›

Exposure Tolerance vs HDRComposition

Alan Kubiak
in PixInsight
If I have multiple timeframes of one filter.    If I  combine them in WBPP using the Post Calibration Exposure tolerance to create one master for that filter, how does that compare to making a master for each time frame and then combining them with HDRComposition.    Which way is preferred??

For example...  Lum 30S, 60S, and 120S sets of masters.   If I use an exposure tolerance in WBPP post calibration large enough, WBPP will combine these three timeframes into one master.   How does the process differ from HDRComposition??

Thanks

Comments

  • BlockHeadBlockHead
    edited April 2022
    Hi Alan,

    When you combine (integrate) multiple frames into a single one- you are attempting to treat each frame as identical after proper normalization and weighting. It is as if you are combining a bunch of frames to get the best measurement (signal-to-noise result). 

    When you use HDRComposition- each frame is treated independently. The only reason you use HDRComposition- is that an object is so bright that the sensor will saturate quickly- requiring hundreds if not thousands of short exposures to reach a particular signal-to-noise level of a normal integrated image.

    The Orion Nebula is one of about 5 objects that really requires HDRComposition. Look at the faint part of the nebula (the outer parts). You want to expose there for 3 hours to get a nice result in the outer parts that isn't noisy. But if you want a single frame per exposure that captures ALL OF THE NEBULA without saturating (the inner stuff)- your exposure will be 10 seconds long. This equals 1,080 frames. Now...what if you do take 18 x 10minute exposures...so now you have 18 frames. These frames will have the center region blown out and useless. But the outerparts are exposed properly.  And you also take 20 frames at 10 seconds. Look at how few frames you have!  Finally... in HDRComposition all parts of the image that are saturated in the integrated 10 minute exposure frame are SUBSTITUTED by the linear fitted same parts of the integrated 10 second exposures. This is NOT an integration... this is a substitution. The linear fitted results give you the values in the heart of the Orion nebula for what they would have been had the sensor not saturated. You are extending the dynamic range of the data by using this method. Hence H D R Composition. 

    For 99% of astronomical objects you are going to pick the longest exposure that doesn't saturate and ends up with a reasonable number of frames to integrate. For HDRComposition.... TWO different exposures is actually enough... but to help make the transitions between the various linear fitted exposures go together better... sometimes 3 exposures are used. 

    Please rate this response.
    Thanks,
    -the Blockhead
  • Alan Kubiak
    Excellent explanation, thank you, the new NSG 2.1.6 videos are excellent too. 
  • jlis
    Adam, have you done a video using Pixinsight HDRComposition? I searched and could not find one. I'm interested in trying it, but there are a lot of variables in that script that I don't understand. 

    Regarding an aspect of the OP's original question above regarding integration of various exposure times... I have obtained images on various targets using a ZWO OSC camera (ZWO ASI183 MC Pro, which has small pixels - 2.4 um) and at low Focal Ratio (1.9 using Hyperstar on a Celestron C11).  Depending on conditions (wind, transparency, etc.), I'll shoot at various exposure lengths (currently doing a mosaic of Witches Head Nebula, shooting at 60 seconds and 120 seconds) and at gain=0, to get the "deepest well". I guess it's not ideal to mix shorter and longer exposure times, but if I didn't do the shorter exposures on certain nights, I might as well not even image that night. 

    So, I get much more signal at 120 seconds, but along with that more saturation in the stars (and I think the bright stars tend to be more bloated). 

    Using WBPP (using the latest version), I typically will just integrate all exposures together (I open up exposure tolerance to 60 seconds) and I get decent results. 

    I wonder if I could do better with HDRComp. For example, would I get smaller stars (from short exposures), but keep the high SNR of the longer exposures?  Or, just stick with integrating them all together?  

    Your feedback will be appreciated!
    Thanks,
    John
  • Gordon Aspin
    On nebula targets I use longer exposures to capture the faint nebula details but remove the stars in processing and replace with stars captured with shorter exposures to try to get the best of both worlds.

    Gordon
Sign In or Register to comment.