| || |
Re: Is this what the screen looks like?
I'm cross-posting my reply to this thread here because it touches on the differences in displays and what I've observed.
I'm sorry, but I'm really going to have to beg to differ on this. When I got my Nexus 4 and showed it the my girlfriend next to my Galaxy S II Epic 4G Touch Long Name phone, she instantly said, "The old one looks a lot better." I told her that the SAMOLED screen was merely hyped up because of all the usual reasons, but the difference DID bother me. I told myself, "No, this is more natural. It's not washed out. You're just used to the old screen," but I can't do it any longer. My screen IS washed out; it IS inaccurate. I'm not sure what to do about it.
Originally Posted by 2defmouze
I'm a semi-pro photographer (meaning I've made money from my work, but it's not my day job; see my stuff here) and thus really picky about image quality and fidelity. I also used to be a DVD reviewer for a couple of major sites and had to calibrate my home theater setups to properly grade A/V quality. I just recalibrated my main computer's display - a Dell UltraSharp U2410 IPS-panel connected via DisplayPort using a GretagMacbeth Eye-One Match 3 calibrator - and compared my current wallpaper (NSFW, so not linked; skin tones are a good source to judge because of the subtleties of complexions) from mine and other Flickr users work and the differences are there and problematic. I even dug out my OG EVO 4G to compare because it was its too-cool/bluish display that paled (no pun) in comparison to the GS2 was what made me thunk down $500 to buy the latter unsubsidized. I had to use the Movie setting in Display properties to get the least-hyped picture because Standard and Vivid were just acid-trip crazy.
Compared to the GS2, the N4's colors are flatter, paler and washed out, especially in highlights on skin which are getting close to being blown out to white. It looks cooler until you throw the EVO into the mix and see what really cool, bluish color cast looks like. When looking at all three, the GS2 looks the closest to the PC monitor; the N4 is clearly less saturated and skewed toward yellow (note: I do NOT see the yellow color cast on pure white that others have reported; it just looks like more red is needed); the EVO looks really cool (blue).
Checking another photo (again NSFW; tanned partial nude woman in vivid magenta leather jacket) shows similar issues. The N4 is flatter in contrast and the highlights are blown out; the wall in back doesn't look creamy, but almost white; the jacket looks pink, not magenta; the burgundy wash on her hair is almost unnoticeable. I couldn't check with the EVO because it's having a cow about updating (gee, it's been offline for 14 months, could that be the problem?) but since it's obsolete, who cares? The GS2 looks really close to the monitor and the N4 doesn't.
Using the LCD test patterns here - LCD monitor test images - the most damning one isn't the Black Level one (browser color profiles seem to mess with it) or the Gamma test (all the phones look waaaay out of whack) but the White Saturation pattern which demonstrates how brighter values are rendering. On my computer, I can see the difference between 254 and the 255 background. The highest the N4 shows is 247, meaning everything from 248 on up looks the same as pure white, 248-254 = 255 = not good. The GS2 goes four steps up to 251 and the difference between getting 96.9% of the way and 98.4% makes a difference. As someone who usually shoots black-clad musicians in dimly-lit clubs with black walls, the difference of a few steps is the difference between seeing a guy in a black shirt in a black room playing guitar or seeing a floating guitar and arm. (This is an all-time worst case example.)
The hardest thing to do in publishing is color management and printers and pre-production houses spend thousands of dollars trying to get their workflow together so that what designers see on the screen will match what's on the page when it's printed. It is said that a man with one clock knows what time it is but a man with two clocks is never sure and what may look fine in isolation may suffer in comparison to something else. (Like how your significant other may be cute, but stand them next to the latest Sexiest Man/Woman Alive and you realize you're in a relationship with a mortal.) The trick is to determine what is CORRECT. For clocks, you get one of those atomic deals and set your watch to that. For displays, you color calibrate and use test patterns.
This is what I've done and while there is a element of personal taste involved - like when you adjust the tone controls on your sound system to boom the bass or whatnot - I'm not interested in preference, but accuracy, and the screen on my Nexus 4 simply isn't accurate. Maybe it's isolated bad unit; maybe it's an issue with 25% of units; who knows? All I know is that some people say it's washed out; others say it's as good as the best-in-class HOX; some try to rationalize what they're seeing by attributing it to the differences between LCD and SAMOLED; and I just want a faithful version of what's being displayed.