Fade shadows/highlights? Or leave blacks and whites at 100%?

Discussion in 'Open Discussion' started by noahtruth, Jun 11, 2014.

  1. noahtruth

    noahtruth Mu-43 Rookie

    15
    Jun 11, 2014
    I have been browsing various online catalogs and obviously everyone and their mom is using VSCO presets, which tend to add light or heavy fade.

    My question is, even when not using presets and simply making edits manually, is it still a good idea to use subtle fade for aesthetic reasons?

    Also, is it a good practice to fade both shadows and highlights together? Or do you sometimes fade just shadows/blacks and leave highlights/whites alone?

    I was editing a shoot last night and had the hardest time trying to decide whether or not to leave my blacks at 100%

    What do you all think?
     
  2. My idea of an ideal histogram (if such a thing existed) is one that stops just short of the white point and gives the black point a small nudge.
     
  3. jamesgehrt

    jamesgehrt Mu-43 Regular

    166
    May 20, 2014
    Easthampton, Massachusetts
    James Gehrt
    I go by the numbers. Just as in scanning negatives. I aim for a shadow value of 12 and a highlight of 242. I then tweak the mid-tones with a curve adjustment. After that a bit of burning and dodging to taste. This allows for a bit of a buffer and I won't blow out my highlights or block up the shadows. I try to expose to make sure I am not loosing anything with the histogram and then adjust the curve in Adobe camera raw, to fix my highlights if needed.
     
  4. lightmonkey

    lightmonkey Mu-43 Veteran

    480
    Dec 22, 2013
    some pictures suit it, some dont. some i do the technically perfect exposure (scale picture to blackest and whitest point in photo). some i intentionally mask or blow out. some i lift blacks to intentionally reduce contrast.

    its like treating beef. some you can eat raw (tartare), some you barely cook (steak), some you boil or smoke to death (stew / bbq). no single answer
     
    • Like Like x 1
  5. ean10775

    ean10775 Mu-43 All-Pro

    Jan 31, 2011
    Cleveland, Ohio
    Eric
    I tend to fade the shadows a bit simply because I like the look, but I rarely if ever fade highlights.
     
  6. noahtruth

    noahtruth Mu-43 Rookie

    15
    Jun 11, 2014
    This may be a silly question, but how did you decide on these specific numbers?
     
  7. noahtruth

    noahtruth Mu-43 Rookie

    15
    Jun 11, 2014
    I agree with your points about it being subjective but can you elaborate on this statement? What exactly do you mean by scaling to "blackest and whitest points?" And how exactly do you achieve this in LR?
     
  8. kevinparis

    kevinparis Cantankerous Scotsman

    Feb 12, 2010
    Gent, Belgium
    I have to ask.... what are you talking about?..... what do you mean by 'fading' highlights and shadows.... genuinely have never come across this phrase

    cheers

    K
     
  9. jamesgehrt

    jamesgehrt Mu-43 Regular

    166
    May 20, 2014
    Easthampton, Massachusetts
    James Gehrt
    Sorry, yes. Not a silly question at all. These are values set by the Federal Agencies Digitization Initiative. (FADGI)

    A value of 0 is pure black and a value of 255 is pure white. 12 and 242 represent the same value that would have been the transparency value when using film and photographic paper. I don't remember the scale for transmission or reflection on a densitometer. However, these are kind of like the zone system in traditional black and white. Now, instead of 11 zones we have 255. I think Ansel would be proud! :smile:

    In my work, I like to have a nice clean bright highlight with detail and I especially love dark shadows that you can get lost in.I tend to find a process that works for me and I stick with it. Be it a film and developer combo or pixels and curves. I am trying to manipulate and control the tonal range to express how I visualized the image to feel. It might be a bit boring, but I like things predictable. Digital has given me the control that I have always dreamed of.
     
  10. lightmonkey

    lightmonkey Mu-43 Veteran

    480
    Dec 22, 2013
    In PS you can use the eyedropper to click on the pixels that represent the 2 end points you'd like to scale the picture too.

    In lightroom you can hold down Alt (or Ctrl?) While controlling the sliders. The hot pixels will indicate those that exceed the 0/255 scale
     
  11. noahtruth

    noahtruth Mu-43 Rookie

    15
    Jun 11, 2014

    Probably a poor choice of words. I mean using values greater than true black (0) and less than true white (255).
     
    • Like Like x 1
  12. noahtruth

    noahtruth Mu-43 Rookie

    15
    Jun 11, 2014
    Ah, gotcha. Thanks so much for this detailed response.
     
  13. noahtruth

    noahtruth Mu-43 Rookie

    15
    Jun 11, 2014
    Okay, that's what I figured. Thanks for clarifying.
     
  14. RobWatson

    RobWatson Mu-43 Hall of Famer

    Too much shadow at black (zero) just looks weird to my eye.

    I wouldnt be surprised if some of my favorite images have shadows at black because it works in the context of that image. Guess one can get away with things depending on circumstance.
     
    • Like Like x 1
  15. MajorMagee

    MajorMagee Mu-43 All-Pro

    Feb 24, 2011
    Dayton, OH
    I wasn't confused about what you meant before, but now I am. So when you "Fade" are you taking everything from 12 and below and making it all 0, or 12? Is everything 242 and above becoming 242 or 255?
     
  16. David A

    David A Mu-43 All-Pro

    Sep 30, 2011
    Brisbane, Australia
    Let's go back to basics.

    Files store luminance data, the data about how bright a pixel is, with a fixed number of bits depending on the file format. JPEGs use 8 bits per colour channel and an 8 bit file records that brightness data in terms of a range of 256 steps from 0, blackest black, to 255, brightest white. Other file formats use more or less bits. If you shoot JPEGs in camera, your straight out of camera files will have 256 steps per colour channel. If you shoot RAW your files will have 4096 steps per colour channel for a 12 bit image and even more if the file uses more bits than 12.

    So, when you process your image you effectively tell your computer how you want the finished image to look by adjusting things like contrast, black and white points, and other controls like your highlights and shadows sliders in Lightroom. Even adjusting the exposure sliding changes the brightness of pixels in the processed file your software application produces. What you are doing is effectively telling the computer at what value in the processed file you want it to display a given pixel in the file the camera generated. It's become a convention for processing applications to give numerical values for brightness levels in terms of the 256 steps that exist in 8 bit images, even when the processed file uses more than 8 bits and contains more than 256 levels of brightness per channel.

    You have control about how the application treats the levels from the camera's file. You can raise or lower those levels in the processed file. If you push some levels lower or higher, you end up compressing or expanding the range of values a certain tonal range in the camera's file has in the processed file.

    So, if you push some values down to 0, say values of 16 and below in the camera file down to 0 in the processed file, and you also raise some values, say from 239 in the camera file, up to 255, what you're doing is "stretching" the 223 step range from 16 to 239 in the camera file to a range from 1 to 254 in the processed file. The opposite happens if you raise the lowest values in the camera file and lower the highest values. What you end up with is a range of values in the camera file being "compressed" into a smaller range in the processed file.

    What you can't do is push any values to below 0/0% or raise them above 255/100% because those values define the lowest and highest luminance value the file format can record, There are no blacks lower than 0 or whites higher than 255 simply because those numbers are the lowest and highest values of the scale being used to display brightness levels. You can raise or lower the brightness value of a pixel in the camera file to another value in the processed file but the value in the processed file still has to fall between 0 and 255 or 0% and 100% because that's the only values available for recording data in the processed file.

    So, what your question is really about is whether or not you should take advantage of the maximum range available for recording brightness when you process a file from your camera. A lot of the time the answer is yes, especially for scenes which had a wide brightness range. On the other hand there are scenes, like say a scene in fog, where the original scene had a very low brightness range with no really dark areas and no really bright areas. You could process the file from the camera so that the darkest areas are lowered to a value of 0 in the processed file and the brightest areas raised to a value of 255, but then your darkest areas are going to appear darker than they were in the original scene and your brightest areas will appear brighter than they were in the original scene. That may actually look good, and you may want to do that for artistic purposes, but it may look bad and you may want to use a much smaller range of values for the processed image than 256 steps in which case you most certainly don't want to push some values down to 0 and some up to 255.

    When you're processing your files, adjust the value of the darkest parts of the image and the brightest parts of the image so that they are as dark/as bright as needs be to make the image look the way you like. In the process you'll also be adjusting the brightness values for the areas of the image between the darkest and the brightest parts. You don't have to make the darkest parts pure black and the brightest parts pure white. Sometimes that's good, sometimes it's bad. Don't worry about what the numerical values your processing application displays are, worry about how the image looks on your display and make it look the way you want it to look. All the numbers are is a readout of the values used to store the data for the image. If the image looks the way you want it to look, the numbers will be "right". If you try processing by ensuring that there's always some pixels with a value of 0 and some with a value of 255 then a lot of the time you will end up with a file with data numbers that look the way you thnk the numbers should look but the picture is not going to look the way you want the picture to look. That's not the outcome you're trying to create.
     
    • Like Like x 4
  17. jamesgehrt

    jamesgehrt Mu-43 Regular

    166
    May 20, 2014
    Easthampton, Massachusetts
    James Gehrt
    Great explanation David A! Well said.
     
  18. noahtruth

    noahtruth Mu-43 Rookie

    15
    Jun 11, 2014
    Wow! Thanks so much David. You pretty much gave me all the information I needed. Your detailed answer is much appreciated.