Although Samsung can indeed use a technology like that, it is highly improbable that this is the difference between C-PVA and PVA. This must be a coincidence of two abbreviations because
- The mentioned works study MVA matrixes; although MVA is similar to PVA, there is no specific mention of PVA in them.
- None of the mentioned works is published by Samsung; their authors work at other institutes or research centers.
- The abbreviation CP-VA occurs in one paper only; otherwise, the abbreviation CP is used along with LP for the sake of briefness.
- The letter P stands for Patterned in “PVA” and for Polarization in “CP-VA”.
Thus, I am inclined to think that this is a pure coincidence of two abbreviations. It is hard to tell if the circular polarization technology is used or going to be used in some versions of MVA or PVA matrixes unless you take the LCD panel apart. CP technology does not have distinguishing external features.
So what is the difference between C-PVA and S-PVA? Although only Samsung’s own people know this for sure, I decided to scrutinize LCD matrixes with special tools.
First of all, I made macro photographs of the screens of a Samsung SyncMaster F2080 (C-PVA) and SyncMaster 215TW (S-PVA) monitors with a Canon EOS-350D camera with Canon EF-100/f2.8 macro lens that delivers a scale of 1:1 (that is, the image projection on the camera’s sensor is the same size as the original object).
S-PVA pixel structure
You can clearly see the characteristic dual subpixels of S-PVA. They consist of two zones, usually denoted as A and B, and one zone is turned on at high brightness only. So, the first picture shows red subpixels of roughly rectangular shape while the second picture shows two small pieces that represent one zone of each subpixel, the second zone being completely turned off.
It is this two-zone structure that differentiates S-PVA from older PVA matrixes which used to have a monolithic subpixel divided into four domains. An S-PVA matrix has two zones with four domains in each, for a total of eight domains per each subpixel. This helps fight the gamma shift effect which occurs when not only the contrast ratio but also the gamma (i.e. the correlation between the video signal sent to the monitor and the resulting screen brightness) changes when the screen is viewed from a side. The pixel zones of S-PVA matrixes have such shape, position and voltage (in the most expensive matrixes that are installed into some TV-sets, the two zones of one subpixel can even be controlled independently) as to mutually compensate the gamma shift effect for each other.
Unfortunately, the gamma shift effect is not absolutely eliminated even in S-PVA matrixes. Besides, these matrixes have one more difference from PVA. Their viewing angles are asymmetric: the gamma shift is bigger from one side.
But it’s time we got back to C-PVA.
С-PVA pixel structure
As you can see, there is no sign of the subpixel being divided into zones. It is monolithic at any brightness. Besides, the subpixel has very uniform brightness. Particularly, it does not have the dark dot in the center which can be seen in the photo of the S-PVA.
Thus, the pixel structure is a return to the ordinary PVA matrix: one zone and four domains. Does this worsen the viewing angles?
Unfortunately, we at X-bit labs do not have methods and tools that would allow us to accurately measure the gamma and color shift when the screen is viewed from a side, so I relied on a subjective comparison and scrutinized a SyncMaster 215TW (S-PVA), SyncMaster 245T (S-PVA), SyncMaster F2080 (C-PVA) and SyncMaster F2380 (C-PVA) from all sides.
First off, I can’t say that the gamma shift is bigger on C-PVA. On the contrary, the asymmetry typical of S-PVA is eliminated, and C-PVA looks even better in this respect. The general impressions from the two matrixes are similar. Both show the gamma shift effect of about the same value. It should not be a problem in most applications.
Second, the color shift is the same, too. The tonality of colors changes somewhat with both C-PVA and S-PVA when you look at the screen from a side. The changes are similar with both matrixes.
Third, the loss of dark halftones when the screen is looked directly at can be observed with C-PVA like with any other type of VA matrixes. It means that if you look directly at a photograph, you may lose some details in shadows. But as soon as you deflect your head to a side, those details will show up as if out of nowhere.
Interestingly, both S-PVA based monitors proved to be better in this respect than the C-PVA models, but I would like to give you more info on this issue as it is often criticized in discussions of S-PVA. So, I first used a highly sensitive photo sensor to record the very beginning of gamma curves – not higher than dark gray – of three monitors: SyncMaster 245T (S-PVA), SyncMaster F2380 (C-PVA) and SyncMaster 2333SW (TN).
The X-axis of the diagram shows color (black is 0, white is 255), the Y-axis shows the measured level of brightness. The green curve is the ideal curve for gamma 2.2.
You can see the effect clearly enough: the curve of the C-PVA matrix is nearly flat in the darkest section of the diagram, meaning that the difference between two adjacent tones is very small, much smaller than it must be. Then, the curve rises up quickly, obviously to meet with the ideal gamma 2.2 curve in lights.
The curve of the TN-based monitor goes higher due to its lower contrast ratio in comparison with C-PVA, but is shaped almost exactly like the gamma 2.2 curve. Indeed, TN matrixes are free form the loss-of-darks effect.
The curve of the S-PVA based monitor seems to provide a clue as to why the SyncMaster 245T looks better than C-PVA. The curve goes very high, which indicates a high level of black and, consequently, low contrast ratio, and has a queer shape. It is almost a straight line rather than an exponential function.
It looks like SyncMaster 245T developers sacrificed the monitor’s contrast ratio, deliberately increasing the level of black and thus taking the gamma curve out of the area where the difference between adjacent tones is small.
To check out this supposition, I tried to move the curve on the SyncMaster F2380 in the same way. This could be easily done in the graphics card’s control panel. So, I opened the ATI Catalyst Control Center (Nvidia’s driver offers a similar setting, too), switched to the color settings screen and increased the Brightness parameter by 25. Black became somewhat brighter while the monitor’s gamma curve (marked as C-PVA (corr.) in the diagram above) got much closer to the ideal one in shape. The loss-of-darks effect has got weaker and moved to the area of nearly-black tones.
Thus, the difference between S-PVA and C-PVA in this case is only due to the different setup of the tested monitors. It is also clear that the loss of darks on PVA matrixes can be opposed by simply increasing the level of black. Unfortunately, this is done at the expense of contrast ratio and I should say that the developers of the SyncMaster 245T overdid it a little.
The effect can also be eliminated without losing in contrast ratio by fine-tuning the shape of the gamma curve in the area of darks, but it is difficult to do this manually while popular hardware calibrators have insufficient sensitivity on such tones. I guess Samsung engineers should pay more attention to this issue and write appropriate settings into the monitor’s LUT, but the settings must be more accurate than those of the SyncMaster 245T.
Of course, the adjustment of the level of black and the beginning of the gamma curve does not solve the problem of darks getting brighter when viewed from a side, but at least you can see all details when looking directly at the screen.
Summing it up, I should say that, despite the obvious simplification of the matrix structure (which must be the reason for the lower manufacturing cost), C-PVA is not fundamentally different from S-PVA in terms of viewing angles. The difference I have observed is due to setup peculiarities of specific monitors. On one hand, C-PVA has such unpleasant effects as gamma and color shift, and the loss of darks when looked directly at, but on the other hand, its parameters are not visually worse. Moreover, C-PVA has symmetric viewing angles, as opposed to S-PVA.
C-PVA is superior to modern TN matrixes in terms of viewing angles. The difference is obvious as soon as you put two such monitors next to each other. The viewing angles of C-PVA are almost identical from all four sides whereas the image on a TN matrix gets dark when viewed from below. The color shift is also more conspicuous on TN matrixes when the screen is viewed from left or right.
The other parameters of the two new monitors based on C-PVA matrixes will be measured and discussed in the next sections using our traditional methods and tools.