Poor display calibration results since ML install.

Discussion in 'MacBook Pro' started by hifirobbie, Jul 26, 2012.

  1. hifirobbie macrumors newbie

    Nov 24, 2011
    Ok, until last night I was able to calibrate my early 2011 MPB to 96% sRGB using a Datacolor Spyder4 Elite system. This system works by modifying the output signal from the video card to achieve display accuracy which is in turn, tested with the supplied Datacolor Colorimeter.

    So far today when I tested it, the closest I can get is 70%.

    This is potentially a MASSIVE issue for me as I work in photography and am often called to edit photos on site, using the MBP display.:mad:

    Does anyone know of any changes they may have dome as far as video output profiles etc.?
  2. ixodes macrumors 601


    Jan 11, 2012
    Pacific Coast, USA
    All of the early adopters at work, were dealing with this today. We use the exact same equipment. While it's too soon to give you help, I'm not concerned.

    The temporary solution we employed, was to break out the new spares (conventional 2011 MBP's) we keep for emergencies, so these guys could get back to work. It reminded me, of how valuable having those on hand can be.

    I'm sure at some point, Apple will get around to addressing this. Hopefully that's not just my optimism speaking :)
  3. hifirobbie thread starter macrumors newbie

    Nov 24, 2011
    Thanks for the reply.

    My pulse has returned to the sub 200bpm realms now, knowing it's not just me.

    Here's hoping they get it sorted soonand it's not just another one of their "Oh, yeah, you'll have to upgrade now because that machine is obsolete as a result of Mountain Lion".

    In the interim, this sucks.

    Fortunately I have an external monitor.

  4. thekev macrumors 604


    Aug 5, 2010
    If you use this for work, you should absolutely never upgrade on Day 1. Wait for the bugs to be worked out. Being productive means that you want a stable environment, not the latest thing. I'd suggest this with Windows or OSX. Stick with something stable. Test the latest thing after the initial bug fixes. Maintain a backup in case you need to revert. Also notebook displays aren't entirely ideal for viewing photos. They're usable, just not ideal.

    Anyway on to the problem. When you calibrate in Datacolor's software, it should strip the profile and use an unmodified behavior. It would feed a generic set of numbers to the framebuffer to measure the behavior of the hardware rather than start from the prior working profile. It's unlikely that you can do much beyond post in their forum to alert them of the bug. Others may have already reported it. A fix will eventually be issued. Datacolor is pretty good about that. My issues have always been more with the long term stability of their hardware, although I've had X-rite problems too.
  5. hifirobbie thread starter macrumors newbie

    Nov 24, 2011
    Yeah 20/20 hindsight. Life goes on.

    First early adopter OS upgrade since moving to Mac.

    Overall, do you folks find the Spyder fairly reliable?

    Maybe I'm just too used to having my retinas scorched by overly bright and saturated displays in the past, but even at 6500/2.2/120 on my ASUS PA246Q, I can't help but feel that it looks rather flat when calibrated.


  6. thekev macrumors 604


    Aug 5, 2010
    Well the drop in gamut is pretty significant, and that shouldn't happen due to a change in OS revision. My past experience with Spyders was that they were too inconsistent between measurements and had a tendency to drift over time or break. The new ones have been quite popular though, and what I've read suggests that they've improved with each revision. I use an X-rite colorimeter. In the $200 realm there's only so much you can expect in terms of manufacturing quality and consistency.

    Regarding your Asus, displays are all optimized differently. Some of them hold up at lower brightness levels better than others. Your experience may vary based on ambient lighting in your work area and the display's actual response. Once it's warmed up does it compare favorably with a print? My question is far from scientific. I was just wondering if it looked dull relative to the same thing printed. If you're viewing a consumer grade display jacked up to maximum brightness, that's just not a stable target. I'd encourage a stable target so that your visual adjustments remain consistent over time.

    Anyway I didn't mean to chastise you over your choice to upgrade, but if it's just not working out, I suggest you revert your OS until things are smoothed out. It's not worth messing up your workflow. You can also consult Datacolor for advice. Typically for professional use, I'd suggest that at the very least, you retain a backup of the older OS. Ideally when you sit down to work, you want to work rather than beta test. People complain about this with Windows, yet day 1 is bad no matter what OS.
  7. hifirobbie thread starter macrumors newbie

    Nov 24, 2011
    Completely agree!

    I have worked in custom home electronics for years previously and for the last 10 years, that's exactly what we, the consumers, have been:

    BETA testers.

    It used to be unfathomable that so many expensive products were released in to the market with so many potential flaws, but now they seem to have been able to steer us in their direction and it's for the most part, expected.
  8. thekev macrumors 604


    Aug 5, 2010
    It's annoying isn't it? There are a lot of problems with limited testing. There's just no way to imagine every possible bug prone scenario. When you release something to the general public, you can collect data from a pool of millions of users. Unfortunately if this is something critical for your work, stability is a prime factor even more than absolute accuracy.

    I comment on displays and calibration quite a bit on here, and I mention stability and uniformity as major factors. Manufacturers play with a lot of specs. Describing gamut as a percentage is fairly laughable, and details about its volume don't really tell you how well it works given the number of addressable values or how well it actually matches a given reference gamut in three dimensions. If you're trying to analyze a representation of continuous tone media, the further differentiations in values increase beyond 1 DeltaE, the more difficult it becomes to evaluate. In that regard higher gamut and contrast ratios aren't always better. Gamma encoding is basically a way of making things aesthetically pleasing over a lower contrast ratio. If you had a much greater range of saturation and contrast, the potential would be there for it to look really good assuming high gamma value encodings became a thing of the past and the number of bytes allocated to describe the gamut increased exponentially without introducing noise in the per channel luminance gains, yet today some of the extreme contrast ratios are only good for marketing. Much of the time they just start to clip shadow and highlight detail.

    I tend to favor stability and detail over everything else. If you're looking at photographs, being able to see the full range of details over your display without major color shifts from one week to the next is a major factor. If you have that stability and detail yet things tend to print a bit darker, you can at least learn to mentally factor for it. If they're too much of a moving target visually, you've got nothing as the display can make anything look good.

    It was fun writing that.

Share This Page