Do you use Active D Lighting on your Nikon camera

13»

Comments

  • ggbutcherggbutcher Posts: 392Member
    edited March 2015
    Two images, first with a normal exposure, the second with ADL-ExtraHigh, aperture-priority, produced from the NEFs with Raw Therapee and its Neutral profile:

    1/10sec, f4:
    image
    1/20sec, f4:
    image

    Now, the raw histograms, same order:

    image
    image

    So, the exposure shift is obvious, based on the shutter speed change. So, are the lower values being compressed, or am I mis-interpreting?
    Post edited by ggbutcher on
  • sevencrossingsevencrossing Posts: 2,800Member
    edited March 2015
    Like it or not, it is a documented feature that ADL lowers the exposure value, as much as a full stop depending on circumstances. That is irreversible.
    Well I have tried to make ADL change the exposure on my D800 and for the life of me, I can't
    It changes the histogram and the the jpeg preview but the exposure remains exactly the same
    Post edited by sevencrossing on
  • IronheartIronheart Posts: 3,017Moderator
    What mode are you in? If M it will only change the meter, not the exposure (unless you have auto ISO enabled)
  • WestEndFotoWestEndFoto Posts: 3,745Member
    After reading this, it occurred to me that ADL was on, and since I proudly have a prime raw fetish, I turned it off.
  • sevencrossingsevencrossing Posts: 2,800Member
    Aperture priority ISO fixed @ 100
  • WestEndFotoWestEndFoto Posts: 3,745Member
    I don't quite have your ISO fetish, but will go to a huge effort to keep it below 1,000. I shoot about 95% with aperture priority. The rest is manual and shutter priority. Still haven't figured out what "P" does yet.
  • spraynprayspraynpray Posts: 6,545Moderator
    edited March 2015
    This morning I tried ADL bracketing with my D750. Manual mode - no auto ISO - all five shots displayed the same on my PC. Aperture Priority - still no auto ISO - the five shots each displayed darker as the ADL setting got higher. Interestingly, the shutter speed increases as the setting goes up so it is a change to the exposure before it is shot rather than after as Ironheart said all along.

    I really don't like what it did to the shots. I'll give it a go in a really high contrast sunny day scenario, then if I don't like that either, it stays off as I have a fetish about weird s#it being done to my images.

    This has been an interesting and useful discussion - Thanks guys ( @Ironheart @PitchBlack ).
    Post edited by spraynpray on
    Always learning.
  • donaldejosedonaldejose Posts: 3,692Member
    "change to the exposure before it is shot rather than after as Ironheart said all along"

    Isn't it both when shooting a jpg image? Before the shot the exposure is reduced so as to not wash out highlights and after that slightly modified exposure the shadows are boosted or recovered a bit as the camera makes a jpg out of the raw data.

    When shooting raw all you get is the reduced exposure which is reduced more as you move from ADL low to ADL high.

    When shooting jpgs the amount of exposure reduction and shadow recovery increases as you move from ADL low to ADL high.

    At least that is how I understood it to work.

    So I don't see a benefit to the reduced exposure when shooting raw. Why not just use your exposure compensation setting if you want to reduce exposure a bit? I only see a benefit to ADL when shooting jpgs because it allows for a small bit of "post processing" automatically in camera just as picture contols also do.
  • ggbutcherggbutcher Posts: 392Member
    So, I'm still not completely clear on where step #2, shadow recovery, is being done. On my rube goldberg test documented above, it would appear its being done somewhere around the AD converter on the sensor, as I believe my raw histogram represents the measurements taken there. If so, ADL munging in total is in your NEF. That would also imply that some of your pixels are sampled at a higher ISO than others in order to achieve the effect.

    So, I think I"m going to put it in my U2 and 'speriment with it in high-contrast situations. But for the majority of my shooting, it'll be off.
  • spraynprayspraynpray Posts: 6,545Moderator
    "change to the exposure before it is shot rather than after as Ironheart said all along"

    Isn't it both when shooting a jpg image? Before the shot the exposure is reduced so as to not wash out highlights and after that slightly modified exposure the shadows are boosted or recovered a bit as the camera makes a jpg out of the raw data.

    When shooting raw all you get is the reduced exposure which is reduced more as you move from ADL low to ADL high.

    When shooting jpgs the amount of exposure reduction and shadow recovery increases as you move from ADL low to ADL high.

    At least that is how I understood it to work.

    So I don't see a benefit to the reduced exposure when shooting raw. Why not just use your exposure compensation setting if you want to reduce exposure a bit? I only see a benefit to ADL when shooting jpgs because it allows for a small bit of "post processing" automatically in camera just as picture contols also do.
    Yes, it probably is as you describe Donald, but I was surprised to learn it actually changed the exposure before the shot - I just thought it worked using the latitude within the RAW file to achieve the results.
    Always learning.
  • ggbutcherggbutcher Posts: 392Member
    I hate not knowing specifically how something works, so I went digging. Of great interest was this flickr post from about 6 years ago on technology Nikon is supposed to have procured from a UK company called Apical, generically called Adaptive Digital ISO, "which tries to selectively boost ISO within localised sections of the sensor." Here's a link to the post:

    https://www.flickr.com/groups/479891@N21/discuss/72157611112048962/72157611149547992

    Now, this is just Flickr-talk, but it sounds reasonable. I visited Apical's website, and they do such things. And, if Nikon is using their engines in image capture, the histogram behavior above bears it out, if I've got the histogram set right. I'm probably going to do a dcraw dump of the unmosaiced data and see if I can build histograms from that, just to know my transform tools...

    Nikon's reference literature is disappointing here. They really should be more specific about the mechanism so those who shoot raw can make a proper decision about using the tool. About the only hint given is this note:
    "Noise (randomly-spaced bright pixels, fog, or lines) may appear in photographs taken with Active D-Lighting. Uneven shading may be visible with some subjects. Active D-Lighting can not be used at ISO sensitivities of Hi 0.3 or above." (ref. D7000 manual)
    If the shadow compression were being done after the image capture, noise would be uniform across the image consistent with the set ISO.
  • IronheartIronheart Posts: 3,017Moderator
    I knew it dropped the exposure a few notches in everything but full manual, where obviously all it can do is drop the meter so you end up setting a lower exposure (just like exposure compensation). I didn't think it affected the raw file beyond that. However from @ggbutcher's analysis it looks like it might be doing something even more. Tone mapping, selective ISO, who knows? I'm shooting a few tests tonight.
  • spraynprayspraynpray Posts: 6,545Moderator
    Love it. We should write Manuals for Nikon (maybe replace the marketing department too). :P
    Always learning.
Sign In or Register to comment.