Stumped by the Screen: Newsletter Challenge (December 2017)

There are lcd cells for each primary color, blue, green, and red. These cells have polarizers on both sides, crossed so that no light normally comes through. These cells are turned on (made transparent) by applying a voltage. This voltage causes the cells to become more transparent because it rotates the plane of polarization of the light passing through. The amount of rotation is determined by the path length of the light and when viewed at an angle, this path length, and hence rotation is greater. This causes the color shift.

If the above is true, the effect should be symmetrical in different directions, and a little experimentation proves it is not.

I believe it has to do with the color mask and the parallax (color leakage between adjacent cells).

I think this paper explains it.

https://www.deepdyve.com/lp/wiley/p-185-color-shift-analysis-of-ltps-tft-lcd-viewing-for-large-angles-CqqKrpvSDr

An microscopically thin alignment layer is placed on the glass to give the liquid crystal material a pre-tilt, to help the liquid crystal take the correct orientation when power is applied.

This pre-tilt is what causes the asymmetry in the optical phenomena. Ideally the LC fluid would undergo a 90 degree rotation between the full on and full off states. In reality, without the pre-tilt (in a TN LCD) some of the molecules could rotate 90 deg 'left' and some could rotate 90 deg 'right', and the optical switching would not be as good. The pre-tilt helps ensure they all rotate the same, and that they all orient in the same direction when unpowered.

The pixels and screen are designed to be viewed straight on, altering the viewing angle allows different wavelengths to be seen...

http://blogs.plos.org/mitsciwrite/2013/05/02/your-computer-screen-from-an-angle/

Rixter's answer is sort-of close, but not exactly right.

The 'normally black' LCD he describes is generally of the 'In Plane Switching' type (there are various names for it). These all have excellent color rendition out to extreme angles. You're not going to see pink become blue for these LCDs.

The older 'TN' (twisted nematic) color LCDs are 'normally white' (white when unpowered; you apply power to get gray shades, including black). These older TN type LCDs don't have as good contrast or color at extreme angles compared to IPS type LCDs. So at extreme angles you will see color shifts. The color shift is due to two phenomena: 1, the path length difference through the LCD which affects the net cross-polarization effect, and 2, the fall-off in contrast at steep angles. I.e., the blacks are less black at steep angles.

The net result is a color shift which can turn pink into blue.

I'm using my laptop screen as an example. I don't see any color shift left or right. Looking from the bottom, it appears to darken and from the top, it appears to wash out. My telephone appears to not change at all at any angle. The TV is also good at about any angle. So the laptop has a TN display and the phone and TV are probably IPS.

So why doesn't the laptop screen show the same difference when viewed from the top as from the bottom? I'm thinking that with no electric field, the molecules are all oriented parallel to the screen plane with the long axis towards the top and bottom of the screen. The molecules are polarized so that an electric field tends to attract opposite ends toward the front and back of the screen.

https://en.wikipedia.org/wiki/Twisted_nematic_field_effect

When the electric field is completely turned on, the molecules align front to back allowing the light to pass through without rotating the polarization. The light is extinguished because of the crossed polarizers front and back of the LCD. When the electric field is partially turned on, the molecules will be at an angle such that the optical path from the bottom of the screen will be looking at the molecules from the end (dark condition, no optical rotation) and the optical path from the top of the screen will be looking at the molecules from the side (light condition, more optical rotation). At extreme angles from the bottom of the screen, you are actually looking at the "bottom side" of the molecules, so the optical rotation is reversed resulting in a negative image. This would result in a color reversal on a color LCD display.

https://encyclopedia2.thefreedictionary.com/IPS+panel

"I'm using my laptop screen as an example. I don't see any color shift left or right."

Agreed. My Laptop (a 15" MacBook Pro with "Retina" screen) shows virtually no change in color at any reasonable viewing angle (up to 70° or 80° from normal). There are changes in apparent brightness and contrast. but not in color.

"They" have come a long way in display quality! I've been amazed for many years how they managed to achieve uniform brightness over such large areas, especially back when they used two fluorescent tubes for illumination. I have no idea how many LEDs are used in the current backlights, but I see absolutely no grid or array of brightness.

This guy explains it pretty well...

Thanks! This is a very different video that shows several different aspects of the technology.

Yes, TN LCDs show 'image reversal' in one direction. Most are designed for either '6 o'clock' or '12 o'clock' use orientation.

LCD manufacturers have used various techniques to minimize the image/color reversal. Typically they use polarizers with 'optical retardation' layers built in. Light travelling straight through the LCD is linearly polarized and yields the desired full on / full off/ or intermediate gray shade desired.

But light travelling through the LCD at an angle becomes elliptically polarized and does not fully 'switch' on/off. To compensate for this, the retardation provides a 'negative' amount of elliptical polarization which then results in the light being linearly polarized once it has fully traversed the LCD. The LCD cell gap is 'tuned' for green light (due to various properties if the human visual system), so the technique works best for green. The bi-refringence of the retardation layer is color dependent, so the net result is that some color shift still occurs. And again, the pre-tilt and other asymmetries of the LCD structure cannot be fully overcome.

This might be a dumb response: where is the figure in pink at ? My cell phone has a liquid crystal display, when I look at it straight on, I can see an image or writing. When i view from an angle, I can still see the writing or the image, there is no change in color.

If the figure in pink is only visible on a blank screen & the figure changes to blue when the screen is tilted, that is neato, but why would I want to look at a blank screen ?

Your phone, like mine, most likely has an IPS (In-plane switching) display and you don't see any color change. With IPS displays, the liquid crystal molecules rotate in a plane parallel to the display surface.

The color change with viewing angle is only a "feature" of TN displays (twisted nematic). TN displays are cheaper, needing only 1 transistor per pixel versus 2 transistors for IPS displays, but obviously IPS displays are much superior for color rendition.

Here is a description of how the IPS LCD display works...

"Implementation[edit]

In this case, both linear polarizing filters P and A have their axes of transmission in the same direction. To obtain the 90 degree twisted nematic structure of the LC layer between the two glass plates without an applied electric field (OFFstate), the inner surfaces of the glass plates are treated to align the bordering LC molecules at a right angle. This molecular structure is practically the same as in TN LCDs. However, the arrangement of the electrodes e1 and e2 is different. Because they are in the same plane and on a single glass plate, they generate an electric field essentially parallel to this plate. The diagram is not to scale: the LC layer is only a few micrometers thick and so is very small compared with the distance between the electrodes.

The LC molecules have a positive dielectric anisotropy and align themselves with their long axis parallel to an applied electrical field. In the OFF state (shown on the left), entering light L1 becomes linearly polarized by polarizer P. The twisted nematic LC layer rotates the polarization axis of the passing light by 90 degrees, so that ideally no light passes through polarizer A. In the ON state, a sufficient voltage is applied between electrodes and a corresponding electrical field E is generated that realigns the LC molecules as shown on the right of the diagram. Here, light L2 can pass through polarizer A.

In practice, other schemes of implementation exist with a different structure of the LC molecules - for example without any twist in the OFF state. As both electrodes are on the same substrate, they take more space than TN matrix electrodes. This also reduces contrast and brightness.^[17]

Super-IPS was later introduced with better response times and colour reproduction."

https://en.wikipedia.org/wiki/IPS_panel

Because the human eye can't distinguish pinks/reds in peripheral vision.

The LCD is pink when straight on because it is not blue (but stills contains blue wavelength). The optical path to your eye at an oblique angle with respect to the display is far longer, and you see the absorbance color of the plastic over the LED display which is decidedly blue. Not only this, but the grazing angle provided a different reinforced wavelength of emitted light as viewed obliquely.

Thanks to all users for chiming in to this thread. The answer has been posted.

"...you are viewing light escaping before it has passed through the liquid crystals and polarizing filters..."

I'm not at all sure I can accept that answer as correct. How can light from behind escape without passing through those other layers?

As JS hinted, light traveling obliquely through the liquid crystal follows a longer path than light passing through normally, so is presumably rotated more, causing a different fraction to pass through the second polarizing filter, and at sufficiently wide angles, causing the light to pass through a different sub-pixel of the color filter.

Thanks for posting that the "answer" was available.

Aside from the above, I'd like to know how different intensities of each primary color are achieved. Do different control voltages produce different amounts of polarize rotation, or is it Pulse Width Modulated, or something else?