exposure differences x10-x20: what could be the explanation?

Discussion in 'Fuji X10, X20, X30, XF1, and X-S1' started by pniev, Mar 26, 2013.

  1. pniev

    pniev Guest

    I like the X20 (at least as much as the X10) but there is one thing that keeps puzzling me... The exposure differences of both cameras. After some (unstructured) testing in 12MP/A-priority/Auto-ISO, I am still puzzled. For a moment I thought it has to do with the DR auto mode: when I use DR auto on both cameras, the X10 exposes a full stop (sometimes 2/3 stop) faster than the X20 by adapting the ISO value (both cameras choose DR100, which makes sense of you shoot the same object, both cameras show same shutter speed). When I use DR100, sometimes I get a 2/3 stop difference and sometimes a 1/3 difference. I tried this about 20-30 times so there is some consistency.

    BTW: to get good results I consistently need to push the exposure of the X20.

    Could there be an explanation other than a camera hardware issue? Did I overlook something?

    Any help is much appreciated!

  2. depscribe

    depscribe FujiXspot Regular

    Feb 17, 2013
    Not to sound too much, I hope, like a broken record, but Fuji has a habit of shipping cameras before the firmware is quite done. This happened with the X100 (which they never did fully fix, as to, for instance, manual focus, suggesting that the issue might be beyond firmware; yes, you can kludge around it but that doesn't negate the fact that it never has worked quite correctly), the X10, and, in my estimation, the X20. I hope the X10 is not the template, or we will have several months of Fuji denying that there is a problem, then a fix that fixes nothing, then a fix that does solve the problem(s). Of which I think there are several, though others disagree. I spoke with Fuji USA yesterday, who said they had not actually seen X20s yet, "but we've been hearing some things." So.
  3. flysurfer

    flysurfer X-Pert

    Feb 1, 2013
    Rico Pfirstinger
    Are you talking about exposure measurement differences? Sure, since Fuji has improved Multi Metering in the X100S and X20, there have to be differences. Without differences, there wouldn't be any improvement. Maybe I misunderstood the question?
  4. pniev

    pniev Guest

    That has pointed me in the right direction! Good to know that Fuji changed the multimetering system. Just tried spot metering and exposure is comparable. If the differences are resulting from changes in the multimetering system, I am not sure if it is an improvement (yet). The results of the x10 do look better if I don't push the exposure of the X20. I will do some testing outside the house too.
  5. flysurfer

    flysurfer X-Pert

    Feb 1, 2013
    Rico Pfirstinger
    Fuji changed Multimetering in existing cameras through firmware udates, too. They just didn't officially say so. But I know it, because I often have two bodys of a model for testing purposes, and when I do a firmware upgrade, I first only upgrade only one and then compare both cameras for different behavior. ;)
  6. flysurfer

    flysurfer X-Pert

    Feb 1, 2013
    Rico Pfirstinger
    Forums are full of often strange discussions about "inflated ISO". Strange because these discussions usually conclude that Sony, Nikon and everybody else are competent, capable and honest about ISO, and Fuji is not (as in incompetent, dishonest etc.).

    The funny aspect of this discussion is the fact that both Sony and Nikon (and several other makers) are using Apical Iridix as their black box DR expansion method, and this solution is changing exposure (depending on its settings) w/o adapting ISO in the EXIF data accordingly. So it's a fact that Sony and Nikon are liars when it comes to correctly representing ISO (while Fuji and others dutifully change their ISO settings along with DR settings), but nobody cares because a) Sony and Nikon are perfect and can do no wrong, and b) the matter is somewhat complicated and requires a few hours of research to fully grasp it, which doesn't run well with c) the fact that most users prefer obvious culprits they can blame. ;)

    To compare the real exposure of 2 different camera bodies you need to set them both to M mode and DR100%, set identical ISO, aperture and shutter speed values, set the same focal distance (yep, that makes a difference), use the same focal length (yep, that makes a difference, too) and (of course) use the exact same lens without any filters, protection glass or hoods. And, of course, you have to shoot the very same subject from the very same distance at the very same angle in the very same light.

    But even then, you will often see differences in brightness between two cameras. Weird, huh? However, a deviation of 1/3 EV from the actual brightness is considered normal and acceptable. So if you happen to have 2 cameras, one with a deviation of -1/3 EV and one with +1/3 EV, both cameras will still be within specs, but the images you compare will show a difference of 2/3 EV.

    There's more, as the deviation can change depending on the ISO that has been set. So it may be different at ISO 200 and at ISO 6400. That's because at higher ISOs, internal processing comes into play: The images are pushed by analog and digital means. Some cameras may push a bit less in high-ISO situations than would be mathematically required – probably in order to preserve highlights (as you know, DR tends to decrease with increasing ISO signal amplification). This is all firmware (or metadata in the RAW file, telling external RAW converters what to do exactly, hoping the external converter fully supports those commands).

    Finally, each sensor is different, and official ISO declarations are just ballpark numbers. For example, dpreview found in their review that the X-Pro1's sensor is more of an ISO 160 sensor (not ISO 200), just like Fuji's Velvia film was more of an ISO 40 film (not ISO 50 as it said on the packaging).
  7. speedandstyle

    speedandstyle FujiXspot Regular

    Thanks for this info! Well written in an easy to understand way and without getting too long.
  8. pniev

    pniev Guest

  9. carlb

    carlb FujiXspot Veteran

    Feb 6, 2013
    twin cities, minnesota
    Hey Peter, I'm noticing seemingly underexposed jpegs when using DR200 or DR400 on the X20. I've been assuming it's the camera underexposing to allow for the additional headroom we ask for with DR200. The X10 wouldn't do this (I'm guessing) because of its 6MP extended dynamic range approach.

    Just a working theory based on what I've read and experienced with the X20.

    I'm not worried about these underexposed jpegs, I push them to the right in post via a combination of highlight push, ambient push, masking, etc. No noise problems pushing the dark areas as needed, so no problem for me.

    It may well be a bit of a firmware burp, but nothing to beat on the tables about.
  10. nippa

    nippa FujiXspot Regular

    Feb 26, 2013
    Cheshire UK
    There certainly seems to be a fair degree of underexposure during the 24 hours I've had the camera ; most noticeably in low light.
    So far I've only used Multi Segment although on one occasion I used Spot Metering to correct things and got the same result.
    Still climbing the learning curve :)