1. Reminder: Please use our affiliate links for holiday shopping!

Can someone explain this in easy terms about A7II and "isoless" test

Discussion in 'Sony Alpha E-Mount Cameras' started by shaolin95, Jun 8, 2015.

  1. shaolin95

    shaolin95 TalkEmount Top Veteran

    942
    Jul 3, 2013
  2. xXx1

    xXx1 TalkEmount All-Pro

    Jan 15, 2013
    I think he is trying to measure photon noise in dark parts of image. That way you could compare sensors in one important aspect 'what is maximum ISO where dark parts of image don't have too much noise'.
     
  3. robbie36

    robbie36 TalkEmount Veteran

    488
    Nov 21, 2014
    Yes, I understand the concept he is showing and what it means but I certainly dont understand the finer details.

    The idea is when you cant get enough light on your camera, do you raise the 'iso in camera' or 'underexpose' at a lower iso and then 'increase the brightness in post'. The idea is that their are 2 types of noise from your sensor. One is that if your sensor doesnt get enough light then when you 'amplify the light' it increases the pixel noise. The second is known as 'read noise' that is created when converting the light to digital. If you have highish read noise and convert at a low iso, then that read noise will be amplified if you increase the brightness in post.

    So at its simplest his results show....
    1) At iso 800 and above it makes no difference whether you choose a higher iso to expose properly 'in camera' or you 'underexpose' at a lower iso and increase the brightness in post.
    2) Looking at the numbers you could also go further. It really doesnt make a much difference if you 'underexpose' at iso200 or even iso100 and then increase your brightness in post.

    Choosing a lower iso and underexposing also has the advantage that you have more room to protect your highlights.

    So here is a practical example....

    Clipboard Image (142).
    Here is an image (on the left) taken at iso100 but underexposed. On the right is my LR adjustment with exposure increased 3 stops (effectively iso800). Would I have been better off shooting at iso800 in the first place. Well his results show that I would have gained half a stop in the shadows (if this had been a 800iso to iso6400 nothing would have been gained.) But by underexposing I have minimized the extent to which the highlights are blown out (by a factor of 3 stops).

    A secondary conclusion is that ETTR only makes sense at base iso. In other words, it is only worthwhile getting as much light on your sensor as your sensor will allow at its lowest iso. It is rather pointless getting as much light as you can on say iso400 when the same exposure at iso 200 wont make any difference.
     
    • Informative Informative x 3
    • Like Like x 2
  4. shaolin95

    shaolin95 TalkEmount Top Veteran

    942
    Jul 3, 2013
    So at 800 ISO, I am better off protecting highlights and recovering shadow detail than moving to a higher ISO according to his tests and ETTR will only be useful when at 100. That will be the summary, correct? :)
     
    • Agree Agree x 1
  5. Dan Euritt

    Dan Euritt TalkEmount Regular

    191
    Jan 11, 2014
    i was just looking at isoless today, saw that post by jim... looked up a more practical explanation here: http://fujilove.com/isoless-photography-with-the-fujifilm-x-series/

    it was argued at length here, with some people not overly impressed with the results that they saw: http://www.dpreview.com/forums/thread/3515504

    as near as i can tell, isoless processing means tweaking only certain exposure levels in the raw file, vs. lightening or darkening the entire file.. basic stuff that we all do as we gain experience, regardless of how we shoot raw.

    isoless shooting, with the goal that everything iso3200 or lower should result in better pq during processing, means that you lose wysiwg ability in the evf(??), when "setting effect ON" is selected in the menu... the jpg that the evf bases its display on will be underexposed?? if true, that would be a deal killer for me, isoless would only be useful for landscapes and such.

    ev comp might enter into the equation somewhere, but since i've never really tried isoless shooting, this is all speculation.

    what i do know is that turning up the iso in your camera does not change the exposure, all it does is turn up the brightness(gain)... which probably lowers the dr, the latitude range, and even the resolution, if we look at the smoothing effect that comes with cleaning up iso noise.
     
  6. MAubrey

    MAubrey TalkEmount Top Veteran

    The results of ISOless processing are really only useful if the sensor itself is ISOless. Most Sony sensors are or are very close to being so. Some sensors show an improvement at ISO 400 and then they are ISOless from there on up (i.e. there is a noise difference between 100 and 200 pushed to 400, but there is no noise difference from 400 pushed and higher ISO's). And Canon sensors aren't ISOless at all. For them it's because to change the ISO in camera.