HP z27x Review
by Chris Heinonen on December 2, 2014 4:00 PM ESTFor calibrating the HP z27x, we use CalMAN software from SpectraCal. Measurements are made with an i1Pro and a Klein K-10A using APL 50% test patterns. For pre-calibration measurements, I used the sRGB mode on the HP z27x and set the brightness to 200 cd/m2. Our targets are 200 cd/m2 with a Power 2.2 gamma and 80 cd/m2 with the sRGB gamma. We target the sRGB color gamut for these tests.
Pre-calibration the grayscale has a slight blue tint to it. The average color temperature is over 7000K, which corresponds to a blue temperature, and the graphs confirm it. The gamma is very good except at the low end, which is due to the HP using an sRGB gamma instead of the Power 2.2 gamma that we target. The dE2000 error levels stay below 3.0 for almost the whole range from 0% to 100% intensity, so it is still good despite these issues.
More impressive are the color error levels. The color checker and saturations are very good, and would be excellent even if this was after a calibration. Nothing creeps over a dE2000 of 3.0 so these are errors you will not notice in moving content. Pre-calibration, the HP is very good.
CalMAN has the ability to directly control the HP z27x, so it performed the calibration in the display LUT and the video card LUT. This provides the most accurate images possible and allows you to correct color errors that cannot be done in the video card LUT alone.
Post-calibration the sRGB numbers are virtually perfect. There are a couple issues in the grayscale but those are actually reading errors and not real errors. If they were actual errors the numbers around them would be higher as well. I ran into this a couple times with the Klein, probably due to my unfamiliarity with it compared to my usual meters. Gamma tracks almost perfectly, and the RGB balances are basically perfect.
Color errors are practically non-existent, with average dE2000 levels below 0.5. For professionals where accuracy matters, the HP z27x delivers that after calibration.
Results are the same when we target 80 cd/m2 and an sRGB gamma curve. You can see a couple of reading errors, but overall the display is virtually perfect. Contrast ratios fall a bit as it corrects 100% white, but there aren’t any major issues in the display.
Pre-calibration the numbers from the HP are good but they get to be reference class after calibration. Anyone using this display is going to calibrate it, because that is their target market, and it offers the ability to be completely dialed in.
Facebook
Twitter
RSS
47 Comments
View All Comments
Olaf van der Spek - Wednesday, December 3, 2014 - link
Cheap? It might not be as expensive as this one but the ZR2704w isn't cheap.DiHydro - Tuesday, December 2, 2014 - link
To me, that uniformity error looks like problems I've seen in monitors with bent or squashed frames. Sometimes in shipping, the inner metal frame can bend or get squeezed, and this will cause the diffuser to get moved slightly closer to the screen or away. This will case atypical uniformity along the edges of the monitor. Does this monitor have a very sturdy outer casing? Was it picked up by one side, or maybe bent just a little while unpacking? It's certainly worth investigating for such a strange result.joelypolly - Tuesday, December 2, 2014 - link
Would be interesting if liquidmetal will ever be used as internal framing due to its more crystalline structure which is less likely to bend or warp.anactoraaron - Tuesday, December 2, 2014 - link
I was thinking the same thing, especially since it's in the top corner.mapesdhs - Tuesday, December 2, 2014 - link
Amusing that so much is being made of 4K issues for professionals, when studios are
already working with 8K. :D For example, see the article on creativecow entitled,
"Quantel Unveils 8K 60p Pablo Rio Workflow" (they use quad-4K monitors for preview,
three Tesla K80s for processing).
Ian.
tyger11 - Tuesday, December 2, 2014 - link
Maybe _some_ are working with 8K, but most high-end cinematographers say they don't want more resolution than 4K, they want much more dynamic range. I saw a discussion panel with 3-4 very high-end hollywood cinematographers, and every one of them were disinterested in resolutions higher than 4K, but were salivating at next year's big expected jump in dynamic range (in the area of 20-21 stops of DR). If we can get rid of the Bayer sensor once and for all, 4K resolution will have a noticeable amount of increased detail; no need for any more than that, even in a cinema setting. More DR - especially if we can get to 20-21 stops - will be huge.Doach - Tuesday, December 2, 2014 - link
I really dont know what to say. I dont think you realize what display you are reviewing. Or know the history behind the dreamcolor display. Know this. A retina display from a 5k imac cannot come close to the performance of the display you just reviewed. Not even in the same ballpark. Do you even now that this is a dreamcolor display? From the article it appears you do not. They really dont need any introduction as most movie production companies use these displays.You just reviewed a display that was damaged in shipping, which would impact all aspects of your review. Black levels, uniformity etc.
That monitor was damaged in shipping. I work for dreamworks and we have several of these. We had one with the same problem. The box it came in was damaged on one side.
You should have called tech support before writing this review and got another one. Saying that this shipped from the factory like that to your readers implies that HP has a quality issue is amatuer at best without knowing for sure.
There is a reason we use dreamcolor displays and not NEC displays. It seems you need to reeducate your self on why rec 2020 is important in a professional display.
Poorly written review. You guys need to do better if you want to be taken seriously by the professional community. Anand was more through and would contact the company first before writing such a review. You just threw them (HP)under the bus over a shipping error.
These monitors are calibrated and tested for accuracy and yes uniformaty issues before they leave the factory. If they were not they would not be used very long by the professional community. As it stands now HP dreamcolor displays own the professional market.
You should know this. If you do not you should not be writing articles for anandtech.
Techinator - Tuesday, December 2, 2014 - link
I'm sorry, but there's an awful lot of arrogance in your comment. How do *you* know the box was damaged in shipping? You're making a pretty big assumption there, and you know what they say about those. But let me just make sure you understand what I'm saying: you are being a complete ASS in your comment. There. We're now clear.I could say more, but there's not much point. You're so biased that you can't even objectively point out problems with the review and resort to simple blanket statements. "Poorly written" -- what part, the part where he ran all the tests, talked about why Rec 2020 and other wide gamut options matter, and all of the other text? Or just the part where he showed that HP sent a reviewer a display that clearly has a uniformity problem?
anactoraaron - Tuesday, December 2, 2014 - link
Yeah, but he may have a point about it being damaged in shipping. None of the other $1k+ displays reviewed here are THAT off in one corner. A quick email to HP about this should have been done.cheinonen - Wednesday, December 3, 2014 - link
If I were to request a new sample from HP, then I'd be accused of having HP pick me out a perfect review sample that isn't anything like what they ship to people buying one. If I don't, then I get these comments. As I said, the only real negative about the HP is the uniformity on mine. This is my 2nd sample, as the first developed a pink stripe and HP replaced it. However, I've not had the same uniformity issues with samples from NEC in their professional lines. Perhaps they hand pick them, or perhaps their packaging (which is the best out there, IMO) helps to prevent these issues.There's a very good chance this is just a sample issue and most don't have this problem, but unless I get in 20 samples to test I really won't know.