Total exposure and dynamic range tables ZWO ASI1600 - long post

[Shiraz (Ray)]

Hi

"How long should I expose for?" has been asked many times. As a rather hopeful attempt at an answer, please find attached some luminance calculations of sub length, total exposure time and dynamic range for the ASI1600MM-C. These apply to galaxy imaging under two different sky conditions and for two scope fnumbers. They show what can be expected as the gain varies and were derived using a moderately complex end-to-end astrograph model.

The sub length calculations are based on the Moore/Smith 5% RN criterion used in previous tables. The starting point for the exposure time calculations was a personal assessment that a stack of 84 subs of M104 was at a minimum acceptable SNR in regions where the surface brightness was about 23 mag/arcsec^2 (assessed from published photometry). This scaled conveniently to a starting point of 12.4 SNR after 1 hour at f4 under mag 21.5 sky - at gain100. Of course, this is a personal assessment and others will have other criteria. The (objective) sub length and (subjective) time to the chosen mimumum SNR could be combined to estimate the dynamic range (DR) after stack (max possible signal/sky noise). The model provided DR estimates that matched very well with the measured DR from a number of real stacks - to some extent validating the procedure.

Only a limited range of parameters have been investigated and the tables should be taken as showing general trends rather than as a definitive guide on how long to expose for. Nonetheless, they provides a way to illustrate some of the characteristics of imaging with the camera.

How to read the tables:
Choose the sky condition nearest your own and then look for the fnumber nearest yours. That section shows the sub length needed for sky limited imaging, the total time needed to get to a SNR of 12.4 on a galaxy region with surface brightness of 23 mag/arcsec^2 and the dynamic range of resulting stack - all for different values of gain.

Some observations:
1. dark sky is of extreme importance. going from mag 21.5 to mag 19 increases the imaging time to the same SNR by about 8x. This doesn't mean that galaxy imaging cannot be done under bright sky - just that dimmer regions may be difficult to reach in a reasonable time.
2. Under dark sky (mag 21.5), provided that the sub length is chosen to give sky-limited imaging, the total time and dynamic range is pretty much independent of gain for values below about 200. At gain 300, the optimum sub length is less practical and the dynamic range has fallen off considerably. ie - it doesn't matter much what gain you use up to about 200 - provided the number of subs is manageable. The widely held assumption that the 1600 has best dynamic range at low gain doesn't really hold after stacking and in dark skies. The subs will individually have more dynamic range at low gain, but the stack will be about the same for gains up to about 200.
3. Under brighter sky (mag 19), the total imaging times are so long that optimum (short) subs will be fairly impractical. The table is based on the assumption that 1 minute subs will be used throughout under bright sky (even that is a bit short for 30+ hours of imaging). Clearly the dynamic range can be much greater than under mag19 sky, but at the cost of ~8x the total integration time. The dynamic range falls off as the gain increases, so it is probably best to use low gain under bright sky and at least get some advantage from the conditions.
4. Using a slow scope at native resolution for galaxy imaging does not seem to be all that practical with the 1600 under brighter skies - long exposures are required to get to reasonable depth.

Please treat these tables as general guides only - the sub length calculations can be traced back to objective measures, but the total exposure and dynamic range are based on a subjective assessment. However, the general trends might be useful.

Thanks for looking. Cheers Ray

[glend (Glen)]

Thanks Ray, always great to see this sort of analysis from you. One question, in your previous sets of ASI1600 charts, as in the Optimum Broadband Sub Length chart (which included max electron count after stack at various Gain settings), your sub time recommendations were slightly different. Not sure I understand why they changed, and I am ignorant of the calculations required. In my experience the original chart times, were pretty good and usually I was able to adjust (for local conditions) to the recommended sub background sky ADU, which I think was 400 at the time. Can I still go by sub background sky ADU? It seemed a simple test that even I could manage.

[Imme (Jon)]

Looks good Ray, plenty of work put in to that for the benefit of others!

A question for you - when applying these exposure time to the asi1600 OSC do we need to extend times to compensate for the Bayer matrix?
I've heard conflicting views. Your thoughts would be appreciated.

[billdan (Bill)]

Wouldn't aperture size vary these figures, e.g a 12in F4 will collect light faster than an 8in F4?

[h0ughy (David)]

so what would F2.2 do to the results?

[JA]

Originally Posted by h0ughy

so what would F2.2 do to the results?

It (f/2.2) is about 2 photographic stops (EV) faster than f/4 so I'd expect all of the times in any f/2 tables to be one quarter of what they are in the Shiraz's published f/4 table. Just like the published f/4 table has approximately one quarter of the times that are in the published f/8 table, which also differs by 2 stops. It's really no different to equivalent photographic exposures

I'm not sure if the f/2.2 f number takes in to account the central obstruction, but if not that would need to be corrected for if one wanted to know any exact result which took account of the true light collection area. Not sure how much or whether that is already factored in to the f number.

Best
JA

[Shiraz (Ray)]

Originally Posted by glend

Thanks Ray, always great to see this sort of analysis from you. One question, in your previous sets of ASI1600 charts, as in the Optimum Broadband Sub Length chart (which included max electron count after stack at various Gain settings), your sub time recommendations were slightly different. Not sure I understand why they changed, and I am ignorant of the calculations required. In my experience the original chart times, were pretty good and usually I was able to adjust (for local conditions) to the recommended sub background sky ADU, which I think was 400 at the time. Can I still go by sub background sky ADU? It seemed a simple test that even I could manage.

Hi Glen. this does not take the place of the earlier sub exposure tables, but is an attempt to extend that analysis to provide an estimate of the total exposure time and the dynamic range as well. The original tables are objective and seem pretty reliable, whereas these total time and dynamic range estimates include a subjective assessment and are less reliable. Suggest that you continue using the original sub exposure tables if they are providing useful information.

I tried to ensure that there would not be any clash between earlier data and this - if there is, I may have used slightly different assumptions in this analysis. Apologies, will cross check. However, the results should still be valid. Cheers Ray

Originally Posted by Imme

Looks good Ray, plenty of work put in to that for the benefit of others!

A question for you - when applying these exposure time to the asi1600 OSC do we need to extend times to compensate for the Bayer matrix?
I've heard conflicting views. Your thoughts would be appreciated.

The Bayer filters absorb ~2/3 of the photons before they get to the sensor surface, so the OSC will need about 3x as long to get to the same result.

Originally Posted by billdan

Wouldn't aperture size vary these figures, e.g a 12in F4 will collect light faster than an 8in F4?

Yes, the aperture size makes a difference but it is directly compensated for by an equal and opposite sampling effect. To illustrate, consider the change from an f4 scope of a given aperture to an f4 scope of twice the aperture. The bigger scope aperture will collect 4x as many photons from any part of the sky. However, in the bigger scope also has 2x the focal length, so the area of the sky from which an individual pixel can collect light decreases by a factor of 1/4. If the pixel size remains the same, the only thing that matters in determining the number of photons in a pixel is the f/number (for an extended object).

Cheers Ray

[Shiraz (Ray)]

Originally Posted by h0ughy

so what would F2.2 do to the results?

Originally Posted by JA

It (f/2.2) is about 2 photographic stops (EV) faster than f/4 so I'd expect all of the times in any f/2 tables to be one quarter of what they are in the Shiraz's published f/4 table. Just like the published f/4 table has approximately one quarter of the times that are in the published f/8 table, which also differs by 2 stops. It's really no different to equivalent photographic exposures

I'm not sure if the f/2.2 f number takes in to account the central obstruction, but if not that would need to be corrected for if one wanted to know any exact result which took account of the true light collection area. Not sure how much or whether that is already factored in to the f number.

Best
JA

yep, just did some model runs to make sure there are no hidden gotchas - the subs are about 1/4 as long and the total time is about 1/4 as long at f2 cf f4. The dynamic range remains the about same. There is a minor downside of having very short subs - the download and dither settling times become significant compared to the imaging time and somewhat limit how quickly you can image. However, that is not a bad problem to have. Cheers Ray