Answers about the new iPad and Color

iPad2 gamut over iPad3 gamut

Please see the previous articles in this series:

Color Management and the iPad

Questions about the new iPad and Color

And for further information about iTunes and color conversion for iPads see:

More Answers about the iPad and Color

Apple described the gamut of the new iPad as “44 percent greater color saturation.” Which left us without much of a standard of comparison. After receiving an iPad3 today, I tested its color gamut, whitepoint, brightness, and screen detail against an iPad2. The graph above shows the increased color gamut. It also represents what I would call a “normalized” color gamut, in that the primaries are now more closely aligned to what we expect red, green, and blue to be. I haven’t attempted to do the math to see if the larger gamut is actually 44% larger, as this would only be for this specific iPad3, over this specific iPad2, so would not necessarily be representative in the way that the larger sample Apple could test would be. But it is significantly larger. Which has two noticeable effects: more saturated colors at the extremities, and different color in other areas which both tablets can reach.

sRGB gamut over iPad3 gamut

The second illustration shows the gamut of the new iPad with the gamut of the sRGB colorspace superimposed on it. I chose this order of layering, as there is more area in the iPad’s gamut that is outside of sRGB than the other way around, though as you can see at the bottom of the illustration, there is a sliver along what’s known as the purple line (from saturated red to saturated blue) where the new iPad does not appear to reach the limits of sRGB. On the red to green edge of the triangle, the match is excellent, and the mismatch on the green to blue edge is in the iPad’s favor; though without color management, no mismatch in any direction is actually favorable.

The ideal way to reproduce colors on a device is to have a larger than needed gamut, and to use color management to display the correct colors out of that larger range. The easy way to avoid color management is to attempt a one-to-one match between the two gamuts in question. The illustration above shows that Apple has gone for the easy fix, and created a screen with a gamut very close to sRGB. As long as the files sent to it are in sRGB, this should produce a reasonable match, certainly a much better match than the smaller and skewed gamut of the earlier iPads, and other iOS devices.

Viewing sample images on both iPads confirmed this effect. I first calibrated each iPad with a Spyder4, then viewed test images in the Datacolor SpyderGallery app, where the color profile can be turned on and off at will. The following are my comparison notes:

* With color management off on both screens, the color did not match between them. The iPad3 was considerably closer to matching the same image on a calibrated desktop display, however.

* Turning color management on made a big difference (as it has always done) on the iPad2.

* Turning color management on made a smaller difference on the iPad3; and often in the other direction.

* Matching between the iPad2 and iPad3 was excellent for in-gamut colors once color management had been turned on for both.

* Matching for out-of-gamut colors was closer, but not identical between them with color management.

* Yellows, in particular, look different on the iPad3 than on the iPad2, to the degree that memory colors, such as the logo color of SpyderGallery, look considerably different. Since there is no color management at the OS level, this cannot be corrected or changed.

*Density corrections in SpyderGallery still improve shadow detail on both iPad2 and iPad3; apparently there was not change in default gamma, even with the change in gamut.

* Default whitepoint on both iPads is very close; unlike the white points of my iPhone 4s versus my iPhone4, where the difference took some time to get used to.

*Maximum brightness of both iPads is, amazingly, identical. Earlier iOS devices have had a significant variation in max brightness, so its basically a coincidence that the iPad3 I received happened to be exactly as bright, today, as  my year old iPad2 is. My iPad2, when it was new, scored the highest brightness of any iOS device we tested, so having it still be as bright as a new iPad3 today is not surprising.

Next; the other factor that everyone has been waiting to hear about: the “retina display” screen. Let me start here by saying that an iPad 2 and an iPad3 are very similar to hold, with the weight difference being pretty much indistinguishable, and the difference in the lip profile minor. In your hands, the easiest way to tell which is which is by the larger camera filter on the back. On a desk, the easiest way to tell is by the slight thickness increase with the iPad3.

Given the degree of similarity, I mixed the two devices up repeatedly while handling them for comparison. So when I did my image detail comparison, it was truly a blind test, as I couldn’t even remember which iPad was which. The answer may not be what you are hoping to hear: the increased resolution is real, but not large. At 24 inch viewing distance or greater it is more or less indistinguishable (with good vision, or reading glasses to correct it). At 12 to 18 inch viewing distances, where close viewing of the iPad is more likely to occur, it is clearly discernible in type, but less so in images. Depending on where you images came from, and whether they were downsized for the iPad (as images coming in through iTunes are), there may be no difference at all. I expect that iTunes may start creating larger images for download to iOS devices. Certainly when viewing images from my Facebook library on both iPads, the resolution and compression Facebook has chosen for my images is low enough that there is no additional detail or sharpness with the iPad3. On images imported via the iPad Camera Connector, sharp and detailed images did indeed show sharper detail on the iPad3. You almost have to look for it to see it, but once you see it, its quite obvious.

So, my conclusion, which is surprising even to myself, is that the increased resolution in the iPad3 is nice, but not killer. While the increased gamut is much more noticeable. My second conclusion is that images on the iPad3 are closer to correct color automatically, thanks to the closer-to-sRGB screen gamut, while images viewed through SpyderGallery with a custom profile are more accurate still, though the difference is less with the iPad3 than it is with the iPad2. And my recommendation is that people  probably don’t need to run out and upgrade, unless they are using their iPads as photographic portfolios and other color critical uses, where the extra gamut, and closer match even in non-color-managed apps, matters more than in most other uses.

And as a footnote: The SpyderGallery wireless communication, calibration functions, color managed viewing functions, and remote viewing functions for Facebook and Flickr all seem to work excellently on the iPad3. I don’t anticipate any need to for an iPad3-specific update to the app.

Credits: C. David Tobie, Copyright 2012.   Website: CDTobie.com   Return to Blog’s Main Page

Published by cdtobie

This blog covers a range of issues of interest to photographers and those involved in the digital photographic workflow, digital tools and platforms, and fine art output.

18 thoughts on “Answers about the new iPad and Color

  1. Excellent iPad color review! Probably the best on the web (at least the earliest :).
    And your final conclusion also surprises me a little regarding the retina display differences, or non-differences. Even though I wasn’t planning to rush out to buy the iPad3 since my iPad2 is less than 6 months old, I don’t think I’ll be sprinting to the Apple store right away.

    And just to confirm, an iPad2 app will appear at the same size on the iPad3 but with lower pixel density (PPI), correct? I think I got that wrong on my earlier post somewhere here.

    Good review.

  2. Interesting review, CD. With the extra resolution screen, is most of the extra processing power being eaten up by the graphics density, or is the “new ipad” noticeably faster in showing images?

    When looking at compared images on the two units, were you feeding both units at the native resolution of the screen, or did you use the same pixel size image on both which was then re-sampled (up or down) to full screen size by the image viewing software?

    Thanks

    1. The graphics and a lot of related navigation is definitely a bit snappier on the new iPad. It would be a good trick to re-size the same image to the native rez of both devices and compare responsiveness, but it would be an academic test, as thats not how real world usage happens. I used an image with a much higher rez than either, and let the device display as it chose. I wasn’t really testing the processing speed; the screen is more my specialty.

  3. Interesting article! I’ve read a lot more about the resolution changes than the color changes, and it was useful to read about those here. The details are interesting, such as the difference in the yellows between the two iPads. Did you feel that one had a “better” yellow than the other?

    One note about wording: the iPad discussed here is “the new iPad” (the updated version of the iPad 2), not the “iPad 3.” This could be confusing to people who are looking at this article after the actual iPad 3 is released sometime in the future. One wishes that Apple had stuck with their “S” terminology and called this the iPad 2S, as this “new iPad” phrasing is incredibly awkward journalistically and prone to produce these sorts of confusions.

    1. One gets accustomed to color looking a certain way; so the difference in color on the new screen does look wrong, when iOS logos and other items you have always seen looking one way suddenly look another way. The comments I’m seeing elsewhere, about the new iPad’s color being wrong may be based on that, or on unit variation, with specific new iPads having different color than the ones I’ve seen.

      All I can address is the single new iPad I have measured and tested. It has far more accurate native color than any of the earlier iOS devices. I base this, not on icons, logos, and wallpaper, but on known test images shown in SpyderGallery, with and without color management. Getting similar results would depend on getting a new iPad with similar color characteristics to the one I tested (I don’t have a sense of the variability of the new screen yet), using precorrected test images imported in a manner that will produce correct image color (such as making sure they are in sRGB before sending them to the iPad), and viewing in an environment where the ambient light is significantly lower than the iPad’s screen brightness, so your eye’s white balance is controlled by the iPad, not the environment.

      Having stated all the fine print, the short answer is: What I’m seeing for color on the new iPad is closer to correct than what I’m seeing on older iPads (more fine print: in all locations except in SpyderGallery, where what I’m seeing, since it is color managed, is the same on both, except for the few saturated colors which are out of the old iPad’s gamut; these colors are more accurate on the new iPad even under color management.) Sorry that there is so much fine print involved in giving a straight answer…

  4. Great followup post! I’m glad to hear that this supposed 44% increase in saturation doesn’t render the device a “for-consumers only” toy. Finding out that this “44% saturation increase” is actually just marketing speak for “better reproduction of the sRGB IEC61966.2.1 colour workspace” is welcome news indeed.

  5. hi, i was wondering how you got the .icc color profile out of the ipad to do your comparison-charts?

    best,
    simon

    1. The short answer is that I develop calibration and profiling products for a living, so I have specialty tools and custom utilities at hand which the average iPad owner would not have.

  6. What did you deduce about white balance? I have had two iPad 3’s – one with a pink cast and one with a yellowish cast to white backgrounds when reading B&W text in comparison to iPad 2. Yes, colors are more saturated but the off-color whites are disconcerting to put it mildly. The display runs dimmer as well. I agree the retina display does not make an overwhelming difference.

    1. Yes, there is not just device variation in iOS Whitepoint, there have even been cases of new firmware changing the whitepoint on a given device. My iPhone 4s has a visibly different whitepoint than my iPhone 4 had… Since Apple is very averse to any color management for iOS, SpyderGallery is the only way I can see of controlling such functions at the moment. I hope to see whitepoint and gamma choices in a future Gallery build, but I can’t say more than that at the moment.

Leave a comment