconds the mount’s behavior was exemplary. Yes, I had the C8 reduced to f/3.3, but even then, the “magnification factor” inherent in such a small chip—M13 is almost too big for the frame—means that tracking has to be darned good. It was.

So, the ZWO had acquitted itself well as a deep sky camera fully capable of fulfilling the wants and needs of those of us who don’t aspire to being the next Bob Gendler. Those who, like your correspondent, are happy with simple deep sky postcards, mementoes of our cosmic journeys to show off to our wives and girlfriends and mates. If that is all you want, this camera and a 600mm or shorter focal length telescope is all you need.

Which got me to thinking…how could Joe or Jane Cashstrapped-amateur kick it up a notch and maybe spend even less? One way might be with a DSLR, a really cheap used DSLR. I still have my old (unmodded) Rebel XTi, which I use as a backup (terrestrial) camera body. It’s been far outpaced even by the Canon Rebels in the last five years, but I thought that it might be an improvement over the ZWO in some ways.

ZWO Stephan's Quintet:  at least they are there...
One big way is its much larger imaging chip, a 10-megapixel APS-C senor that produces large 3888x2592 images as compared to the ZWO’s 1280x960 pictures. More importantly perhaps, the Canon’s individual pixels are larger, 5.7µm in size compared to the ASI120’s 3.75µm. Usually, larger pixels result in more sensitivity and less noise.

Before I could try the old Rebel on Emma, I had to round up a couple of items. First was the long exposure shutter module, the DSUSB from Shoestring Astronomy. Unlike more recent Canons, you cannot enable long exposure with the USB connection to the camera. You have to have a DSUSB or a remote shutter release connected to the camera.  

I also needed my good, old SCT prime focus adapter. Which wasn’t in the big tackle box I use to store my small astro-gadgets. In fact it wasn’t anywhere. I probably had not used it since last January when I was shooting the comet, since we haven’t had much prime-focus-DSLR-worthy weather since. I am pretty maniacal about putting stuff back in its place, so it not being in the tackle box was a bad sign. Indeed, after much ill-tempered searching I had to admit it was just not anywhere. Well, maybe it was somewhere, buried in the tall grass at the club dark site. 

What to do? I’ve got a couple of old Newtonian/refractor prime focus adapters, but they would put the camera way too far back from the Edge f/7 reducer. What would I do? What would I do? Still looking for the adapter, I opened a cabinet in the shop and spied my ancient Meade off-axis-guider. Heck, that would work. I’d plug up the eyepiece port and just use it as a prime focus adapter. Threaded it onto Emma’s reducer, mounted the camera, and I was ready to roll.

Since I’d be shooting with the Canon, the image acquisition program would as usual be Craig Stark’s fantastic Nebulosity 3. All I had to do was select “Canon,” enable the DSUSB option, set an exposure and have at it. I thought that to be fair in comparing Canon and ZWO I probably should have given the Canon shots a little more exposure than 15-seconds, since the Xti’s ISO tops out at 1600—which I thought might make it less sensitive than the ZWO. A look at the first M13 to appear on my screen said otherwise.

Not that the ZWO isn’t a sensitive little camera; it’s just that the Canon is more than sensitive enough. More importantly, perhaps, the backgrounds in the images were obviously smoother and more noise free than those of the ZWO. No, the Canon isn’t cooled either, but its makers had gone to great lengths to keep noise down even in long exposures and it shows. It also helped that the camera subtracted a new dark frame from each image automatically. If M27 looked a hair dimmer on screen than with the ZWO, and I wasn’t sure it did, I could tell it would be easier to pull out of the background glow during processing.

Canon M27 and ZWO M27 (inset)...
One other good thing inherent in the Canon? A bigger chip meant more field. While every object I’d requested had been in the frame of the ZWO thanks to the VX mount’s outstanding goto, some were on the edge. Everything was well within the bounds of the frame of the Canon and most of the objects didn’t even need centering.

This is the time to give a shout-out to my mount control program, my planetarium program, the ever popular TheSky. In my case, that's TheSky6 Professional. Yeah, I know TheSkyX is even better, but I am cheap and I am used to 6 and it does everything I need. One odd thing? I have it hooked to the VX using an ASCOM driver. That being the case, there is no way it should be any more accurate in gotos than any other ASCOM capable astro-soft. BUT… For some reason, with my setup gotos using TheSky are better. Go figure. Anyhow, I love the program, and the combination of it and Knightware’s Deep Sky Planner is a powerful one.

Since for comparison purposes I would be shooting the same objects on this night that I’d shot on the previous one, I didn’t spend a lot of time outside. Once it was clear the scope, mount, camera, and computer intended to behave, I’d start a sequence of 50 15-second frames, and immediately retire to the den for the 25-minutes (including time for the darks) it would take to complete that. There were no surprises. I did M13, M92, M27, M2, and a couple of others before calling it a night.

What were my findings the next morning when I processed the images using the same software and procedures I’d used on those I'd got with the ZWO? The ZWO images were easy to process, but those from the Canon were easier. I could play around with levels and sharpening without having to be as careful not to pull out more background noise.

NGC 6934 with the ZWO...
Otherwise? On the globulars it was a tossup. The Canon shots looked good, but in the ZWO images, the globs nearly filled the field of the camera and just naturally looked very impressive. Yes, I could shoot at a higher focal ratio, f/10, with the C8 and make the globs bigger, but even in 15-seconds you can run into problems with trailed stars at f/10 if you don’t want to do a drift alignment, and I do not.

The Dumbbell Nebula was another story. While I didn’t have quite enough subs to make it look really smooth given my suburban skies, there was no doubt the nebula showed considerably more detail than in the ZWO shots. It’s a large object for that small chip, and this time the object being bigger hurt more than it helped.

The big takeaway for me, however? This observing run showed that a DSLR in the C8 at f/7 can do a good job on brighter objects, even in suburbia, and that guiding is not required. Neither is a drift alignment. My stars were round at 15-seconds, and probably would have been fine at 30, too. At 15-seconds I had to throw out exactly zero frames.

My success with the Canon got me to wondering how far I could push this. What would the C11 do? Its longer focal length would give larger pictures. But how about guiding? Would that heavy OTA require it at 15-seconds with my CGEM? I thought it would be fun to find out. In the months I’d had the CGEM I had yet to take on single prime focus shot, guided or unguided. You hear a lot about the mount’s capabilities or lack of them in that regard, so maybe it was time I found out.

There was more than just idle curiosity to discovering how the C11/CGEM would do for deep sky imaging. What is the biggest problem facing the backyard imager? Those darned light pollution gradients. Use an SCT with a focal reducer and it gets worse. The sky background isn’t just bright or uneven; you have a “porthole” effect. There’s light falloff toward the edge of the frame due to the reducing lens, the telecompressor. You expend most of your effort in processing trying to suppress that while bringing out as much detail in the object as possible.

What can help? Keeping exposures short helps with the gradients, but that makes background noise worse—TANSTAFFL as Bob Heinlein used to say. “There ain’t no such thing as a free lunch.” What else? Flats can also work if you’re not too lazy to take them. If you are, an excellent Photoshop plugin, Gradient Xterminator, can make your shots a lot more even. So can a little more focal length, which tends to spread out background skyglow. Course, there is the TANSTAFFL factor there, too, since your image scale increases with increasing focal length, which, as above, isn’t always a good thing. Still, I thought some more millimeters via the C11 might help.

The Crescent:  not bad considering the light pollution...
I had planned to disassemble the C8 after this night’s run anyway, and convinced myself it wouldn’t be much more work to not just pack up the Edge and the VX, but to replace them with the C11, Big Bertha, and her CGEM. Ha! A C11 ain’t a C14, but don’t kid yourself; it is a handful at close to 35-pounds. And so, of course, is the CGEM head (the normal tripod is almost identical to the VX tripod and is no heavier). By the time I was finished, I was soaked to the skin with sweat and wondering if the effort had been worth it.

I planned to reduce Bertha’s focal length via the good, old Jim Riffle designed f/6.3 reducer/corrector, which would take her from about 2800mm to 1764mm which is a darned sight better than nearly 3000, but still… My experience over the years has been that as soon as you pass 1500mm, getting good tracking becomes harder. And as you approach 2000mm it becomes absolutely tough. Even at short exposures in the 15 – 20 second range. Would the CGEM, which is most assuredly in the “bargain mount” category, be up to the task? Stop back by next week to find out. 

Another good post, I think I have spent a large portion of the previous couple of months reading all your blog archives!

This is especially relevant for me seeing as I recently upgraded to a C11/CGEM combination from a C6/CG5 (non go-to). I had a try at imaging with the C6 but found finding targets the most frustrating thing, especially after using a focus mask on a bright star - it was too difficult to find targets again in regions where there weren't too many nearby bright stars. Also polar alignment was tricky with only a polarscope and the lack of patience to drift align

Now that those issues should be easier to solve, I have been tempted to try imaging again. I haven't got a focal reducer just yet, but was thinking of getting an OAG to get round stars at f/10 without too much expense.

Looking forward to the next post!
This comment has been removed by the author.
Great write up! I have been using the ASI224, Smarteq Pro, Celestron C5 and both a SCT 6.3 and .5x reducer. I use it with Fire Capture and the live stacking AstroToaster for semi-live viewing from my ridiculously light polluted Manhattan rooftop(you commented on my Ring Nebula image on Cloudy Nights.) It was quite a learning curve! I love that I can see things that would otherwise be impossible from my location.
Your images have nice round stars to the edge, my images have some serious coma from all the reduction. Any tips for better results? Here is a link to my thread:
Your pictures are excellent. I'm not doing anything special...but the old Meade f/3.3 does a better job than some reducers...
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?

stats counter Website Hit Counters