FD May 19 - first higher res with 3nm CaK

[SolarMathmo]

On Sunday I took the Altair Wave 125 out to see if I could get higher resolution and closer to critical sampling with their prototype 3nm CaK filter. I made a bit of progress and eventually got a rig running at around f/19.5 using an 80mm mask (perhaps a bit optimistic on seeing) together with the APM 1.6 Barlow, and a Hypercam 26M running with an ROI of 4168*4168. This was a testing morning for seeing and I spent a "happy" couple of hours watching live video zoomed in to show spots changing size and shape. About half way through this I got Sharpcap's feature tracking working and later got the seeing thresholding/filtering working. (I still have a lot to learn!). I made a few runs and this was the best of a tricky group. It was a 60s run producing 157 frames and this is 4% of those stacked in AS4, with some wavelets added in Registax, and then Pixinsight's solar toolbox.

Click this link for the full res version: https://www.astrobin.com/full/m42v9p/0/

May19CaK4k.png (19.37 MiB) Viewed 999 times

The spot detail remains a bit scrappy, but for a first serious attempt with this combo I was pleased to at least see a bit of spot detail within a full disc. Like many of you folks seem to do routinely! This is 120m up in the Thames Valley, Oxfordshire UK. The contrast is not of course what you would get with an Angstrom-class filter.

[Carbon60]

Hi,

157 frames in 60s is really slow and 157 isn’t very many to get ‘lucky’ with lucky imaging. Add to that the small number of stacked images (only around 6 at 4%) and you’re making it tough for yourself to get the best from your kit.

You’d be better with a faster camera and take 500-1000 frames in a shorter span of time time. Many of us use ImPPG for deconvolution and sharpening after stacking.

Hope this helps.

Stu.

[marktownley]

Nice disk!

I was struggling with the seeing at 80mm @ 393nm on Sunday, to give a context I was shooting 30000 frames (10x3000 frame runs) per feature and stacking the best 100 from each to get a 'best image' from a sample of 10 to choose from, even then not all of them were coming out as resolved as I would like. I'd probably stick to the 533 camera, you'll get a 1000 frame run in less than a minute, and at this scale full disk there won't be any temporal distortion, shoot 5 x 1000 frame runs stacking 10% and then pick the sharpest disk.

Unless you're shooting just after dawn this time of year you'll find you get better results at 60mm for the full disks.

[SolarMathmo]

Thanks guys. I think this camera frame rate is a bit low - this was an experiment to explore the 4k^2 zone and try out feature tracking and seeing thresholding at the same time. I think I will put the 533 back on at 60mm and do some 2400 frame runs taking about a minute, with feature tracking on and add just the seeing trigger for each to start, rather than filtering the frames. I'll let AS4 sort out the quality selection.

Mark - are you shooting in a higher bit depth? - I get 40fps in MONO8 so can get 2400 in a minute. I keep reading different things about 8 bit vs higher. I can certainly get 1000 in a minute at 14bits.

Stu - agree about the very small size. The 157 was prefiltered for seeing by Sharpcap though so it's like taking about 15% of a thousand. My mistake on the low % though.

[marktownley]

I'm shooting in 8 bit, i've experiment numerous times, with numerous cameras over the years and find 8 bit fine, I prefer the higher frame rates it brings. I'm shooting around 27fps for my full disks, 160fps for my closeups.

[Montana]

It's a very beautiful disc with lots of details, you should be pleased with the result

Alexandra

[christian viladrich]

This is an interesting test. Resolution is quite good in some parts of the image.
Processing is very fine !

I would suggest the following :
- go back to the IMX533 (Player One ?). Indeed, 4 fps is really too low,
- extend acqusition to 120 s acquisition. AS3 or 4 can easilly handle the very small solar rotation during these 120 s,
- forget about % in AS, and just use an absolute number. For the IMX 533, use the following stacks : 7, 15, 30 and 60 frames. Usually, the 7-frame stack is the best one,
- stick to 8-bit acquisition with the IMX 533 for the time being. Later on, and if you have the Player One, and if you have good seeing, you can indeed get an increased dynamic range using 14-bit acquisition. But I suggest you leave this for latter one.

[christian viladrich]

Thinking back about it, you might also like to keep this camera and just extend acquisition time to 120 s.

[SolarMathmo]

This is an interesting test. Resolution is quite good in some parts of the image.
Processing is very fine !

I would suggest the following :
- go back to the IMX533 (Player One ?). Indeed, 4 fps is really too low,
- extend acqusition to 120 s acquisition. AS3 or 4 can easilly handle the very small solar rotation during these 120 s,
- forget about % in AS, and just use an absolute number. For the IMX 533, use the following stacks : 7, 15, 30 and 60 frames. Usually, the 7-frame stack is the best one,
- stick to 8-bit acquisition with the IMX 533 for the time being. Later on, and if you have the Player One, and if you have good seeing, you can indeed get an increased dynamic range using 14-bit acquisition. But I suggest you leave this for latter one.

Thanks Christian - that is helpful advice especially on the stack sizes. I have the (cooled) Altair 533M and that does 40FPS in 8bit (I think the P1 claims 43 but I already had the Altair for DSO work). The 120s is good to get a confirmation on - I was never quite sure about the video exposure length on an FD. I'd recently started to notice that the stacks with small numbers were doing best so will go to the absolutes, especially now I am trying sometimes trying seeing filtering which makes the % useless.

[DavidP]

Well hell. I think this is a pretty nice image with good detail..
I also use 8 bit on my 533 because of frame rate. I’ve compared it to 14 captures and have yet to seen improvement..