HDRComposition

[jeffweiss9]

  I tried to use HDRComposition on 4min and 1min stacked, registered, linear M31 images where the 4min M31 core (plus ~20 stars) were clipped in highlights but the 1min M31 core with only about 3 stars were just ever so slightly clipped.  With the default HDRComposition settings, I got a 17.8bit image but the mask just showed some stars and avoided the core I was trying to improve.  Choosing a preview 'fitting region' around the core didn't get the mask to cover the core.  Lowering the threshold and growing the masks didn't do it either.  Am I misunderstanding what HDRComposition is supposed to do in this situation (i.e. combine different exposure images) or am I just using it incorrectly?
 Thank you.
-Jeff

[Juan Conejero]

Hi Jeff,

You're applying HDRComposition correctly. To help you, I need to take a look at the two images you're trying to combine. Could you please upload them? Thank you.

[vicent_peris]

Hi Jeff,

are you sure you get a burned out galaxy core with just 4 min exposures?? If it's not saturated, you just need to make a good histogram adjustment and apply HDRWavelets to bring out the detail.


Best regards,
Vicent.

[jeffweiss9]

Hi, Juan-
 Files sent via yousendit.com
Thanks,
-Jeff

[Juan Conejero]

Hi Jeff,

Thank you for the images. Vicent has just described what happens in this case: you really don't need a HDR composition, as you have no saturated areas (except a few stars).

To perform an effective (although useless) HDR composition with these images, try lowering the binarizing threshold to 0.02. Then you'll see how the inner core of M31 gets replaced with data from the short exposure. However, this really makes no sense at all, since 0.02 is far from saturation. All you need is, as Vicent has pointed out, a simple histogram stretch followed by HDRWaveletTransform.

If you want to make a true HDR image, then you need much larger exposures where the central regions of M31 should be completely saturated, along with a set of decreasing exposures to cover the saturated areas. Ideally, the shortest exposures should show little more than a star-like galaxy core.

[jeffweiss9]

Vicent, Juan-
 Thanks so much for the response.  You're right that, when I look at the stacked images, the core in neither one is, in fact, saturated, which leaves me bewildered.  When I was collecting the data, I noticed that the galaxy core in the 4min. individual subs was definitely saturated, which is the reason I started collecting the 1-minute data, as I observed in the field that that brought the core pretty much out of saturation on the subs.  So that leads to the question of why the stacked images have no saturation while the 4min subs that went into that stack did?   I must be missing something else because that implies I don't know from looking at individual subs if I have the right exposure to avoid saturation in the final image.
-Jeff
P.S. I've refreshed my memory at the Starizona Optimum Exposure web page (http://starizona.com/acb/ccd/advtheoryexp.aspx) and Craig Stark's SNR Calculator spreadsheet (http://www.cloudynights.com/item.php?item_id=1973) on how to set ideal exposures.  I realize now that sky background flux is the most important factor.   For data from my camera and my m31 subs, including the SNR target object taken as the faintest part of the m31 galaxy that I'd like to bring out, Craig's SNR calculator basically gave a knee of the curve of stacked SNR vs. exposure length that is somewhere in the 2 to 4 minute range.  The Starizona "Ideal Exposure Calculator" gave 2.5 minutes.  So these methods indicate I should have used a subexposure in the 2 to 4 minute range for the sky that I had and my camera.    This, however, doesn't address my question above of understanding why I saw the core saturated in the subs but not in the stack of 20 of the same subs and, more generally, how one sets multiple exposure times to increase the dynamic range of an image.

[Yuriy Toropin]

Hi Juan,
I also have problem using HDR. Files in M31HDR.zip archive (sum of 10 minutes M31 and 2 minutes M31 with DSLR at f/2.8 ) are uploaded (well, in progress now) to my folder on pixinsight ftp site.
Could you please check why I can't combine HDR from those input files?
Thanks a lot in advance,
    Yuriy