Our Beautiful Universe - Photos and Videos

[collinsmark]

Here's an image I captured last Tuesday night [Edit: actually, Monday night/Tuesday morning], a little less than a week before opposition.

Rhea, one of Saturn's moons, can be seen on the left. Saturn's north is "up" in the image.

If the rings' shadow and inner rings look a little blue to you, it's less to with them actually being blue and more to do with the fact that I did not break out my atmospheric dispersion corrector (ADC) during acquisition, mostly out of laziness. Earth's atmosphere affects blue wavelengths more than other colors.

Equipment:
Celestron C14 EdgeHD telescope
Skywatcher EQ8-R Pro mount
TeleVue 2x Powermate (a fancy Barlow lens)
Astronimik RGB filter set
ZWO ASI290 (monochrome camera)

Software:
FireCapture (for acquisition)
AutoStakkert! (for lucky imaging processing)
WinJUPOS (for derotation, RGB combination)
PixInsight with RC Astro plugins (sharpening and misc. processing)

Acquisition (using FireCapture):
Location: San Diego, USA
Date/Time: 2025-09-16, from 08:33.9 to 08:54.6 UT
(Midpoint time: 2025-09-07 08:47.4 UT)
Atmospheric seeing: somewhat on the better side of "meh"
Sub-frame exposure time: 25 ms.
Acquisition video length: 60 sec per color, alternating R-G-B-R-G-B...
21 minutes of total acquisition time.

Lucky Imaging with AutoStakkert!:
Best 50% frames kept.
Drizzle/Resampling: 3x Drizzle

Sharpening (after lucky imaginging processing) by PixInsight:
MultiscaleLinearTransform
DynamicCrop (to remove stacking artifacts at frame edges)

Derotation:
For each color channel (R, G, or B) 7 sharpened images were derotated and stacked using WinJUPOS.

Final adjustment processes with PixInsight:
ChannelCombination to combine the R, G, and B into a single color image.
HistogramTransformation did the heavy lifting for color balance
CurvesTransformation for color balance fine adjustments
NoiseXTerminator
DynamicCrop for final crop

 

[Andy Resnick]

It's interesting you are reporting this, and I think I know what you mean, actually.

I'm not done trying to locate the blemish, hopefully off-axis illumination (an idea borrowed from 'ultramicroscopy') will reveal the problem. I'd be surprised if it was inside the lens, but I can imagine a few scenarios (condensation, for example).

I am not brave enough to take the lens apart :)

 

[Andy Resnick]

Have you tried flat frame calibration?

I'm a stickler for flats. Every deep sky image I've ever posted here I've processed with flat frame calibration as part of the process (planetary doesn't count; I don't do flats for planetary). I take new flats for every new target for each filter I use for that target. Also, if I ever do adjustments to the telescope's optical train (e.g., detach the camera from the telescope, adjust the collimation, etc.), I'll take new flats.

Flats... yeah, I've tried those. Unlike dark/bias/bad pixel frames, I spent at least a year (maybe even longer) trying to get flats to work with my telephoto lens and in the end just up. I'm not exactly sure of the underlying cause(s), but using flats always increased the amount of post-processing work I had to do:

1) Any dirt present on the flats and not in the lights results in not just incomplete correction of the lights, but the introduction of spurious bright spots.
2) The other image defect is hard to describe, but consisted of concentric rings of color, almost as though the colors of an archery target were superimposed over the image. I think it's related to the changing sky hue as it gets dark and also as the city light pollution commences.

Now I may try again with this lens, but not try too hard... :)

 

[collinsmark]

Flats... yeah, I've tried those. Unlike dark/bias/bad pixel frames, I spent at least a year (maybe even longer) trying to get flats to work with my telephoto lens and in the end just up. I'm not exactly sure of the underlying cause(s), but using flats always increased the amount of post-processing work I had to do:

1) Any dirt present on the flats and not in the lights results in not just incomplete correction of the lights, but the introduction of spurious bright spots.
2) The other image defect is hard to describe, but consisted of concentric rings of color, almost as though the colors of an archery target were superimposed over the image. I think it's related to the changing sky hue as it gets dark and also as the city light pollution commences.

Now I may try again with this lens, but not try too hard... :)

1) Yes, this is true. Any blemishes on the lens/optics must be consistent between lights and flats.

That's one reason why I take flats fairly frequently (about 1 set of flats for each deep sky target, and if I revisit a target after a significant amount of time has passed, that's a whole new set of flats). But yes, it adds another step in the post processing.

2) Hmm.

Not sure what caused that. But again, make sure that lens' aperture is identical between lights and flats. If the light source used for flats is too bright, and you try to compensate by stopping down the aperture, your flats are doomed. The aperture must be identical between lights and flats.

Also, when in post processing, make sure the flat compensation is multiplicative and not additive. For example, the master dark is subtracted from the lights (an inverse additive process). Don't do that with flats. The inverse of the calibrated flats (master flat) must be multiplied, not subtracted.

And it can be tricky in theory to set up and configure correctly. For example, bias frames can be used to calibrate lights, flats and darks. But if you've already [you haven't] calibrated your darks with bias frames, you want to be careful that you avoid calibrating your lights with bias frames too. Otherwise when calibrating your lights with the calibrated master dark frame you'd be double-doing (over correcting) your bias frame (i.e., read noise) calibration. Your software might automatically take this into consideration, but it has to be configured properly. (But it is okay to calibrate your flats with bias frames independently of all this.)

You mentioned you used APP (Astro Pixel Processor). I don't have any experience with that software. But I expect it can work with flats, as long as everything is set up and configured correctly.

Before I finish talking about flats, I can't imagine taking deep sky astrophotos without flats. Yes, it's an extra step or two, but one gets used to it. It's just a (relatively easy) part of my process now that I just expect and do without thinking much about it (because I'm so used to it).

---------------

When looking for the dust mote culprits, you can probably rule out the front objective of the lens. Getting dust and muck on the front of the lens can reduce contrast and light gathering ability, but it probably won't have a significant effect otherwise. So the front of the lens probably isn't the place to look.

The back of the lens -- the side of the lens that points toward the sensor -- is a different story. That could be the place to look.

As a general rule, the bigger the dust mote looks on the exposure, the farther away it is from the sensor (for a given aperture setting). Dust on the sensor itself will just look just like a distinct piece of dust. Dust farther away will look bigger. Dimmer and blurrier, but bigger. (In my case, dust on my camera's sensor window [a few millimeters from the sensor] will appear a little bigger, rounder and blurrier than dust directly on the sensor. Dust on a filter on the filter wheel will appear even bigger than that, because it's farther away. If dust was on the objective itself, it would be infinitly big [or in practicle terms, vastly bigger than the entire sensor], thus it wouldn't look like a dust mote; it would just reduce contrast.)

Yeah, let's hope it's not inside the lens assembly.

 

[collinsmark]

By an uncanny coincidence, Cuiv, the Lazy Geek released a video a few hours ago introducing calibration frame process flow.

And he brings up a good point that might explain (possibly, maybe) what may have been the culprit with @Andy Resnick's flat frame trouble. (See around 26:30 in the video.)

 

[Andy Resnick]

1) Yes, this is true. Any blemishes on the lens/optics must be consistent between lights and flats.

[...]

Yeah, let's hope it's not inside the lens assembly.

My test run using flats for 55mm imaging was a disaster. Regardless of the underlying reasons (possibly correctable reasons), my conclusion is that for my particular setups/imaging conditions, flats cause more problems than they correct. Just the flats, tho- darks and bias frames pose no problems for me.

I did spot the offending blemishes when looking off-axis; unfortunately they are (apparently) indeed within the lens and not on an external-facing surface. S'ok, tho...

 

[Andy Resnick]

By an uncanny coincidence, Cuiv, the Lazy Geek released a video a few hours ago introducing calibration frame process flow.

And he brings up a good point that might explain (possibly, maybe) what may have been the culprit with @Andy Resnick's flat frame trouble. (See around 26:30 in the video.)

I guess it's possible? It doesn't really change my conclusion that for me, using flats is a super-frustrating PITA that isn't worth my time.

 

[collinsmark]

I guess it's possible? It doesn't really change my conclusion that for me, using flats is a super-frustrating PITA that isn't worth my time.

For me it's pretty invaluable. There's a lot of "bang for the buck," once the initial setup/workflow is debugged and figured out. Again, flats aren't just about correcting for vignetting. It's the best way of virtually eliminating dust motes from your data (when possible; where dust motes are consistent between lights and flats).

 

[Andy Resnick]

Consecutive nights of a crescent moon:

The first night was a complete disaster- I am amazed I got anything at all. In the best shot, Dorsum Oppel is barely visible as a bright curve on the left running roughly parallel to the outer perimeter of Mare Crisium:

(Nikon D810 + Nikkor 800/5.6 @ f/8, 1/15s ISO 200)

 

[Ibix]

It's interesting how messy the terminator is. Some of those mountains must be pretty high, because there are a few bright bits quite deep into the night side.

 

[Andy Resnick]

Hot off the press- the International Space Station, as seen from Cleveland, OH (300%):

(Nikon D810 + nikkor 800/5.6 @ f/8, 1/2000s ISO 1000)

 

[Andy Resnick]

A cropped and downscaled image of the Veil nebula and NGC 6940, taken using a 105mm lens:

Deets: Nikon D810 + 105/1.4 (shot at f/2) on Losmandy GM-8, 30s subs, 8.75 h total integration time. Stacking etc. using APP.

As it happens, for whatever reason(s) flat frames worked really well with this lens. I honestly don't know why I struggle so much with flats- I first thought it was a f/# issue (my lenses are really fast compared to most telescopes) but since I shot this at f/2...? Whatevs...

The full frame shows some of the tricky parts about this lens:

The green hue along the left side is chromatic aberration due to a slight tilt of the image plane with respect to the CCD plane. It appears when I forget to set focus to a point displaced to left of center to compensate. This optical effect does not appear in 'normal' photographs, it appears only due the extreme tone mapping required of astrophotography.

Happily, the flats did indeed cancel out the dust. Funny story- partway through the week I needed to changes lenses, and when I did I spotted the dust particles. I *almost* gave in to the reflexive urge to wipe the lens clean, but I remembered the flats!

 

[DennisN]

105/1.4

That is seriously fast for a 105mm lens!
Methinks it ought to be great for e.g. flower shots (providing lots of background blur) from a bit of distance, and portraits, of course.

 

[collinsmark]

The Jellyfish Nebula (a.k.a. IC 443, Sh 2-248) scooped up from my back patio, Feb. - Mar. 2025. The nebula is a supernova remnant found in the constellation Gemini. It lies about 5000 light-years away, almost in the very opposite direction of our galaxy's center.

To me it looks more like a brain, but I concede it kinda' looks like a jellyfish too (sort of).

Equipment:
Explore Scientific 80ED-FCD100
Sky-Watcher EQ6-R Pro
Orion Field Flattener for Short Refractors
Off-axis guider (OAG) with guide camera
Baader 3.5/4nm Ultra-Narrowband filter set
ZWO ASI2600MM-Pro main camera

Software:
N.I.N.A.
PHD2 guiding
PixInsight with RC-Astro plugins

Acquisition/Integration:
Location: San Diego, USA
Bortle class 7 (maybe 8 ) skies
All subframes binned 1x1
Stacked using the drizzle algorithm
SHO mapping
SII: 96×480s = 12.80 hrs
Hα: 124×480s = 16.53 hrs
Oiii: 72×480s = 9.60 hrs
Total integration time: 38.93 hours.

 

[Andy Resnick]

That is seriously fast for a 105mm lens!
Methinks it ought to be great for e.g. flower shots (providing lots of background blur) from a bit of distance, and portraits, of course.

It is a fantastic lens for low-light level work, and vanishing depth-of-field images. Fully open, tho, the image is a little soft for astrophotography.

 

[Andy Resnick]

Tonight: Rabbit in the moon

 

[bruha]

Hello, I attach my last atempts, this saturday was clear moonless so I get out telescope after long time.
Images are made with phonecamera on eyepiece adapter-little gimp corrected.
trapezium in orion nebula and saturn (rings was pointed to earth so was seen just almost as line..)
Lot of succes

 

[Andy Resnick]

For me it's pretty invaluable. There's a lot of "bang for the buck," once the initial setup/workflow is debugged and figured out. Again, flats aren't just about correcting for vignetting. It's the best way of virtually eliminating dust motes from your data (when possible; where dust motes are consistent between lights and flats).

Just on the off chance you have some ideas... I have a good example of the underlying problem I am having with trying to use flats flats when imaging with my 400/2.8 lens. Below is a single-night stack, representing an average of 353 individual frames taken over a 3 hour period:

The problem should be obvious. To be sure, I maximized the saturation to clearly show the color gradations that occur over the field of view, and I have no explanation for why they are present. They are not present when using my other lenses (and using flats for those make a lot of sense), and what makes flats unusable is that there are night-to-night variations in "spatial hue", and this is why I get bulls-eye hue patterns when I try and use flat field correction.

Anyhow, like I said, just in case you (or anyone, I suppose) have some ideas! thanks in advance?

 

[collinsmark]

Just on the off chance you have some ideas...
[...]
Anyhow, like I said, just in case you (or anyone, I suppose) have some ideas! thanks in advance?

In addition to ideas in previous posts (e.g., make sure focus, rotation, etc. is consistent between lights and flats), also

• Make sure that the use of the lens' lens-hood/dew-shield is consistent between lights and flats. For example, if you use lens hood when you take your lights (which you should), don't remove it to take your flats.
• Make sure the device used to produce the flat frames (dedicated flat panel, tablet, artists' sketch pad thingy, etc.) is large enough to simulate the entire sky. (It should at least extend to the edges of your lens hood.)
• Consider placing diffusion material, such as tissue paper on the screen of the tablet (or whatever you use to generate the flat field). The screens of tablets, by themselves, can be very directional dependent. Their intensity flux can appear much brighter face-on then it does even tilted at a small angle. This effectively can cause vignetting in the flat frames themselves, ultimately causing overcompensation of in the light frame calibration process. Using diffusion material such as tissue paper should (hopefully) make the light coming off the tablet/panel more omnidirectional.

Those are some ideas I came up with. I almost included the idea of taking "sky-flats" but that's a "whole 'nother" can-of-worms.

 

[Andy Resnick]

In addition to ideas in previous posts (e.g., make sure focus, rotation, etc. is consistent between lights and flats), also

• Make sure that the use of the lens' lens-hood/dew-shield is consistent between lights and flats. For example, if you use lens hood when you take your lights (which you should), don't remove it to take your flats.
• Make sure the device used to produce the flat frames (dedicated flat panel, tablet, artists' sketch pad thingy, etc.) is large enough to simulate the entire sky. (It should at least extend to the edges of your lens hood.)
• Consider placing diffusion material, such as tissue paper on the screen of the tablet (or whatever you use to generate the flat field). The screens of tablets, by themselves, can be very directional dependent. Their intensity flux can appear much brighter face-on then it does even tilted at a small angle. This effectively can cause vignetting in the flat frames themselves, ultimately causing overcompensation of in the light frame calibration process. Using diffusion material such as tissue paper should (hopefully) make the light coming off the tablet/panel more omnidirectional.

Those are some ideas I came up with. I almost included the idea of taking "sky-flats" but that's a "whole 'nother" can-of-worms.

Thanks, but these don't address the problem that spatially varying hue causes. For my other lenses, I've had excellent results by placing the lens against a flat panel display monitor: there is no problem correcting for the spatial brightness.

Again, as you can see from the image I posed, the problem I have with that specific lens is the color variation within the field of view. It's not constant night-to-night and is not replicated with obtaining flats using a flat panel display. I do not understand the origin of this phenomenon, it is not a lens aberration but could be due to terrestrial lights.

 

[collinsmark]

I do not understand the origin of this phenomenon, it is not a lens aberration but could be due to terrestrial lights.

I don't know the origin either. At this point I can only speculate.

But I'm confident that it's possible to find the origin/root cause with proper experimentation, perhaps starting with very controlled conditions. Bound the problem.

For example, start with a batch of flats and stack them to make a master flat as you normally would. Now take that same batch of flats (duplicating if necessary) and pretend they are lights. Stack the "lights" (which are really flat frames) as you normally would. If all works out well, the result should be a nearly perfectly monotone image without any significant variation. (Be careful of using interpreting "autostretch" here, as it will exaggerate even the most minute variations/noise in the final image. My point is that it should be nearly variation free even with a pretty aggressive stretch.) If it's not, something is wrong with the process.

If that works out, try taking your "lights" of a blank wall and see if the bad-pattern/bad-color-variation reproduces itself.

Or try taking flats using a different light source and see what changes in the end.

And yes, it is quite possible (as far as I can tell at the moment) that the culprit might be glare from terrestrial sources (e.g. street lights/porch lights). Also, moonlight glare could be a contender. Different lenses handle glare differently. Some lenses are more prone to lens flare and internal reflections than others. I'm sure there are experiments to test the lens' flare performance. (One thing's for certain though, if you take your lights with the lens hood on, make sure you take your flats with the lens hood on too. You're almost guaranteed to get weird problems if you change things up between lights and flats.)

For what it's worth, I regularly block out the most egregious sources of glare with a black bed sheet that I string up every night that I get the telescopes out (this is for light frames, of course). And even then, my results are not completely flat after initial processing -- there are gradients left. The flat frame calibration helps a lot, but I still have to do some gradient removal in post processing.