Moonshots, compared

[DavidWright2010]

I remember reading, many many year ago, that the Moon was best observed (or photographed?), with a 5 inch aperture telescope, as atmospheric turbulence was generally worse at a larger scale.

Also, looking at the turbulence at 16x magnification with my 9.25" edgeHD, more often than not, the Moon image was jumping around as if it was being viewed through boiling water.

I recently got a new camera that could help in testing whether smaller apertures are better, a M43 from Olympus. So now my two setups are:

1. The E-M5.3 Olympus (20mp) with a Celestron C-5 spotting scope and mounted on a Celestron AVX drive. (This scope is really for visual work, and has a pretty strongly curved focal plane, but for less that full illumination, I could keep the terminator away from the edge of field.)

2. A Celestron 9.25EdgeHD telescope with a Pentax K-1 (full-frame 36mp) camera, mounted on a Sky-Watcher EQ6-R mount.

#2 has a slight advantage in 'pixels across the target", 1.26 as many as #1. But #1 has a 'secret weapon'. And, it's a whole lot cheaper.

So last night I took some shots with the Olympus first. It doesn't have the features I use on the Pentax (or maybe I can't find them), so I settled for 'Continuous shooting'; which only allowed up to 10 images, but has a delayed start to let the gear settle down after I pressed the shutter,

Then I took a couple of sets of 16 images with #2, using 'mirror up" and delayed shutter with a corded remote shutter button.

I was about to pack up, but thought about the 'secret weapon', and went back to #1 and took several HR images (9 images, with about 100 ms between exposures, and pixel-shift of the sensor giving an 80 mp result.)

I developed the Pentax raw files in DxO PhotoLab with 'xtra deep prime NR', and then Auotstakker! (using the sharpened resultant image). I developed the Oly raws in the Oly studio (Nobody else knows how to develop the HR images.) All 3 80mp results looked about the same, so a selected one at random. These Oly images were noisier than the Pentax, because they didn't get the 'deep prime NR, so I fussed around a bit, and after sharpening them with the wavelet feature of Registax 6 (because the Pentax set got sharpening with AutoStakkert!) , I then passed them through both 'Neat Image', and Topaz Sharpen AI (too soft - very noisy mode)

So here we are:

Oly/C-5 on left, Pentax/edgeHD on right

The main difference I see is the the large-scale shading in the flat plains is missing. That's probably due to too much NR on the first wavelet step, but I really don't how to get the best result from the Registax app.

FWIW, I'd put last night's stability at 8 (10 being completely steady), as viewed at 16x magnification.

David

 

[Alen K]

Given the great differences in processing, I’m not sure that you proved anything about the effect that aperture has on the visibility of air turbulence in images.

But I do think you forgot that the K-1 has its own “secret weapon” called Pixel Shift Resolution System. The images it gives don’t have more pixels but they do have lower noise and they can have higher actual resolution, provided the objects in the image don’t move during the process of taking the four images.

But because of air turbulence, they can move when you are imaging the moon. (Tracking is assumed to be happening.) The processing can actually take into account any movement, but typically by reverting to using pixels from only one sub-exposures for the parts of the image that moved between them.

Those considerations should also apply to the HR images from the Oly. Lucky for you that the seeing was so good that night or they could have been greatly affected.

 

[DavidWright2010]

Given the great differences in processing, I’m not sure that you proved anything about the effect that aperture has on the visibility of air turbulence in images.

But I do think you forgot that the K-1 has its own “secret weapon” called Pixel Shift Resolution System. The images it gives don’t have more pixels but they do have lower noise and they can have higher actual resolution, provided the objects in the image don’t move during the process of taking the four images.

But because of air turbulence, they can move when you are imaging the moon. (Tracking is assumed to be happening.) The processing can actually take into account any movement, but typically by reverting to using pixels from only one sub-exposures for the parts of the image that moved between them.

Those considerations should also apply to the HR images from the Oly. Lucky for you that the seeing was so good that night or they could have been greatly affected.

You are right, of course. This test proves nothing because there were too many differences. I was just pleased that the set up number one did so well.



Right at Sunset, the seeing was even better. I judged it to be a nine out of 10. But I waited an hour to use the set up #2 because the contrast gets better as the sky gets darker.

i’ve tried the Pentax Pixel shift on other occasions, when the seeing wasn’t so good, and it didn’t work at all.



David