Investigation: Active 3D TVs, Full HD 3D Guaranteed?
Two types of 3D technology are fiercely battling for market domination—active 3D and passive 3D, which use different kinds of glasses. While active 3D glasses are heavier and more expensive, they have the major advantage of displaying 3D images in Full HD ... or, at least, that's what we thought!It's pretty much certain that the future of 3D technology lies in glasses-free 3D. You'll have to be patient, though, as even if one glasses-free telly was released at the end of 2011 (the Toshiba ZL2, which costs around £8,000), the technology isn't yet mature enough to be rolled out fully. In fact, we wouldn't expect to see decent-quality glasses-free 3D TVs before 2015 at the earliest. In the meantime, two types of 3D technology are currently available, each of which uses a different kind of glasses.
- Passive 3D glasses are light and very cheap to produce, but they halve the resolution of 3D images (to 960 x 1080 pixels) as pictures are effectively split in two.
- Active 3D glasses are heavier and require a power source like a battery or a cable. These aren't as comfortable to wear, but they do deliver better image quality, as 3D images maintain 1920 x 1080 pixel Full HD resolution.
However, it now seems that not all active 3D TVs can deliver Full HD 3D images in 1920 x 1080 pixel resolution!
So what would happen if active glasses were no longer able to deliver the Full HD 3D images promised? Would they still have anything going for them? Well, no, actually. Victory would be declared for passive 3D pioneer LG, and other manufacturers would slowly abandon active 3D to follow LG's lead. In fact, this process already seems to be under way, as Philips and Toshiba are now using both active and passive 3D in their TV ranges.
Eight Active 3D TVs Plus One Passive 3D TV Put To The TestWe decided to find out whether active 3D TVs really do deliver Full HD 3D images as promised in their tech specs and by the little logos that proudly embellish their packaging.
We might as well tell you straight away that we decided to investigate this after receiving several user complaints about Samsung TVs (2011's D6xxx models in particular). Even though the problem seems to originate from Samsung TVs, that doesn't necessarily mean that it only concerns Samsung models—but that's something we'll have to check. Since Samsung largely dominates the TV market and, in particular, the active 3D market, it seems logical that any potential problem would be detected in a Samsung model first. However, it inevitably means that Samsung's D6xxx series was the starting point for our investigations, so we hauled a couple of these TVs back into our labs together with selected competitor models (more will be tested soon) to take a closer look.
Here are the eight active 3D TVs and one passive 3D TV we tested:
The test results are quite surprising.
Results: Samsung D6xxx Series Lags Behind
Both of the Sony TVs, the Sharp model and the Samsung UE46D7000 display our 3D 1080p test image perfectly.
The Philips active 3D model (46PFL9706H) displays the 1080 lines very well, but not the 1920 columns. No matter how we adjusted the manual settings, we couldn't stop the TV processing the image and stretching it very slightly horizontally. This interpolation degrades parallel lines on our test image (see bottom test image), in turn making these areas look less sharp onscreen.
The Panasonic 42GT30 plasma TV is one of the best choices for 3D picture quality at the moment, as its pixels are much less subject to ghosting than LCD models—in fact, we measured just 8.5 ms compared with 10 to 20 ms for LCD TVs. This faster response time ensures that the 3D images for each eye don't cross over into one other, a defect that's known as crosstalk. No LCD can match this plasma TV on that point. The good news is that this model displays 1080 lines but, unfortunately, the resulting image still isn't perfect—it's slightly degraded in a similar way to the Philips TV, although there's no overscan (stretching) going on here. On the test image below, the segments of the circle aren't as clearly defined or as sharp as on an LCD TV. The second test card will show us why. Note: we'll be carrying out the same tests on other sizes of this model to establish whether the problem only affects the 42" version.
Things are even more problematic on the Samsung entry-level and mid-range TVs. The two D6xxx TVs drastically reduce the resolution of our 1080p and 720p test images. The loss in detail is as noticeable in the columns as it is in the lines. In practice, that means the resulting 3D image is of a lower quality than an equivalent 3D image displayed on the Philips model, and even more so than on the Sony TVs.
Here are the results from our first test image:
The test images above show that while the Philips, Sharp and Sony TVs, plus the Samsung D7000, display the 10 segments required to render the edge of the circle, the Samsung D6200 and D6500 only use five segments to display the same curve. Some detail is therefore lost in the image.
To find out why, we used a second set of test images, one with 1920 alternate black and white columns and one with 1080 lines.
The Sony HX923 displays the 1920 columns and 1080 lines perfectly (degradation in the top-left image comes from our camera, as it's pretty difficult to capture these fine lines).
The Philips TV does a very good job of rendering the 1080 lines. However, certain zones in the image are degraded—in the middle of the screen, for example—because the image is stretched slightly horizontally. That also explains the slight loss in quality that's visible on the segments of the circle in the first test image.
In the Panasonic 42" plasma TV, white light from white pixels spills over into black parts of the image, which in turn reduces their intensity. The slim lines that are just one pixel wide on our test image highlight this defect only too well. It also explains why the segments of the circle looked blurred in our first test image, as while big blocks of black look nice and deep on this screen (with contrast reaching over 5000:1), at the edges of these dark areas, the adjoining colour bleeds into the black slightly (here, the adjoining colour is white, which shows the effect at its worst).
A similar effect can be seen in the Sharp TV, but on vertical lines this time. While horizontal lines are perfectly reproduced, white vertical lines aren't defined as well and look more grey than white. This could make the overall image look less sharp than on other TV screens.
Finally, the Samsung D6200 and D6500 display the 1920 columns perfectly, but the 1080 lines are barely even visible! In fact, these two TVs fuse the horizontal lines together perfectly, effectively halving the resolution. That explains why the Samsung models use five segments rather than 10 to render the curve in the first test image (see above). All this boils down to the fact that when watching 3D content on these TVs, you'll be watching 1920 x 540 lines rather than the 1080 lines required for Full HD (1920 x 1080).
In Practice: 3D MoviesThis difference in display quality picked up by our test images is also visible in 3D content. Here's a scene from a 3D film, showing a bicycle wheel as it should be displayed in Full HD 3D:
Bike wheel in Full HD 3D on the Sony EX723, HX923 and Samsung D7000
Here's the same image on the Philips 46PFL9706H
While it does stretch the picture, the Philips TV does so very discreetly. So while the effect is picked up on our test images, it isn't strong enough to be really noticeable when watching a 3D movie. The onscreen image is effectively 'perfect' to the eye.
The same image on the Panasonic Viera TX-P42GT30
The 3D picture on this Panasonic TV isn't quite as sharp as on the Sony models or the Samsung D7000. However, some of this degradation could come from our photograph. It's much more difficult to take a photo of a plasma screen than an LCD, as bursts of light from the pixels are emitted at very high frequencies. The above picture is as close to reality as we could manage, but the defects are accentuated and they aren't shown exactly as we saw them onscreen. In reality, the onscreen image looked closer to that of the Philips TV. However, we still noticed that while blacks and colours are very intense on this 42" Panasonic plasma TV, the sharpness can't quite match the best LCD TVs, as lighter tones do bleed into darker areas where they meet. That's something we'll testing on other sized versions of this TV soon.
The same image on the Sharp LC-46LE830
As expected, the fuzzy vertical lines do make the image on the Sharp TV look less crisp than on some other models, especially compared with the excellent high-end Samsung TV and the two Sony models tested above. There are 1080 lines, but some detail is lost.
All in all, the images reproduced by the Philips, Sharp and Panasonic TVs are still very good quality. Unfortunately, the same can't be said of the TV Samsung D6xxx TV.
The same image on a Samsung D6xxx TV
The loss in sharpness and overall picture quality is only too visible, and in turn makes the colours all bleed into one another.
In light of this, it's fair to ask whether the image is still any better than that of a passive 3D TV. Up close, passive 3D images have dark horizontal lines across them, which disappear as you move further away from the screen. Take a look at the same image on a passive 3D TV:
The same image on a passive 3D TV
Samsung seems to be aware of the issue too, as the Full HD 3D logo has recently been discreetly removed from D6xxx product pages on the Samsung website. If you look at the picture in the top right corner of this page, you can clearly see the logo on the box of the D6200 one of our readers bought on the high street. Note that this logo hasn't been taken off the firm's higher-end TVs (D7000 and D8000).
Samsung even admitted that the D6200, D6500 and D6700 series TVs do actually deliver 540 lines per eye, while D7000 and D8000 models display 1080 lines in each image.
Samsung went on to explain that the Full HD 3D logo that featured on the packaging of its D6xxx TVs when they launched has now been removed. This, they explained, helps show customers the difference in quality they can expect in both 2D and 3D between the D7000/D8000 models and TVs further down the range.
Therefore, the removal of this logo shouldn’t be taken as a sign of Samsung admitting the problem outright, as the firm still considers the D6xxx series to be ‘Full HD 3D’ TVs—they're just a different kind of Full HD 3D TV, apparently, as according to Samsung, human perception of two images with 540 lines is effectively equivalent to Full HD.
Based on that logic, you could argue that passive 3D TVs should be considered Full HD 3D models too ...
However, we don't think that having different categories of Full HD 3D TVs is such a good idea. First of all, customers could find this kind of manufacturer's justification rather infuriating, but it could also be detrimental to the TV market itself. At a time when 3D is still trying to find a place for itself in the market, any added confusion could just serve to put people off completely. While competitor active 3D TV makers that deliver 1080 lines per image will no doubt be patting themselves on the back for being superior to Samsung, all of this does put something of a damper on active 3D as a whole, further strengthening the appeal of passive technology that's more versatile, more affordable and effectively gives the same visual result (at a distance of two to three metres from the screen)!
We can only hope that the technology is harmonised as soon as possible. Hopefully that day will come when TVs start to switch to passive Quad HD panels. We may even spot some first models on display at the CES 2012 tech show in January.
We tested a selection of TVs to check this problem out for ourselves. Our sources were a PC hooked up to the HDMI port used to display our 1080p and 720p test images, and then a 3D Blu-ray player to display 3D movies. In the future, we'll be carrying out these tests systematically.
We'd also like to thank one of our readers, a certain Alexandre, for his help and for lending us his TV to test.