New Script: SkyLimitedExposureTarget

[mschuster]

Most broadband imaging sky limited exposure calculators give an exposure time, which depends on setup and observing conditions.

This script works differently. It gives a background exposure target which depends only on detector characteristics. Whatever your optics, filters, and observing conditions, you expose just long enough to ensure that frame mean background is no less than the exposure target. By doing so you guarantee that the loss of background signal to noise ratio due to detector read noise is bounded (typically 5%).

With a fixed exposure target, exposure time depends simply on per pixel flux. Examples: add a narrower bandwidth filter, more time is needed. Add an optical extender, more time is needed. Observe in darker skies, more time is needed.

Here is an example for my camera unbinned. Input is detector read noise, gain, offset (mean of bias frame), and acceptable background SNR loss.

Mike

https://dl.dropboxusercontent.com/u/109232477/PixInsight/Scriptbox/SkyLimitedExposureTarget.0.2.zip



Here is the background SNR loss due to read noise curve for my camera. Background exposure needs to be at least 1000 DN to get loss below 10%, below 5% requires about twice the exposure.

[IanL]

Timely as I am just trying to figure out exposure times for my "new" (in 2006) camera.  Do I have it right that if I take an exposure of the target area, measure the background DN and divide by the exposure time, I just have to divide the script result by that DN per unit of time value to get the desired exposure length?

Might be nice to add a box to input the measured DN and exposure time of the test image have the script calculate the exposure time for you? And also some options for whether the test image is calibrated or raw (and if raw have the script subtract the bias offset for you).

If you were being really fancy you could even pick a preview from an image (set on the background) and measure the average DN for the user, though maybe you'd need to know the bit depth of the camera as well to convert from normalised real to DN.

[mschuster]

Hi Ian,

If measuring a raw (not bias subtracted) I suggest using this formula to account for the fixed offset:

targetTime = (exposureTarget - offset) / (rawMeasurement - offset) * rawTme

Yes the script could do this and also read a frame, but so far I've decided to keep it simple. I will reconsider.

Note that 14-bit cameras when read can be normalized to either [0, 0.25] or [0, 1], so frames need to be double checked to see what is actually happening.

Thank you for your comments!

Mike