• Members 184 posts
    Oct. 24, 2024, 9:17 p.m.

    It I understand it correctly, the method you described is what many people use as their default method of capturing HDR.

    In fact it doesn't guarantee you don't blow the highlights, although the chances of that will be reduced. What you describe as "about right for main part of the picture" sounds quite arbitrary, and going -2ev from midtones often means you get a heavily underexposed shot with little to zero additional information, but increased chances of ghosting.

    I almost always use ETTR in landscape shooting. By histogram, I can see if the scene fits the DR of my camera, which is, say, 80-90% of cases.

    When it doesn't, I take just two shots: ETTR and ETTR +2ev. This is enough for HDR blending in 99% of cases. It's much easier to blend just two images instead of three - less chances for ghosting.

    In very high contrast cases (e.g. the sun in the frame) I'd take a third shot ETTR +4ev.

  • Members 4254 posts
    Oct. 25, 2024, 8 p.m.

    Yes, that is spot on.

    Most of the time only 2 shots are required to capture the full dynamic range as you described.

    There is no need to underexpose the highlights by 2 stops, or whatever, since they are not being clipped when using ETTR.

  • Members 725 posts
    Oct. 26, 2024, 9:14 p.m.

    I would be interested in how shooting with a negative EC increases chances of ghosting, assuming a faster shutter speed.

    Perhaps "ghosting" has a different meaning here ...

  • Members 184 posts
    Oct. 26, 2024, 10:23 p.m.

    Any additional image used for HDR merge increases chances of ghosting, no matter how fast the shutter speed was.

    The method described by NCV involves taking 3 shots for HDR merge: some arbitrary exposure that looks right for midtones (call it 0ev), then -2ev and +2ev.

    But note that if the highlights aren't blown at 0ev, there's no actual need to take a -2ev shot because it has very little additional information, but it increases chances of ghosting if used in an HDR merge.

    On the other hand, if the highlights are blown at 0ev, there's no guarantee they won't be blown at -2ev.

  • Members 725 posts
    Oct. 27, 2024, 2:23 a.m.

    OK, that kind of ghosting, thank you.

    Personally, if I merge a stack, I align the images first - so as to avoid "ghosting". I'm a bit surprised that others don't bother with that.

    Um ... "ev" stands for electron-volts ... pardon the pedantry ...

  • Members 184 posts
    Oct. 27, 2024, 5:56 a.m.

    In this context, EV stands for "exposure value": en.wikipedia.org/wiki/Exposure_value

    Ghosting in HDR merge occurs when there's movement in objects in the scene between the shots, e.g. foliage. Obviously the frames must be aligned before merging, and HDR software (e.g. Lightroom) does the alignment automatically, but that doesn't help to fix this kind of ghosting.

  • Members 1963 posts
    Oct. 27, 2024, 9:54 a.m.

    Capture One seems to eliminate ghosting. My landscape shot was shot on a windy day, and has zero ghosting.

  • Members 4254 posts
    Oct. 27, 2024, 10:16 a.m.

    If you're happy with your photos that's fine. No-one is telling you how to do HDR images.

    The point being made in this thread is that for those aiming to maximise the quality of the raw data for a HDR scene before merging, ETTR is the way to do it.

    I see no point in setting -2EV compensation for highlights in situations where they are not being clipped at 0EV compensation. Applying negative exposure compensation when it isn't necessary just reduces the quality of the raw data.

    As Quarkcharmed correctly pointed out, in most cases you need only 2 shots to capture the dynamic range, not 3.

  • Oct. 27, 2024, 12:37 p.m.

    It may depend on your source images and software used. If you can base all your adjustments on raw data and it is linear, then yes, there is no point in underexposing. If any part of image pipeline (starting from sensor readout) makes nonlinear corrections [for highlights], then underexposed images may come handy for restoring better dynamic or color for highlights.
    Like when your raw developer software itself cannot perform local tone mapping (or how the HDR image creation process is called), then you may not have full control over linearity of data. For your Adobe products this likely is not the case, but for some other tools it may be.

  • Members 184 posts
    Oct. 27, 2024, 10:49 p.m.

    The data in raw files is almost always linear. HDR merging/blending typically involves applying algorithmic models that 'recover' the original per-pixel illuminance and then converting it back to digital numbers as if it was a single shot. Underexposed shots may or may not add noise depending on the algorithm.

    Tone mapping happens after the HDR merge. It happens all the time even without HDR.

  • Members 725 posts
    Oct. 29, 2024, 12:21 a.m.

    One of my favorite links ... nowhere does it say "ev". It does allow "Ev" though.

    It's a pity that grammar is going downhill fast as more and more people type with their thumbs and do not give a mojón about correct capitalization.

    now where did i put my bayer cfa nikon d500 ...

  • Members 184 posts
    Oct. 29, 2024, 12:55 a.m.

    If you're that picky about capitalisation, then electron volt is not 'ev' but 'eV'.

  • Members 725 posts
    Oct. 29, 2024, 1 a.m.

    Ouch, I stand corrected! 😠

    My general comment about grammar still stands.

  • Members 399 posts
    Oct. 29, 2024, 12:43 p.m.

    This is an international forum. Different languages have different grammatical rules for capitalization. For example in German ALL nouns are capitalized. Russian has similar rules as English with an important exception, the names of months and days of the week are NOT capitalized. In fact there are even minor differences between US English and British English when it comes to capitalization. So it is not necessary a problem with poor grammar, but more of one often people are using their second or third language to post to this all English forum.

    As far as "eV", when used in physics "eV" is a symbol for a unit of the measure of a physical property used in some (but not all) areas of physics. While eV is a unit of energy used in say particle physics, it is not an SI unit (basic nor derived).

  • Members 725 posts
    Oct. 29, 2024, 4:11 p.m.

    As to grammar, Truman, I was mainly commenting about people typing with their thumbs and not bothering to capitalize for the most part. That is bad grammar in any language.

  • Oct. 29, 2024, 5:30 p.m.
  • Members 725 posts
    Oct. 29, 2024, 7:22 p.m.

    Wow, never heard of that one!

  • Members 549 posts
    Nov. 11, 2024, 3:09 p.m.

    Thanks.

    It was a simple binary typo, probably because I started out with "noise" and then changed it to "SNR", which, of course, is a PITA if you are behind in following a thread and don't notice the catch until 19 days later, and a whole threadstorm snowballs in your absence.

  • Members 549 posts
    Nov. 11, 2024, 3:21 p.m.

    It's a lot easier in the long run to simply qualify the nature of the stacks, so there is no question.

  • Members 549 posts
    Nov. 11, 2024, 3:30 p.m.

    Yes, you can get ruined on either end of the raw histogram, but the highlight end is quite drastic, making every would-be higher value record at the clipping point, whereas, at the black end, there is no actual clipping, as black is an asymptote, not a passable value, and decreasing exposure mainly increases noise, but can still show larger image features when it can no longer show fine ones because of too much noise.