13 replies to this topic

[Coconuts]

I have been meaning to try my hand at planetary imaging for a long time, and only just now got around to it, inspired by Jupiter's imminent opposition.  A few hardware and software details:

• The scope I am using is an Intes Micro (no longer made, Russian) "M715 deLuxe": 178 mm aperture f/15 Maksutov-Cassegrain, with a native focal length of 2,670 mm.  This scope uses a separate secondary ("Rumak"style), not an aluminized spot on the meniscus ("Gregory" style), which offers better correction of optical aberrations.  Field correction doesn't matter for field-centered planets, but spherochromatism does.
• The mount is an iOptron HAE29EC, on an iOptron 40 mm carbon fiber tripod.
• The camera is a Player One "Neptune" 664C, RGB, 2.9 um pixels, so 0.22 arc-seconds per pixel, critical sampling.  I was relieved to learn that the Neptune 664C would also work on Jupiter
• My tropical skies were fairly dark (20.87 mpsas), with no breeze, and good steady seeing.  The time was 3:16 am local time, when Jupiter was about 60 degrees above the horizon.  Had I awakened earlier, it would have been at the zenith.
• I used SharpCap for image acquisition.
• I used Astrosurface as the sole alignment, stacking, and image processing (sharpening, wavelets, and white balance) tool. 
• The attached images apply three differing levels of wavelet processing, and are from a single 6000 frame .ser video file, 68 seconds in duration, with 50% of the frames being used.
• The ROI was 900 x 900 pixels, with Jupiter centered and subtending 242 pixels.
• The exposure time was 10 ms, and the FPS was 88, with an analog gain of 378.  The histogram peak was in the green, at ~ 85%.  The SharpCap CameraSettings text file is attached.
• I was pleasantly surprised at the level of detail I was able to pull out of the image on this, my second night trying planetary imaging.  Two nights ago was windy, and a sequence of frustrations as I puzzled out all of the details.  Finder alignment is critical to get Jupiter in the field.  Last night, I added a 400 - 680 nm UV/IR cut filter, to avoid IR blurring.
• In Astrosurface, I applied three levels of wavelet processing.  It was hard to decide between pulling out the highest level of detail vs looking oversharpened and unreal.
• Feedback on my settings and results, and on which is the more appropriate level of wavelet processing, would be appreciated.

All the best,

 

Kevin

 

Low wavelet processing:

 

Medium wavelet processing:

 

High wavelet processing:

 

SharpCap CamerSettings text file:

 03_16_51.CameraSettings.txt   1.32KB   8 downloads

Edited by Coconuts, 03 December 2024 - 11:48 AM.

[Tapio]

You asked for feedback so I'll give you some.
But first must say that you have a nice scope.

Two things caught my eye: first it seems that for some reason you are overexposed (or processed levels or something), second you should check and adjust white balance.

But overall, and it was first,then excellent result.

[bsturges]

Yes, decrease the exposure and you'll have it. That is very sharp.

[RedLionNJ]

• The attached images apply three differing levels of wavelet processing, and are from a single 6000 frame .ser video file, 68 seconds in duration, with 50% of the frames being used.
• The ROI was 900 x 900 pixels, with Jupiter centered and subtending 242 pixels.
• The exposure time was 10 ms, and the FPS was 88, with an analog gain of 378.  The histogram peak was in the green, at ~ 85%.  The SharpCap CameraSettings text file is attached.
• Feedback on my settings and results, and on which is the more appropriate level of wavelet processing, would be appreciated.

 

I highlighted these four bullets in particular, primarily because the rest are just fine!  I'm not a seasoned SharpCap user, so can't justifiably comment on those aspects.

 

If the seeing/focus/collimation are all really good, then only the finest level of wavelets is necessary to "clear the mist" and bring out the finest detail you captured.

 

More frames during capture would be helpful. I'd be looking to stack at least 2500, so try to capture at least 10,000 over a 2-minute period.

 

That's a large ROI for a 240 pixel Jupiter. Does your mount let you make it any tighter? I'm not sure on the SharpCap terminology, but I'd be going for something like a 320 pixel vertical ROI and a 280 x 320 cutout during capture.

 

Exposure x FPS should equal 1000 - you're missing out on some light/capture time somewhere. The smaller ROI should help with that.

 

Green should never dominate Jupiter's capture histogram and you shouldn't capture at 85% - you're going to burn out the EZ during processing, which is exactly what you did.  Try a smaller gain and/or shorter exposure if you're going to sharpen that much. Don't "white-clip" the histogram during processing, at any cost.

 

As already mentioned, try either adjusting the capture color ratios, or more aggressively color-balance in processing. The SEB and NEB should be somewhere between a light tan and pale red.

 

You're off to a promising start. Where are you located? I was imaging Jupiter at 74 degrees last night, but the conditions were appalling

[Coconuts]

Tapio & bsturges:  Thanks for the feedback.

 

Per "it seems that for some reason you are overexposed (or processed levels or something)", when I was adjusting the analog gain, I kept an eye on the SharpCap histogram; nothing was clipping there.  I made sure to choose a value above 180, which is where HCG kicks in and slashes the read noise.  When I probe the final image's RGB values, I don't see any clipping.  Could my level of sharpening or wavelet processing simply be excessive?  Which of the three variants comes closest to being what would be considered a good image?

 

Per "you should check and adjust white balance", I did use Astrosurface's Auto WB correction.  Previously, the images was very yellow.  By the way, last night I added a 400 - 680 nm UV/IR cut filter, to avoid IR blurring.

 

Per "decrease the exposure and you'll have it", is my choice of 10 ms a bit long to freeze atmospheric seeing?  What exposure time would be right for an f/15 scope and this camera?  Should the initial image in live video be intentionally on the dim side?  I raised the analog gain until the image looked bright enough to my eyes.  If I went with a 5 ms exposure, should I increase gain further, or should the video image appear on the dim side for best results?

 

Thanks again.  And yes, I really like the Intes Micro M715 for planetary and lunar observing.  I plan to try lunar imaging soon.

 

All the best,

 

Kevin

[Coconuts]

RedLionNJ:  Comments below:

 

I highlighted these four bullets in particular, primarily because the rest are just fine!  I'm not a seasoned SharpCap user, so can't justifiably comment on those aspects.

Thanks.

 

If the seeing/focus/collimation are all really good, then only the finest level of wavelets is necessary to "clear the mist" and bring out the finest detail you captured.

So easy does it on wavelets.  Got it.  I plan to check collimation via star test this evening.

 

More frames during capture would be helpful. I'd be looking to stack at least 2500, so try to capture at least 10,000 over a 2-minute period.

I used 50%, so 3000 frames stacked.  The individual image quality plot was pretty flat for well past the middle 50%.  Isn't two minutes long enough to begin to have issues with rotation?

 

That's a large ROI for a 240 pixel Jupiter. Does your mount let you make it any tighter? I'm not sure on the SharpCap terminology, but I'd be going for something like a 320 pixel vertical ROI and a 280 x 320 cutout during capture.

Completely arbitrary choice, not mount limited; I'll tighten it up as advised.

 

Exposure x FPS should equal 1000 - you're missing out on some light/capture time somewhere. The smaller ROI should help with that.

Super helpful rule of thumb; thanks!  But 10 ms and 88 FPS is 880, close, right?  But maybe 5 or 3 ms, and 200 or 330 fps would be better, enabled by a smaller ROI.

 

Green should never dominate Jupiter's capture histogram and you shouldn't capture at 85% - you're going to burn out the EZ during processing, which is exactly what you did.  Try a smaller gain and/or shorter exposure if you're going to sharpen that much. Don't "white-clip" the histogram during processing, at any cost.

For some reason (twice as many pixels?) green was the highest histogram component.  Should I reduce gain on a per-color basis diring video capture to make them about the same?  Great advice that 85% is too high.  What % should I shoot for?  What is "EZ"?  I'm not sure if Astrosurface allows you to monitor the histogram during processing; I'll check.

 

As already mentioned, try either adjusting the capture color ratios, or more aggressively color-balance in processing. The SEB and NEB should be somewhere between a light tan and pale red.

Very helpful!

 

You're off to a promising start. Where are you located? I was imaging Jupiter at 74 degrees last night, but the conditions were appalling

USVI, 18.3 degrees latitude.  Long ago, I was a NJ kid.  Exit 36.

 

All the best,

 

Kevin

Edited by Coconuts, 03 December 2024 - 10:05 AM.

[RedLionNJ]

Further feedback in red (don't take red as being negative!):

 

RedLionNJ:  Comments below:

 

I highlighted these four bullets in particular, primarily because the rest are just fine!  I'm not a seasoned SharpCap user, so can't justifiably comment on those aspects.

Thanks.

 

If the seeing/focus/collimation are all really good, then only the finest level of wavelets is necessary to "clear the mist" and bring out the finest detail you captured.

So easy does it on wavelets.  Got it.  I plan to check collimation via star test this evening. Please make that an in-focus star test - see if you can get concentric diffraction rings around the Airy disk.

 

More frames during capture would be helpful. I'd be looking to stack at least 2500, so try to capture at least 10,000 over a 2-minute period.

I used 50%, so 3000 frames stacked.  The individual image quality plot was pretty flat for well past the middle 50%.  Isn't two minutes long enough to begin to have issues with rotation? Different people say different things. Two minutes is probably fine, maybe even 2m30s, at least if using AutoStakkert to do the alignment and stacking. It builds a reference frame from near the middle of the recording, then aligns the features to that, so it can compensate for probably somewhere in the 2-3 minute range. Don't forget that quality plot is relative to the BEST frame of the entire video. It's all relative. If you get one errant, hyper-sharp frame (or a frame the software THINKS is hyper-sharp), that can throw the whole estimation thing off. I think you get this, though.

 

That's a large ROI for a 240 pixel Jupiter. Does your mount let you make it any tighter? I'm not sure on the SharpCap terminology, but I'd be going for something like a 320 pixel vertical ROI and a 280 x 320 cutout during capture.

Completely arbitrary choice, not mount limited; I'll tighten it up as advised.

 

Exposure x FPS should equal 1000 - you're missing out on some light/capture time somewhere. The smaller ROI should help with that.

Super helpful rule of thumb; thanks!  But 10 ms and 88 FPS is 880, close, right?  But maybe 5 or 3 ms, and 200 or 330 fps would be better, enabled by a smaller ROI. We don't have to work with whole numbers

 

Green should never dominate Jupiter's capture histogram and you shouldn't capture at 85% - you're going to burn out the EZ during processing, which is exactly what you did.  Try a smaller gain and/or shorter exposure if you're going to sharpen that much. Don't "white-clip" the histogram during processing, at any cost.

For some reason (twice as many pixels?) green was the highest histogram component.  Should I reduce gain on a per-color basis diring video capture to make them about the same?  Great advice that 85% is too high.  What % should I shoot for?  What is "EZ"?  I'm not sure if Astrosurface allows you to monitor the histogram during processing; I'll check.   For Jupiter, it's not bad to make all three colors about the same level on the capture histogram. But that's not true for red-dominated Mars, for example. EZ refers to the Jovian Equatorial Zone - the lighter area inside the "tropics", between the equatorial belts.

 

As already mentioned, try either adjusting the capture color ratios, or more aggressively color-balance in processing. The SEB and NEB should be somewhere between a light tan and pale red.

Very helpful!

 

You're off to a promising start. Where are you located? I was imaging Jupiter at 74 degrees last night, but the conditions were appalling

USVI, 18.3 degrees latitude.  Long ago, I was a NJ kid.  Exit 36.  Oh wow, nice location. I assume you're on the main island, though. Supposedly the further south one goes (like down to the southern island), the better conditions can get.  Exit 36, eh? Down near Atlantic City. I'm not so far away, about 25 miles east of Philly, just inside the NJ Pine Barrens. But the seeing is truly bad here much of the time. An exercise in frustration and patience.

 

All the best,

 

Kevin

[gstrumol]

For a first attempt, that is quite good! 

 

Here is your 3rd image with all the color stripped away so we're not distracted by it (and some tweaking in GIMP):

 

 

You have a lot of detail in your image. You need to reduce the exposure (so the center is not blown out, like it is in your pic), and get a handle on the color balance.

[Coconuts]

Last night's attempt: Jupiter very near the zenith, 30K frames, 10K used, 4 ms exposure, 400 x 400 pixel ROI, 122 second capture, 245 fps, Player One Neptune gain 384, less sharpening.  I am still struggling on color balance, despite using Astrosurface's White Balance tool.  I balanced the three colors in the histogram prior to capture; all three were around 50%.  Nice to see the Great Red Spot this time.  Comments welcomed.  Progress, maybe?

 

All the best,

 

Kevin

 

 

 23_28_34.CameraSettings.txt   1.32KB   2 downloads

 

 

[RedLionNJ]

Significant improvement on exposure and color balance, Kevin. Good job, there.

 

It's not clear to me which of the following is still "off", but it's likely between seeing, focusing and sharpening. You can't do much about the seeing, other than wait for better periods and know how to judge (from experience) when they're likely and when they've actually arrived (and when they've departed again!).

 

Focus may or may not have been a tiny bit off - one can't really tell from an image, as seeing could account for the effect, too. You have to catch it at capture time. Afterward is too late.

 

If you look at the image (and this is why I'm questioning sharpening), there's quite a sharp edge all around the planet. Given the amount of detail visible on the disk (lack of sharper edges), this contrast sticks out like a proverbial sore thumb. There's something "not right". I suspect a more subtle (dialed-back) sharpening may have been more appropriate for the conditions here.

 

Could you perhaps share your unsharpened image in PNG or TIFF format and encourage others with experience to take a stab at sharpening, see what kind of results they can pull out of it, so you can get a better handle on the degree of processing to apply?

 

Best Wishes,

 

Grant

[Coconuts]

Grant:  Thanks for your feedback, which is very much appreciated.  I'm not sure about the seeing; for daytime solar, I have an SSM to quantify that, but at night I have yet to develop the skills to assess that.  FWIW, the images seemed pretty stable during the capture, and there wasn't much breeze.

 

I have yet to assess collimation.  I dug out my copy of Suiter, only to recall that he pans real stars, and recommends setting up an artificial one.  But I need to at least look at a bright star at high power.  Soon.

 

I am definitely vague on the focusing.  I took other captures with 90 degree tweaks to the fine focus knob.  At f/15, it is hard to tell where true focus lies.  My approach has been to see how many turns (2-3) it takes to get to clearly soft on either side, and then dial in to the middle of that.  I am shooting from a wooden deck, and so having me next to the scope leads to some bouncing.  I tweak, wait, and squint.  On reflection, does SC have a focus quality score that works on planets?

 

During the capture itself, I retreat to a nearby concrete floor.

 

I need to go out for the better part of the day, but since the PNG takes up so little space, I will attach different focus versions later today, after I stack them.

 

I don't like the ring around the planet either.  That is definitely a sharpening or wavelet artifact, but if I lower those to the point where I don't see it, I end up with a soft image.  But maybe only one of the two steps is the culprit, and I can reduce that one, and increase the other.  In a Nazstronomy video on AstroSurface, Naz claims that the HDR step removes the ring, but that wasn't my experience.

 

I wasn't allowed to attach an IrfanView TIFF, so I used a PNG.

 

All the best,

 

Kevin

 

 

It's incredible that anything of value can be squeezed from images such as the one above...

Edited by Coconuts, 05 December 2024 - 09:07 AM.

[RedLionNJ]

Hi Kevin,

 

That stack isn't too bad at all. A little noisy (and dim), but workable. Is there a way you could force-normalize the stack to, say, 75% during the align/stack process? That would help the dynamic range a bit.

 

I took the PNG and worked it a little in waveSharp:

 

 

Quite acceptable for a seven inch scope!

 

Attempting (and succeeding) to over-sharpen when the data is just not there to support it is a common foil for planetary imagers. A slightly "soft" appearance is always preferred over one which accents noise to the detriment of real information.

 

 

Best Wishes,

 

Grant

 

 

 

 

[Coconuts]

Grant:  Thanks again for your help in shepherding me along the not all too obvious path of planetary imaging.  Very nice processing, and a much more natural image.  A few comments:

• "That stack isn't too bad at all. A little noisy (and dim), but workable."
  • I had been using 10 ms exposures, but was recommended to use shorter ones.  So that image was with 4 ms exposure times.  My Player One Neptune gain has been constant (384), so lower all overall histogram percentages (50 vs 85%), the latter of which was suggested as being too high.  Should I leave the exposure time short, and increase the analog gain, to bring up the image brightness?
• "Is there a way you could force-normalize the stack to, say, 75% during the align/stack process?".
  • Could you elaborate on what you mean by this?  Histogram percentages?  I can check if AS has that capability.
• waveSharp looks pretty interesting.  Are you and one of the developers Grant the same?  v2.0 looks imminent, better I think to wait a few days until it arrives.
• One of my issues may have been slavishly following parameter settings I found online, which may not be generally applicable.  Using lower values for sharpening produced softer but more realistic images, example below, which resembles your processing.  Perhaps I merely need to settle for less extreme processing..
• What (if any) planetary focus quality metrics are available to guide focus quality?

All the best,

 

Kevin

 

Astrosurface minimal sharpening and wavelet processing image:

 

Edited by Coconuts, 05 December 2024 - 08:36 PM.

[RedLionNJ]

As far as forcing the histogram level to a fixed value (say, 75%) for all frames prior to stacking - AutoStakkert has this feature, I don't know about AstroSurface.

 

The Grant listed on the waveSharp splash screen is indeed me. Also there are dcaponeii (Don Capone) and borodog (Mike Owen), who contribute a lot to this forum.

 

Virtually no two captures benefit from the exact same processing - don't take anything you find online in that context as factual. It's totally on a case by case basis.

 

If you capture with SharpCap, it has an autofocus which Mike O swears by. For those who don't use it, just a hands-off focuser (which doesn't move the primary mirror), along with patience to wait for better seeing spells is the way to go for focusing. When you get really close to the best focus (during a period of good seeing), finer and finer details become visible, particularly on Jupiter and Mars.  In the case of Saturn, the Cassini division is at its highest contrast, as are the rings (or their shadow) crossing the globe. There are some nights when the seeing never gets good enough to judge best focus. Those are the nights I've learned to do something less futile