sticky exposure length or sticky gain/offset for sky-limited exposures?
Posted: Wed Sep 12, 2018 9:08 pm
Some advice would be greatly appreciated.
I'm about to start a new project and I'm contemplating which would be the best approach. I'll be shooting the Heart Nebula and wanted to see how long I could collect data for it. Each night, I'll go out and capture my respective mono channels (as 60 minute auto-saved livestacks) i.e LRGB, come back and process a first edition. Then, on subsequent nights, shoot respective filter live stacks again and then come back and doublestack them with the others in DSS, reprocess and release a new edition. When I 'feel' that the latest edition is no better in SNR, looks, detail etc.. than the previous iteration, I'll call it ends and move onto another target.
Now my conundrum is actually how to go about getting my sky limited exposures (as guided by the smart brain analysis).
One way, is to keep the same time in the exposures for each filter and use gain/offset to shape the histogram correctly. This means that I would end up with four exposures, each of them say 30 seconds and each of them having a different gain/offset. This would nescessitate different darks for each filter setting, but all of them 30 seconds. Probabity of matching guiding/tracking behaviour during the same exposure lengths also improves the chance of stars remaining sharp to each other between filters.
The other way, is to keep gain/offset the same, but this time vary exposure length to achieve that good sky limited histogram. However, this means that longer subs might have an increased probability of guiding/tracking issues. There's also a minor consideration of increasing dark noise/amp glow, which proper darks should take care of that. In this scenario, master darks for each filter will share the same gain/offset but have varying exposures. This approach is akin to how many shoot with their cameras set at unity gain and then vary the exposure length accordingly to compensate for the different filter bandwidths.
Considering I'll be doublestacking the live stacks, it means that if I choose to keep exposure times the same and vary gain/offset, then the 60m livestacks for each channel will all contain the same number of frame (120 if we use the 30s example above). If I go the other way, with maintaining gain/offset and varying the exposures (like the unity gain example), my 60 minute livestacks will all have varying number of frames for each filter. The number of frames contributing to the 60 minute livestack are obvious considerations for the sigma clipping and SNR aspects. This then begs more questions, such as how many frames to aim for to maximise my 60 minutes livestack? How many initial frames should you count from the livestacking tab, 10, 15, 20, or the max 25? Then the clipping sliders below come into question, how much pixel percentage should I reject? The goal here with the stack is to produce the best possible 60 minute live stack, for later doublestacking in DSS (or SC virtual monitor).
Also, I should say I'm willing to sacrifice maximum dynamic range, in order to keep my exposures relatively short (for better star shapes), up until say 3 minutes max, but ideally shorter. The goal for each individual exposure is to get a nice fat histogram off the left wall that occupies the good first half of the chart. I'll have to see how this works with sky limited exposures..
I know there are many questions in my text above, but it utlimate boils down to this:
For sky-limited exposures - sticky exposure length or sticky gain/offset?
..and how many frames for the best 60 minute livestack?
Hope this is clear and looking forward to any advice.
Cheers
Minos
I'm about to start a new project and I'm contemplating which would be the best approach. I'll be shooting the Heart Nebula and wanted to see how long I could collect data for it. Each night, I'll go out and capture my respective mono channels (as 60 minute auto-saved livestacks) i.e LRGB, come back and process a first edition. Then, on subsequent nights, shoot respective filter live stacks again and then come back and doublestack them with the others in DSS, reprocess and release a new edition. When I 'feel' that the latest edition is no better in SNR, looks, detail etc.. than the previous iteration, I'll call it ends and move onto another target.
Now my conundrum is actually how to go about getting my sky limited exposures (as guided by the smart brain analysis).
One way, is to keep the same time in the exposures for each filter and use gain/offset to shape the histogram correctly. This means that I would end up with four exposures, each of them say 30 seconds and each of them having a different gain/offset. This would nescessitate different darks for each filter setting, but all of them 30 seconds. Probabity of matching guiding/tracking behaviour during the same exposure lengths also improves the chance of stars remaining sharp to each other between filters.
The other way, is to keep gain/offset the same, but this time vary exposure length to achieve that good sky limited histogram. However, this means that longer subs might have an increased probability of guiding/tracking issues. There's also a minor consideration of increasing dark noise/amp glow, which proper darks should take care of that. In this scenario, master darks for each filter will share the same gain/offset but have varying exposures. This approach is akin to how many shoot with their cameras set at unity gain and then vary the exposure length accordingly to compensate for the different filter bandwidths.
Considering I'll be doublestacking the live stacks, it means that if I choose to keep exposure times the same and vary gain/offset, then the 60m livestacks for each channel will all contain the same number of frame (120 if we use the 30s example above). If I go the other way, with maintaining gain/offset and varying the exposures (like the unity gain example), my 60 minute livestacks will all have varying number of frames for each filter. The number of frames contributing to the 60 minute livestack are obvious considerations for the sigma clipping and SNR aspects. This then begs more questions, such as how many frames to aim for to maximise my 60 minutes livestack? How many initial frames should you count from the livestacking tab, 10, 15, 20, or the max 25? Then the clipping sliders below come into question, how much pixel percentage should I reject? The goal here with the stack is to produce the best possible 60 minute live stack, for later doublestacking in DSS (or SC virtual monitor).
Also, I should say I'm willing to sacrifice maximum dynamic range, in order to keep my exposures relatively short (for better star shapes), up until say 3 minutes max, but ideally shorter. The goal for each individual exposure is to get a nice fat histogram off the left wall that occupies the good first half of the chart. I'll have to see how this works with sky limited exposures..
I know there are many questions in my text above, but it utlimate boils down to this:
For sky-limited exposures - sticky exposure length or sticky gain/offset?
..and how many frames for the best 60 minute livestack?
Hope this is clear and looking forward to any advice.
Cheers
Minos