My seeing is usually only average at best, with the jet stream always overhead. But we had a long unusual stretch of 7 days in April with 20 to 30 C temperatures. This one morning the seeing was rated as excellent, with high transparency, no wind and with a very still hot and humid air mass above. Even without any tuning, my initial look at the Sun was like nothing I had seen before, showing incredible detail. I spent a lot of time getting everything focused and my Etalon fine tuned, waiting for temperature on my sensor to stabilize near 38C, before starting this sequence.

With my ASI120mm-S camera I switched to high speed mode, which for this camera is still the full native 12 bit it is capable of, but running at 116 fps. Given the great conditions I was able to shorten my exposure to 3 ms. With 15 second SER files, this meant I captured about 1700 frames for each time point in the animation. I tested AutoStakkert with 10% and 5% of frames, with the clear winner in terms of resolution being 5%, with very little noise over 10% frames. I used 3X drizzle since I am at prime focus, not using a barlow.

I struggled with Registax to enhance the image, everything seemed to bloat the fine detail. Further, and I have seen this before, for fine detail Registax will pull out pixelation really well, as of course it should, since the pixelation is real detail, if not what one wants.

As many have found Topaz DeNoise AI is very aggressive, and can create artificial sharpening, and thus has to used extremely careful to avoid defects. This was not the right tool to use anyway, as my S/N was very good, with little background noise. 

I found that Topaz Sharpen AI is actually very subtle in sharpening: one can choose to remove motion blur, focus blur, or overall softness. I used Focus blur-very blurry at the AI generated settings and got a wonderful natural looking improvement in focus. One really cool thing about Sharpen AI is that it has obviously been trained extensively with regard to pixelation, most pixelation artifacts were removed, as the training recognizes this as a defect. Any remaining artifacts, there are a few, are from AutoStakkert, though overall AutoStakkert as usual did a great job stacking.

Each SER raw video was taken 15 seconds apart, so the 49 stacks in this animation represent 25 minutes of real time. My final animation is 10 seconds long, so is sped up a factor of 150X over real time. 

I'm going to give Sharpen AI more of a workout in the future, for day and night images, I am much more impressed with it so far than with DeNoise, which tends to be difficult to control. And I'm looking forward to summer, and hopefully some more heavy humid and sluggish air masses. Meanwhile I am back to cold and rainy for now...