1.20
#...
·
4
likes
|
|---|
|
Hi! I'm using a TS65 f6.5 telescope 420mm. With my Sony a7 I am able to get nice pictures fitting for this relatively wide FOV. But when I try to go for smaller motifs like in this example, trying to enlarge by crop, the results get very coarse. My combination of telescope and full format camera with relatively large pixels gives oversampling. To utilize the hidden information I use Drizzle 2X drop .5 (Astro Pixel Prosessor) to effectively double the picture resolution. This allows me to double the enlargement by crop maintaining the details. You do not get anything for free! In this case it is the cost in storage and prosessing of the 4 times bigger files. Also, to maintain the high quality I have done post prosessing in the editor (Affinity) in tiff 32 bit float, which puts high demands on memory and CPU. See this Example: https://astrob.in/hyvasy/B/  |
1.20
#...
·
2
likes
|
|---|
|
You could almost double your focal length with a 200mm newtonian and not break the bank. The longer focal length would like the large pixels. The main problem with a fast (f/3.9) Newtonian and your full frame camera would be coma. So many options and so many trade-offs. We are all addicted to a crazy hobby. You many need to buy a 64 core computer ($$$$). Good luck to all of us, Monty |
3.61
#...
·
3
likes
|
|---|
|
Hi Gustav, +1 for trying to get the most out of your equipment! Some of us are playing with short focal lengths and drizzle can work wonders in undersampled conditions! You'll find examples on Astrobin, e.g. here, here, here. In your case x2 drizzle is equivalent to working with a 860 mm focal length, which can bring out a lot of details. There are drawbacks to drizzle, but workarounds  With 0.5 drops, the signal per (drizzled) pixel is divided by 4 compared to the non-drizzled stack. So in practive your F/6.5 system works at F/13, which is pretty slow. The only solution there is to increase the number frames, over several nights if required. Having a lot of well dithered subs is in any case a prerequisite for he drizzle algorithm to perform at its best. One has to be patient, but it is rewarding! To speed up your processing and/or for tests, you could crop all your subs beforehand and then run the drizzled stacking on these. Given the size of your crop that could be even faster than a x1 stack of the full frame CS Frédéric |
|
1.20
#...
·
1
like
|
|---|
|
Thanks for good advice and knowledge! I have already used the technique of integrating only a lesser FOV in connection with my experiments with a dual telescope with different FOV. Also the advice about the necessity of integration time due to the 0.5 drop size is good. I just wonder if my efforts to shoot two sequences in synchron parallell will alleviate need for dither since the two sessions have different sensors and optics? The plan is to let one session do RGB and the other dual NB. What do you think? |
