Long-term plasma TV tests show color, black-level changes but no major issues yet

Update January 19, 2012: See the latest test results here.

After a year in the lifespans of five 2011 plasma TVs that I've subjected to new long-term tests, all have shown some fluctuation in black level, light output, and color, and in some cases those changes are significant.

A few changes have been for the better, others can be fixed in calibration, and overall none is a "deal breaker" that radically changes my initial opinions of these TVs--all of which scored an 8 (Excellent) or 9 (Outstanding) in the Performance section of the reviews.

Before you read on, know that I aged these TVs in a controlled albeit accelerated process, leaving them on for basically six days a week for up to four months, so far. The charts below encompass a bit more than a year of normal use, assuming the average daily "on" time rate of 5.2 hours. When reading the charts, keep in mind that not all of the TVs have been aged for the same amount of time. See below for more details.

Black-level changes
In general terms of black level, the Panasonic plasma TVs got slightly darker (better) over the first 1,500 hours, and the Samsungs got somewhat lighter (worse). Here's the chart:

To put the chart in perspective, I consider anything 0.0089 or lower "good" and 0.009 to 0.019 "average." In a side-by-side comparison in the VT30 review I said: "The difference in absolute depth of black was quite subtle between the VT30 (0.0061 fL) and the Samsung [PND8000] (0.0071), not to mention the GT30 (0.0082)..."

Note the title "original picture settings," which means the post-calibration settings we publish in the review. The big caveat here is that readjusting the picture settings on the Samsung PN59D8000 at 1,500 hours caused its black level to return to 0.0074 Fl--essentially the same as the initial 0.0071 measurement at 150 hours. In other words, depth of black "drifts" higher on the D8000 over time unless you readjust it periodically, and so far the D7000 appears to be behaving the same way, although I didn't recalibrate it yet to confirm. (This finding corrects an earlier mistake I made regarding Samsung black levels over time; see No. 6 below for details.)

Since you can adjust the Samsungs to reclaim that excellent MLL, and there's ample "room" in the controls to do so, this drift isn't a deal-breaker in my book. On the other hand, it's less desirable than the behavior exhibited by the Panasonics, which don't require adjustment to maintain their darkest black level.

Black on all three Panasonics got slightly darker over time until 1,500 hours, after which it brightened slightly (aside from the VT30, which hasn't aged beyond 1,500 hours yet).

White-level changes
All five plasma TVs showed decreases in overall light output during the measurement period up to 1,500 hours. In other words, they got a bit dimmer; check out the chart:

At 2,000 hours, again, the Panasonic ST30 and GT30 underwent "tweaks" amounting to a slight increase in light output. I wouldn't be surprised if the VT30 underwent similar tweaks when it reached the 2,000-hour mark as well, but there's no telling how the Samsungs will behave. As a side note, the ST30's initial measurement is so much lower than the others because of decisions we made during calibration.

In most cases I calibrate TVs to 40 fL, which is plenty of light for a dim room. Keep in mind that most of these displays can get brighter, although not much more-so--and certainly not as bright as most LCD/LEDs. It's also worth noting that larger plasmas have lower light output capabilities.

Color temperature changes
On all five TVs, the color temperature,which determines the overall color palette of the display, has become somewhat lower (redder) as the TVs aged over the first year-equivalent. Again, the chart, keeping in mind the target of 6500K:

To my eye the change in color temperature is significantly more visible than the black- and white-level changes above. Going from 6500K to 6000K in the first year isn't going to ruin the picture for most viewers, but attentive watchers may notice a redder tinge to skin tones and white/gray areas.

The GT30 and ST30 show signs of "leveling off" after 900 hours, and I assume the color temperatures of all these TVs won't continue to decline at this rate.

Conclusions and thoughts so far

1. Plasma TVs get a bit dimmer and redder as they age, but they still maintain excellent picture quality during the first year. None of these findings spoils my positive impression of high-end Samsung and Panasonic plasmas.

2. I haven't tested any LED/LCDs long-term, so I can't say one way or another how they age. Even with these findings, I'd still take the best plasma over the best LCD, mainly because of the latter's uniformity issues and generally poor off-angle viewing.

3. Calibration is best performed more than once. If you can only do it once, wait until after the first year of use (or as long as you can), after the picture has time to "settle."

4. Adequate picture settings and range are essential for picture quality sticklers. The lack of grayscale controls in Panasonic's THX setting, for example, becomes more of a liability given the reddening we measured during the first year. Samsung's superb picture settings are a major asset and advantage as the panels age.

5. Panasonic and Samsung age basically the same except for black level. Between the two, we prefer Panasonic since recalibration of black level isn't necessary.

6. As mentioned above, I originally indicated that the Samsung PN59D8000 became lighter as it aged (I made this statement in the performance section of the PN59D7000 review). The fact that the lightening can be corrected via calibration contradicts that statement. I will update the PN59D8000's settings and correct the PN59D7000 review accordingly.

7. Panasonic has publicly stated that its plasma TVs' black levels no longer worsen over time because of a design change from 2010 and earlier models. When reached for comment about our findings, Samsung's statement was: "Samsung's testing in the past has not shown significant changes to black levels on its plasma TVs either over time, or for different screen sizes." If either TV maker has anything to say about the results in this post, I'll update it. Update 9/1: Panasonic replied with a comment that reiterated the design change it touted earlier, but said nothing specific one way or another about the tests. Samsung has not yet responded to a request for comment.

8. Neither of the Samsungs I've aged have shown any sign of the peeling issue reported on some other Web sites. Samsung has said it will repair any peeling TV.

9. I will continue to age and test these TVs, as well as the LG 50PZ950, throughout the year.

10. I'd be surprised if the second and succeeding years of aging showed as drastic a change to any of these picture quality parameters as the first year did. In other words, I expect these sets to stabilize over time, but there's no telling when that will happen.

Testing methodology
All of the plasma TVs were aged on a schedule that included approximately 144 hours/six days of "on" time and 24 hours/one day of "off" time per week. Essentially, I'd leave them on for six days and give them a "break" of one day off every week. During the "on" time they were set to the picture settings used for each review and displayed a loop of the 10-minute IEC 62087 test DVD clip (the same used to test TV power consumption) played from an upconverting DVD player set to 1080i resolution (after more than 2,000 hours, or 12,000 loops, that Oppo is still going strong).

About once per month (roughly 500 hours of accelerated aging) I would measure each TV using the same hardware and software used to calibrate and measure the TVs for the original reviews. Measurements were conducted using a Konica Minolta CS-2000 spectroradiometer in our completely dark lab at the picture settings used for each review. For each measurement, I first made sure the TV was warmed up for at least 1 hour, then I caused the SpecrtaCal software's ISF_Advanced workflow to take a "Pre-Calibration" pass using the 21-point grayscale option--which includes data points for both zero percent black and average color temperature, among many others.

Many of the results are recorded in this spreadsheet. If you compare the charts with the spreadsheet, you'll see that I rounded off hours in some cases (+/- 250 hours is the largest difference, with less than 100 more the norm). In the case of the Samsung PND7000 plasma, I didn't measure at 100 hours (the TV was reviewed and calibrated at 385 hours), and I skipped measurements at about 500 hours for the ST30 and GT30 (those null values are plotted on the charts to maintain line continuity).

All three Panasonics are manufacturer-supplied review samples, while both Samsungs were purchased on the open market by CNET.

There are a couple obvious problems with my methodology. First, I've only tested five 2011 TVs. Any of my findings might be different on other samples of the same TV; I just don't know, and testing multiple samples of each model is beyond the scope of this project. Even more important, this accelerated aging schedule might cause different changes compared with how a TV ages normally. I decided to accelerate aging simply because simulating normal use would take too long. I'm not sure how "normal" aging affects these TVs and welcome any information to that effect.