1. Welcome to NikonCafe.com—a friendly Nikon camera & photography discussion forum!

    If you are thinking of buying a camera or need help with your photos, you will find our forum members full of advice! Click here to join for free!

Ersatz HDR via 14-bit and Active-D??

Discussion in 'Nikon DX DSLR Forum' started by Harry Lavo, Aug 15, 2008.

  1. While out shooting this week I usd both 14bit and Active-D for the first time (I had done some backyard trials previously). I was shooting late in the day in an orchard, and for one shot I felt I need all the dynamic range I could get. So, NEF set to 14 bit and Active-D set to High, and here is the result:

    [​IMG]

    My take: the 14 bits captures a much expanded exposure range, and the Active-D manages to reduce it to the limited bits available for print or monitor in a way that far exceeds ordinary exposure. Not really HDR, but IMO much better exposure than otherwise would be possible without these technologies.

    Your thoughts?
     
  2. Panopeeper

    Panopeeper Guest

    The captured DR does not depend on the bit depth.

    ADL does not do anything to the raw image data. If you don't believe it, pls post two raw files shot with the same exposure, from tripod, same illumination, one with ADL on, the other off.
     
  3. I stand corrected on the facts....I'm just beginning to explore HDR and I apologize. But I'm not sure it changes my conclusion....the high bit rate allows many more nuanced choices of color and Active-D shapes the lighting to better control what the eye sees as shadow detail and brightness "finesse" (for want of a better term). I really don't care what it does to the raw file...I do care what it does when that raw file is processed with NX.

    With both these in operation , it seems to me one "sees" seemingly more defined "brights" and more shadow detail than otherwise is the case, all when reduced to the screen or paper dynamic range. There seems to be a synergy between the two that neither shows to the same degree on their own.
     
  4. adrianaitken

    adrianaitken Guest

    Panopeeper - I think it was on this forum (or another!!) , ADL does effect RAW files. It seems to basically underexpose the shot by a stop or so (I seem to remember). You can pull more detail out of higlights than shadows so it exposes for the shadows and thereby gives a seemingly wider DR.
     
  5. Panopeeper

    Panopeeper Guest

    That's ok, but one has to be aware of the fact, that this is not coming from the camera but from the raw processing. The same could be done with any other camera through suitable raw conversion adjustment.

    This is how Canon's Hightlight Tone Protection works, and I though that's happening with the D3 and D300 as well.

    I don't have either of these cameras, I can not test if for myself. However, I got directly comparable raw images shot with and w/o ADL and I did not find any difference in the exposure level.

    This contradicts the other claim because lower exposure is for the highlights, on the cost of shadows.

    Plus, reducing the exposure by one full stop reduces the DR a lot, maybe even by a full stop.

    If you upload two raw images shot as I described above, I can verify what is happening - assumed, that the image pair I received earlier is somehow not truly representative (I could not explain that).

    "Saving the highlights" and "recovery" are no obvious terms. In many cases it can be achieved by a different mapping of the highlights.

    The issue is, that the contrast is very low in the very brightest region, due to the non-linear mapping.

    Let's take an example with the mapping of sRGB, 14bit depth, without any additional adjustment to the mapping curve. In the ninths stop, in the very dark shadows groups of four-six raw levels are mapped on a single sRGB level of 8bit depth. In the third stop (measured from the top) gropus of 40-60 raw levels are mapped on a single sRGB value.

    However, at the bright end, in the last stop, up to 140 raw levels are mapped on one sRGB value. This way the contrast is eliminated, everything appears to be white, like burned out.

    Changing the mapping function (the "curve"), the contrast can be "recovered" (nothing was lost!), and this is what ADL is about.

    Combining this with a one stop underexposure, like my camera does is a crude method, useful only for JPEG out of camera. IMO HTP should never be used when recording raw data. If I am afraid of overexposure, I can use a bias of -1/3 or -1/2 stop, it does not have to be a full stop.

    A crude underexposure is detrimental to the DR.
     
  1. This site uses cookies to help personalise content, tailor your experience and to keep you logged in if you register.
    By continuing to use this site, you are consenting to our use of cookies.