Howling at the Moon, 8

:: home ::
              Go to  1   2   3   4   5   6   7   8   9   10   11   12   SRCH

The knee bone's connected to the...


10/16/2021. In the twilight, I tried several configurations of adapters, spacers, Barlows, and filters as well as the new Imazing 20,000 mah (why not just say 20ah?) battery pack / jumpstarter.

The Clear Sky Clock said seeing this evening would be dreadful. I didn't need good seeing. I used the waxing gibbous Moon to find and record some solid connections between telescope and camera. The seeing was every bit as bad as forecast. I tried an IR-pass filter and a solar continuum filter to see what effect they would have with a small aperture in really lousy air. The IR-pass filter requires about 10x more exposure than no filter; the continuum filter takes about 3x as much exposure as the IR-pass filter. Tonight, the faster exposures, the faster frame rate, and the deeper stacks the absence of these filters allowed mattered more than the filters. In short, the filters might improve average or good seeing but they won't rescue atrocious seeing.

As I was getting started, I looked up to see the shadow of the Earth rising in the east, the "Belt of Venus." I sent this photo to Paul Lewis on Facebook:




10/17/2021. While documenting yesterday's arrangements, I realized how incredibly clumsy and (if I may) stoopid some of them were. So I hauled the kit up to the cul de sac and tried saner connections around midnight. Perfect. I dug the old TeleVue "Big Barlow" out of the case with binocular viewing accessories and used it in place of the GSO Barlow.

Measuring from Copernicus to Eratosthenes yields these plate scale measurements:

  • Native FL: 160 pixels. (That's known to be 506mm F.L.)
  • GSO Barlow only: 324.
  • TV Barlow only: 326
  • TV Barlow + 1.25-inch Svbony Barlow "close coupled:" 542 pixels

I thought the TV was a noticeable improvement. Next time, try the GSO Barlow and the TV Barlow together, reserving the Svbony barrel sans lens as a filter holder. That could be a nifty setup. And think about finally adding one of those 5x Powermates to the kit; it would simplify much and use only premium glass for the high magnification efforts.

The Imazing battery pack will not run more than 30 seconds before unceremoniously turning off the 12v output. It worked well last night, but is useless this night. I plugged into the Subaru's accessory jack since I hadn't recharged the heavy AGM battery. Imazing Power's support line responds quickly but without giving me the impression that they're on top of this issue. I was convinced that the unit was faulty; they apparently agreed and sent me a new IM29 power bank.

[Update 10/23/2021: The new IM29 behaves exactly the same as the first one. That lead me to think that this is not a defect but a design "feature" that needs to be figured out and documented better. I tried the new powerbank on the A-P CP3 and Mach1 sans OTA. It seems happy to run it continuously. So does the old one. Maybe we just need to add this line to the "how-to" sheet for that product: "The DC power outlet will cut off after 30 seconds when no load is detected or when the attached load falls below the threshhold for detection." It makes sense since there's no "off" button. Now, where is the limit? The Meade Autostar 492 draws 120ma when idling and is insufficient. The RA motor of a well-balanced LXD-55 mount carrying a light instrument doesn't add enough. The CP3 and RA motor on the A-P draws about 500ma, maybe somewhat more since the mount was unbalanced. It appears that I got lucky the first time I used the powerbank: I probably had an east-heavy OTA and the LXD-55 drew enough current to let the powerbank know it was there while lifting the tube against gravity. I suppose I could routinely misbalance in that way, or I can add a small load to the kit. The ASIAir might suffice; the cooler on the ASI1600 surely would; so would a dew preventer (which, frankly, I've so seldom needed that I don't own a real one). Anyway, I've got options. I offered to return the new powerbank or preferably buy it "at their best price."]


10/19/2021. I burned 130GB of storage during the previous session and spent much of the next day refining the videos. I got tons of practice in Autostakkert!3 and IMPPG. Here's a selection from the photos in the cul de sac the night of 10/17-18. Each is the best 500 frames out of 2,000.


1200x800 ROI w/stacked Barlows


1200x800 ROI w/stacked Barlows


1200x800 ROI w/stacked Barlows


1200x800 ROI w/stacked Barlows


Plato and Sinus Iridum
2400x2000, stacked Barlows


2400x2000, stacked Barlows


This is coming right along.

Two lessons: (1) I did a mosaic of four captures of Plato and Sinus Iridum using a ROI of 1200x800 and reshot the field with an ROI of 2400x2000. Save the effort; use as much of the chip as you like. Copernicus was also shot with the wider ROI. The only downside is fewer frames per second, and you still have plenty of those, ~40 instead of ~100. (2) With tall stacks, under-exposure is really not a looming problem. Some of those captures live deep on the left side of the histogram, probably 2 stops under what they should've had. Get the exposure right when you can, but if you need a faster than ideal shutter speed, it'll be fine if the stack is tall enough.

A new project: around Halloween and again around Thanksgiving, the Moon's librations will bring Mare Orientele and its environs into excellent position with the Moon high in the morning sky. Be ready. I plan to standardize on the TMB92 for a while longer. I really want to learn the ropes with the little guy before putting my back at risk and courting frustration with the R-C (but I'm thinking that with bigger glass, tall stacks of short exposures with the IR-pass filter will be striking).

Yeah, it will be. While looking for previous attempts to look across Mare Orientele to the far-side mountain ranges, I founf Claude Navarro's shot of that subject. He show this earlier this month using a C14 and a red filter. Cut and paste this into your browser:

You might need to be a CloudyNights member to see it, if so, just take my word for it: it's great. I copied it, sampled it down to what I might get with my smaller glass, and the subject is well worth pursuing.

10/29/2021. Be aware of something called the "Gibs effect" or "Mars rind" which produces artefacts on bright features (high contrast lunar features, limbs...). You've seen this recently when doing preliminaty imaging of Mare Orientele and it may have much to do with the problems you see around bright crater rims, isolated sunlit features, etc. What you think is computational "ringing" may be an exagerated diffraction phenomenon. Partial mitigation might be accomplished by layering with the unsharpened image, using a different blend mode with the emboss trick, etc. Experiments are in order. See the bullet points below for your homework. I had some success with the Hadley Rille data from 10/27 -- enough to be suggestive, not to show off. The rille region appears in at least three frames from that morning with different scales, exposures and sharpness. Play with those clips. Here's a link to a good discussion of at least some of what I'm on about:

Some random thoughts from the shower, jotted down before hitting the sack:

  • are these primarily diffraction effects? If so, they ought to be worse with small apertures pushed hard, which is, after all, exactly what planetary imaging is all about (Charlie Brown). But see if the effect varies in a way that maps to diffraction effects by playing with aperture masks, central obstructions, etc. Look at the brightening inside crater rims and measure whether that corresponds to diffraction. Note that hi-rez efforts are ALL about reaching the diffraction limit, so it stands to reason that artefacts visible at that limit will become increasingly problematic as you approach success. You're trying to image at the scale of the Airy disk. (What does the NSA etc do to get the most from spy sat imagery? The Hubble literature might have something useful and accessible.)
  • how does sharpened defocus relate to what I'm seeing? the edge profile is different, and the effect of defocus should be different. Likewise seeing, likewise differential refraction (which is not an issue with the 10/27 images mentioned above since the lunar elevation was on the order of 80 degrees).
  • What can be modelled or predicted or even isolated in an image can be mitigated to a useful degree even if not entirely eliminated.
  • Think how the emboss trick works and do it better. Play with sharpened and histogram-adusted emboss layers and different blend modes (darken was initially promising in addition or instead of the original soft light -- what is the soft light algoirthm, anyway?). To make a better layer, try edge-detection. Try Isolating the brightest, sharpest features, defocus them, construct a layer consisting of the outlying regions, and experiment with layering and curves. That ought to be enough of a hint without my giving myself (another) headache trying to describe how that could work. If it's a diffraction effect we're addressing, a single good layer ought to be good for anything shot with a specific telescope, right? So it's worth working it out.
  • I don't think this explains the systematic smearing of AS!3 stacks. To get at that, try some intentionally drifted photos and see if the result is worse. Examine the direction of drift in source videos and see if it corresponds to the image degredation. If so, then it's just a matter of aligning the mount better.

:: top ::





                   © 2021, David Cortner