• Removed user
    June 18, 2023, 11:41 a.m.

    Download Phil Harvey's ExifToolGUI.

    There are many, many, many EXIF tags, apps do not show them all.

  • June 18, 2023, 11:57 a.m.
  • Members 457 posts
    June 18, 2023, 11:57 a.m.

    The histogram is computed using ISO, WB, etc settings (JPEG), none of them being part of exposure.

  • Removed user
    June 18, 2023, 12:28 p.m.

    again:
    Phil Harvey says:
    This is a Composite tag calculated by ExifTool using the formula:
    LV = 2 * log2(Aperture) - log2(ShutterSpeed) - log2(ISO/100)
    exiftool.org/forum/index.php?topic=7195.0

    Time this whole insane thread was moved elsewhere, IMHO.

  • June 18, 2023, 12:30 p.m.

    That is not applicable to this. That is a description of 'Exposure Value' otherwise known as EV, given by the formula

    EV = log_2 * N^2 / t

    I think that's definitive with respects to how ExifTool calculates it - that is, it's not a value passed through by the camera. But we should think why that is done. The APEX system says
    Ev = Av + Tv = Bv + Sv
    Here, EV is Exposure Value, as above. Av is 'Aperture Value', 2*log_2 N. Tv is Time value, Log_2 1/T, Bv is Luminance Value (sometimes written as Lv) - i.e. the meter reading. SV is the logarithmic exposure index, scaled by an appropriate value.
    So we can reform the second two clauses as Lv = Av + Tv - Sv, which, when you arrange for 'shutter speed' rather than exposure time, comes down to the formula that Phil used there. The idea is that the exposure meter reading provides for a combination of aperture, shutter speed and ISO - so if you want to know what the meter was reading, you can calculate from those. What it's missing is that it should take into account the EC control - otherwise what you get is the meter reading as corrected by EC.

  • June 18, 2023, 12:31 p.m.

    See my reply to your earlier post

  • Removed user
    June 18, 2023, 12:40 p.m.

    I was hoping for a reference that supports your opinion that "Light Value" is scene luminance derived somehow from in-camera metering.

    Without a credible reference, the opinion carries little weight, sad to say.

  • June 18, 2023, 12:46 p.m.

    You can try following the reasoning. I gave you a link to the Wikipedia article, that explains the Bv is Luminance Value or LV. I showed how the APEX equation comes down to the same thing as EXIFtool uses, and that it would give the luminance value, on the assumption that EV was zero.

  • Removed user
    June 18, 2023, 12:51 p.m.

    Just levels, Donald, not "exposure levels".

  • Members 1738 posts
    June 18, 2023, 12:53 p.m.

    The in camera histogram does not measure scene luminance. It’s affected by other things, too.

  • Removed user
    June 18, 2023, 12:58 p.m.

    Ouch!

  • June 18, 2023, 1 p.m.

    Yes, you irritated me more than a little. I rephrased my response a bit. Try reading that.

  • Removed user
    June 18, 2023, 1:20 p.m.

    Tony, could you tell us what ISO was used for the shot and was any Exposure Compensation applied (if so, how much)?

  • Members 976 posts
    June 18, 2023, 1:29 p.m.

    From ExifTool-12.63/lib/Image/ExifTool/Exif.pm

    lines 2229 and 2230:
    # Wikipedia: BrightnessValue = Bv = Av + Tv - Sv
    # ExifTool: LightValue = LV = Av + Tv - Sv + 5 (5 is the Sv for ISO 100 in Exif usage)

    lines 4472 .. 4484:

    LightValue => {
            Notes => q{
                calculated LV = 2 * log2(Aperture) - log2(ShutterSpeed) - log2(ISO/100);
                similar to exposure value but normalized to ISO 100
            },
            Require => {
                0 => 'Aperture',
                1 => 'ShutterSpeed',
                2 => 'ISO',
            },
            ValueConv => 'Image::ExifTool::Exif::CalculateLV($val[0],$val[1],$prt[2])',
            PrintConv => 'sprintf("%.1f",$val)',
        },
    

    lines 5069 .. 5083

    # Calculate LV (Light Value)
    # Inputs: 0) Aperture, 1) ShutterSpeed, 2) ISO
    # Returns: LV value (and converts input values to floating point if necessary)
    sub CalculateLV($$$)
    {
        local $_;
        # do validity checks on arguments
        return undef unless @_ >= 3;
        foreach (@_) {
            return undef unless $_ and /([+-]?(?=\d|\.\d)\d*(\.\d*)?([Ee]([+-]?\d+))?)/ and $1 > 0;
            $_ = $1;    # extract float from any other garbage
        }
        # (A light value of 0 is defined as f/1.0 at 1 second with ISO 100)
        return log($_[0] * $_[0] * 100 / ($_[1] * $_[2])) / log(2);
    }
    
  • Members 300 posts
    June 18, 2023, 1:52 p.m.

    I know apps do show only some tags of all of them. I used exiv2.

    ExifToolGUI is only for Windows or Mac. But I found another tool called exif.

    With it I get one interesting info: Brightness |4,73 EV (91,05 cd/m^2).
    Is that the same value as Light Value in your picture?
    At least it is a metric value which is better to understand than luminance in cd/ft2) in en.wikipedia.org/wiki/APEX_system#Use_of_APEX_values_in_Exif

    Still I can not get FOV or CoC or Hyperfocal Distance, which were interesting things to know too.

    Edit. Iliah posted when I was writing. I guess this is the answer to my question:

  • Removed user
    June 18, 2023, 2:02 p.m.

    Not my picture, but looking around today they are not necessarily the same.

    They are also calculated values and may not necessarily be in all image EXIFs ...

  • Members 206 posts
    June 18, 2023, 2:49 p.m.

    The CoC is an assumption that determines the Hyperfocal distance, so if I use a CoC of .02 mm (which I do) then the hyperfocal distance increases and the DOF decreases. FOV is also based on the format used, which on many cameras is selectable and that is in the EXIF data too, but that doesn't take into account that often photos are cropped versions of the original.

  • Members 206 posts
    June 18, 2023, 4:55 p.m.

    ISO 450

    I may have applied EC, though when using Auto-ISO I usually don't. Since it was in Manual Exposure mode I think the EC doesn't show up in the EXIF data (I can't find it here or in a couple of other files I looked at).