Post and results updated October 21st with results from Samsung PN50C8000 and Panasonic TC-P50GT25.
Initial testing of 3D-compatible HDTVs has revealed that, in the default picture settings, displaying 3D images can indeed draw more power than in standard 2D mode, especially on plasma models.
We tested five plasma televisions as well as three LED-based LCD models, and the results generally show that more power is consumed to produce images in 3D mode as opposed to 2D. The plasmas were particularly power-hungry in 3D, averaging about double the power used in 2D. The only TV that didn't increase its power consumption in 3D was a Sony LCD--a discrepancy we chalk up to its dimmer 3D picture setting. See the full results after the break.
The main reason for the extra power draw in 3D, as far as we can tell, is that 3D needs a brighter image. The active LC shutter glasses required to view 3D on these TVs alternately block off each eye very quickly, so it stands to reason that the 3D image needs to be brighter than the 2D version to have similar impact. A quick eyeball observation sans glasses confirms that the 2D image is dimmer than 3D on the plasmas. On the other hand, we couldn't see an obvious brightness difference on the LED models.
In addition, the glasses themselves have a slight tint even when "open" and not blocking an image--Sony's glasses seem to have the lightest tint and Panasonic's the heaviest--so the TV must be brighter to overcome it. A Panasonic engineer told us to think of it as "watching TV through sunglasses."
Naturally the different makers can and will use different default picture settings, and different brightness offsets for 3D, but unfortunately we can't compensate for those differences in our testing yet. For example, the Sony's image in 3D (with the glasses on) seemed dimmer than in 2D (glasses off) with the same picture mode, but we can't say for sure and we can't properly test it (see below).
Even with 2D material the 50-inch VT25 used more power than the equivalent 2D TV, the 50-inch Panasonic TC-P50G20; we measured about 28 percent more juice (which costs about $14 more per year) after equalizing for light output. The discrepancy in Samsung's case, on the other hand, between the 2D-only PN50C590 and the PN50C7000 in 2D mode was slight (just $1.18 more per year for the 7000). Of course, plasma in general uses twice to three times as much power as LCD, especially these LED-based models.
Power use aside, in our opinion the Panasonic and Samsung plasmas outperform the LEDs with both 2D and 3D material. It's also worth mentioning that 2D will likely be used much more frequently than 3D on these TVs for years to come, so the monetary and energy use impact of this increased power use will be minor.
Unfortunately a rigorous test of 3D power use, with a comparison that specifies equalized light output in both 2D and 3D modes, isn't in the cards right now. It would require a measurement of light output through a pair of 3D glasses--literally placing the glasses over the lens of our light meter--which introduces variables that are difficult to control. Alternately, we could forget the glasses and just compare the light output of the screen in 2D and 3D mode with the same content, but at the moment we lack the proper 3D test patterns to do so.
In lieu of a test that controls for light output, we've elected to simply compare the power consumption in each TV's default Standard picture mode (with ambient light sensors disabled). We used the first 10 minutes of "Cloudy with a Chance of Meatballs" because that Blu-ray allows an in-menu choice of 2D or 3D presentation (in 3D mode we skipped past the 3D Blu-ray and Sony promos, which don't appear in 2D mode, before we began the 10-minuter period). All TVs were tested using the latest firmware versions. We used our standard test procedure, including hour-long warm-up times, for each measurement.