Blur Busters Forums

Just like the title says people, I just read on wikipedia that LCDs can go as fast as 480hz, is this really true? and does the pixel size matter i.e. can a LCD with smaller pixels have faster refresh rates?

jlafarga wrote:Just like the title says people, I just read on wikipedia that LCDs can go as fast as 480hz, is this really true? and does the pixel size matter i.e. can a LCD with smaller pixels have faster refresh rates?

There's no real defined upper limit, just that the GtG becomes slower and slower when you have less time to inject voltage into pixels.
Blue-phase LCDs can refresh in microseconds, so could theoretically refresh far faster than standard TN / VA / IPS LCDs. However, these are still in the lab.

Did you know? LCD pixel response is actually independent of refresh rate.
We used to have 25ms and 33ms LCDs more than ten years ago -- that's longer than a refresh cycle at 60Hz (16.7ms).
See High Speed Video of LCD Refresh Cycles.

Here's an example of an old 2007 LCD with a real-world refresh that's noticeably slower than a refresh cycle.


When LCDs are too slow, they simply streak multiple refresh cycles into each other's refreshes.
You can have moments where there are three refresh cycles (refreshes displaying "06", "07" and "08") overlapping each other:


So you could theoretically have a 1000Hz LCD with a 33ms GtG.
Or a 60Hz LCD with a 1ms GtG.
GtG has nothing to do with the refresh rate (number of GtG transitions per second, or number of GtG direction changes per second essentially -- as subsequent refreshes can 'interrupt' the GtG momentum, since pixel transitions (GtG) actually physical momentum of LCD crystal (molecules) in motion sandwiched between layers glass. You inject voltage into a LCD pixel, and long after the voltage is finished, the GtG pixel transition is the physical momentum. You could theoretically inject a voltage into an LCD pixel for a picosecond, but you won't get much GtG momentum out of that, unless you raise the voltage hugely. Blue phase LCD's use extremely high voltages (>10 volts) injected into pixels, to speed up transitions. So the limiting factor is how well the LCD can be designed to accept a very high voltage per pixel for the maximum possible momentum imparted into liquid crystal in the shortest possible time, in order to refresh the next pixel (i.e. faster scan out speed). You can also speed things up by multi-scanning the LCD, too (which brings up a host of other problems)

As you can see in high speed video, essentially one pixel is refreshed at a time, but previous pixels are still "fading in" because of the momentum of the pixel transition. So it looks like a fuzzy "wipe" scan from top to bottom.

does the pixel fading time depends on the capacitance of the pixel or the actual liquid crystal?

jlafarga wrote:does the pixel fading time depends on the capacitance of the pixel or the actual liquid crystal?

Good question. I'm not sure. I think both would have elements at play. We're actually moving actual mass (changing orientation of crystals) during LCD grey-to-grey transitions, so I think momentum would be the dominant factor. However, capacitance does appear to play a factor.

When you suddenly stop refreshing an LCD, but keep backlight connected, the LCD retains its image for a few seconds, slowly fading away to its default low-energy state (either solid black or solid white, depending on LCD). That would be the capacitance factor.

like you said, since mass is actually being moved inside the crystal wouldn't a smaller mass be easier to move? hence achieving faster transitions times? do smaller pixels (think retina displays) transition faster or are they driven with smaller voltages or they actually do change faster but compared to larger pixels the difference is negligible?

jlafarga wrote:like you said, since mass is actually being moved inside the crystal wouldn't a smaller mass be easier to move?

Yes, but the wires to each pixel are smaller too, so you can't push as much wattage through smaller pixels. Touché.

That said, there seems to be a point (pixels too big) when things are inefficient and slow, and a point (pixels too small) when things are inefficient and slow. Pixels in large televisions are often covered by a grid of wires to evenly spread voltage over the whole large LCD pixel. The response ratings from the biggest 80" LCD TV, all the way down to the tiniest Retina displays, are all in the similiar single-digit milliseconds now, without too much differences.

Once pixel response is mostly less than half a refresh cycle, it has a fairly negligible effect on motion blur. This is why, nowadays, persistence now makes a bigger difference than GtG response, as shown at 60Hz vs 120Hz vs LightBoost. Milliseconds in persistence is NOT the same as milliseconds GtG. (Whereas I can't tell apart 1ms vs 2ms GtG, my eyes can tell apart 0.5ms versus 1.0ms persistence via Blur Busters Strobe Utility on Z-Series, on TestUFO moving objects at 2000 pixels/second). And all the HDTV makers are adding black-frame-insertion-based technologies to their high end televisions, and the gaming monitor makers are adding strobe modes (LightBoost/ULMB/Turbo240/BENQ Blur Reduction).

cool, btw there's an LG HDTV(I think its LG) advertised as being 480hz, have you looked into that or do you think is marketing bs?