T-Shirt Trick Works and Learning More About Adobe Processing

It took a little while to perfect it, but the T-Shirt trick works for capturing flats. Unlike my previous attempts, I bought a clean, new white shirt. I gently stretched it over my scope and secured it with a rubber band. I made sure to maintain the same focus I used for my light frame exposures. And, the best trick of all, I switched my camera’s mode to “Av” on my Canon Eos Ra (an amazing Christmas gift from my brother!). The mode automatically picks the correct shutter speed to produce the proper exposure for the flat frame.

White T-shirt, somewhat stretched over my telescope and held in place with a rubber band. Makes all the difference.

As for my light source, I’ve been shooting in my backyard lately. So, once I finished taking my light exposures and my darks and I’m getting ready to pick up, I turn on my white LED porch light. From there, all I need to do is point my score at the light. It’s just as good as using an all-white LCD tablet or laptop screen or resorting to a light box. Except, it doesn’t do my a lot of good when I’m not shooting from my backyard. So, I’ll need to find a good portable light source when I’m working remotely.

Here’s the result of plugging my new flat frames into Deep Sky Stacker and doing some newbie image processing on them in Photoshop. It looked so nice, I had it printed and matted at FedEx Kinkos. Ultimately, you can see how that distorted ring of brightness from the last post is gone now, averaged out by DSS with the flats!

Orion Nebula, M42
A couple hours of exposure time with flats and darks.

The next step for me is to learn more about image processing in Adobe. Right off the bat, I can share just how great the “arcsinh” curve stretches are for performing color-preserving stretches of the TIF file data. The guy who designed these stretches for PixInsight, which is the super expensive astrophotography editing software, created these stretches for Adobe and he explains it all in this post on the Cloudy Nights forum here. Definitely pick these up and install them if you’re using Photoshop for Astrophotography; they’re free.

Here’s a quick and dirty bit of imaging I did using the arcsinh stretches to draw out my images with only a couple hours of exposure time at my local observatory. Granted, the EOS Ra also helps a lot with image quality, too. And even still, there’s a lot that could be done to improve these images of M81 and M82, the Bode Galaxy and Cigar Galaxy, respectively.

M81 and M82, after a couple hours of exposure, some arcsinh stretching, and a really good DSLR camera.

I grabbed 4 hours of exposure time of the Whirpool Galaxy, M51, which was the deep sky target I cut my teeth on over last summer. Assuming I can find some time this weekend, I’m hoping to process that TIF file in adobe and try some more advanced techniques I learned, like creating star masks to preserve star color, tricks to remove light pollution, and (not so advanced) ways to “feather” my lasso’s deselections so I can play with the galaxy separate from the stars and vice versa.

DSLR Astrophotography – Vignetting and Flats

“Vignetting” remains a problem for my astrophotography and has become an even worse since using my new astrophotography camera, a Canon EOS Ra. The pricey camera was a surprise Christmas gift and it works really well, so much so it feels like I just got my learner’s permit and someone handed me the keys to a Lamborghini. But it’s so much more sensitive to light with a larger sensor that it’s almost unusuable in my light polluted sky and backyard.

Reduced, JPEG version of my original stacked TIF file of the Orion Nebula, M42. You can make out the ring vignetting and uneven light gradient interfering with my processing.

But what is vigentting? My understanding is inexpert, but I would describe it as the background light in my photographs not being exposed evenly across the frame. So, my corners will be really dark, the middle of my image typically has a brighter rings, and then the middle is the brightest part. This disformity or unevenness in the light gradient makes balancing the background light and stretching the image cleanly almost impossible (at least for me…). As you try to darken the background, especially the brightest parts, you lose the image or you have this blob of overexposed noise.

There are a variety of ways to remove or reduce the vignetting. One tool I attempted to use after hearing it recommended by AstroBackyard was GradientXTerminator. You can download the tool to try for free for 45 days, but past that it costs money. I think it’s a great tool and would be really useful going forward, but I had limited success with my current batch of stacked in images. Here’s what the TIF image above looks like when GradientXTerminator is used on a medium/aggressive setting.

The image of the TIF file above run through a round of GradientXterminator. It’s a great tool, but my current batch of images are a little beyond its capabilities or my experience.

There are also more manual methods that can be performed in Adobe Photoshop that create fuzzy screening layers to even out the background noise or rely on using the radial gradient tool. I didn’t have much luck with either.

Really what I should have done is take some flat frames. These frames are used by stacking software like Deep Sky Stacker to balance out the background light to remove the vignetting. These images should consists of an uneven frame of light, not too bright, across the entire censor so the exposure artifacts and anomalies like dust can be substracted from your final image.

I used to think the flats were just unhelpful and not that necessary. Whenever I did take them, I either didn’t notice a difference or they made the final TIF file worse. But that’s largely because I was taking them incorrectly.

Unless you have a light panel or light box to stick onto the front of your telescope, you’ll be advised to use the “T-Shirt Method”. This method entails stretching a cotton white T-shirt over the end of your telescope and then exposing the end to an even light source like either the dawn sky, a tuned-down all-white LCD screen or light panel, or whatever light source you can at hand that’s both not too bright and also uniform. The T-shirt helps diffuse the light for a more even spread. Once you have the shirt attached and a light source, you snap some frames, like 20 or 30, at the same ISO setting and focus as the frames you took of your deep sky image, but generally with a much shorter exposure. As I understand it, non-technically, you want to return a grayish frame, not an all white one and certain DSLR cameras have an AV mode that will automatically adjust the exposure time for you to achieve this result.

With these flat frame, the frames of your image (light frames), and some dark frames (and potentially some bias frames but I’m not there, yet), you should be able to composite a much more balance TIF file from your stacking program.

That’s all theory to me for the moment, however. My first attempt at flam frames saw me using a rubber band to affix a few coffee filters onto the opening of my smaller refractor telescope and then pointing the scope at a 100-watt-equivalent LED bulb in a shop light. To make matters worse, I still relied on my intervaluemeter to take the snaps and it couldn’t take snaps of less than 1 second of exposure. So, my “flats” were blowing out my image to the point of uselessness.

See this bright white block. That was one of my first “flat frames”. It didn’t turn out so well.

The next not too cold and humid clear night I get, I’m going to give the flats another shot. I’ve actually bough some brand new T-shirts and figure I can still use the shop light with a lower wattage bulb and possible a second T-shirt over it. I’ll post about how well it goes eventually. Will it work or will I fail? Wait and find out.

In the meantime, here’s an image of my ongoing misadventures in learning to process these images. I guess I inadvertently activated the hidden Sailor Moon plugin.

Processing gone wrong. The “Sailor Moon” plugin.

Verified by MonsterInsights