This would probably be a good place to review what it means to compare LCDs and CRTs in ms vs Hz. Also, what the "2ms, 4ms, 8ms, 12ms" garbage is all about.
First off, I imagine most people here started their gaming lives off on CRTs and good ol' fashioned TV with Nintendos, the Sega Genesis, or maybe for the younger crowd the PlayStations, N64s, or maybe even the original XBox. Regular TV, and most CRTs had refresh speeds of 60Hz, and from my perspective, that was plenty (though granted, games have changed a LOT since then)
Some people (see: mpilchfamily or Siggsy) still swear by the old CRTs. But here's some evidence that seems to support tossing pretty much any CRT into the trash:
Equivalence
60Hz = 17ms (actually just a hair under 17)
75Hz = 13ms (actually just a hair over 13)
Um, what? So, really, a 75Hz monitor is only as good as a 13ms refresh speed LCD? Answer: Yes. Sort of.
Understanding the difference between the two types is important for understanding why faster LCDs (read: lower ms) aren't always better and CRT has still managed to hold its own as far as gamer following even though LCDs are obviously listing speeds far exceeding the capabilities of CRT (see Viewsonic's VX922 or whatever has the 2ms response time, which is the equivalent of 500Hz)
First off, CRT. CRTs refresh by generating lines on a screen. If you've ever seen a video of a regular TV on broadcast TV, it looks funny, like there's a big bar of image slowly travelling down the screen. This has to do with the frequency of the recording and frequency of the TV being out of synch. If you have really fast-registering vision and stand back a bit, if you look at a CRT monitor, you can actually see the screen being refreshed, line by line, top to bottom. This is why CRTs are measured in Hz, not ms. Every individual screen is refreshed, so you actually see 60 images per second or 75 images per second (TV is 30 frames/sec and DVDs are really only 24, even though sometimes you'll see 48 frames per second listed. This is because every frame is displayed twice, so even though there are 48 images being shown, half of them are duplicates - this is why moving camera angles look so crappy on detailed images). The nice thing about this is that every time you see a pixel, it's a fresh one.
Now for LCDs. The image is generated in a much different fashion. Whereas CRTs refresh every pixel at once (well, almost at once - line by line, filling the screen on each shot), LCDs refresh an image one pixel at a time. Each pixel is generated with a set of little lights that the monitor changes when necessary. So, for a completely white screen, a CRT would keep generating a white screen over and over again, while an LCD monitor would just turn all pixels on white and leave them alone until they needed to be different. And in contrast to CRT which measures screen refreshes, since LCDs don't ever
do a screen refresh unless all pixels change at once, it uses miliseconds to measure how long it takes an individual pixel to change. And this is where the confusion begins.
Pixel swapping in LCDs isn't quite as simple as on/off. Each time the light changes, the pixel needs to "warm up" to the new color. To explain, I'll analogize it to running a shuttle run (think high school gym class - you started on the middle line, ran 10 yards one way, picked up the eraser or touched the line, ran 10 yards the other way, dropped off the eraser or touched the other line, and then ran back to cross the line in the middle to finish). LCDs make colors by mixing options of blue, red, and green light. White is a mix of all 3, black is all off. But let's say there's a dark color in a game: you don't want to turn the lights on too high, because your color will be wrong. For instance, if you're displaying an image that has to go from purple to green, your monitor has to take that poor pixel that has its red and blue lights on, turn them off, and fire up the green light. It takes time, both for the blue and red to turn off, and for the green to turn on. Say the purple is like running for the one line, and the green is the complete opposite direction. You have to slow down going the one way, change directions, and accelerate over to the other line in your shuttle run.
This is how LCDs manufacturers measure their response time and you end up with absolute trash readings like 2ms. The response time listed for LCDs is only for the black-to-white (0 to 255) transition. This is the same as going from being still at the starting point in your shuttle run and accelerating as fast as you can for the one line, without worrying about having to stop or slow down. We're talking full stop to full throttle. To make matters worse, the posted monitor speed isn't even the value to get all the way to the 255 display. Just 80% brightness or so. And to compound the matter, Overdrive has increased in popularity (not a bad thing in the long run, mind you), which is basically a way to get that 0-255 time down farther by trying to WAY overshoot the desired color and then when you get close, stop trying to produce such brightness. This would be the same as putting the pedal to the metal in your car to get up to speed as fast as possible in a 25mph (about 40km/hr) speed zone. There's no reason for that. Pedal to metal is like shooting for 100mph (160kph) but then pulling up short. This wouldn't be a bad thing if it were accurate, but what ends up happening is that your monitor (only if it has Overshoot technology) displays the wrong color for a while before it eventually balances out and gives you the right one.
So why bother with this garbage if it gives you the wrong color? Well, in the end, when the manufacturers finally get a handle on Overshoot (read: accelerate fast, but don't actually overshoot the desired color), it will be a means of showing colors faster. But there's still a major advantage for right now. By enabling this mechanism, manufacturers can put a label on their box with a super-low timing, even though that's not really the case. Remember, you only need to achieve 80% of desired brightness on a 0-255 pixel transition for your monitor's official timing. Back to your shuttle run: if you only cared about getting from the start line to one side, and didn't really care about the time of your full shuttle run, you could just run pell-mell for the one line, similar to the way a baseball player will run out first base and run through the bag, since he can turn around and still be safe, whereas going from first to second, you can't overshoot the bag and you have to be under control. So many times what you'll see is that a manufacturer using Overdrive has a great response time listed, but the actual quality of each generated pixel is somewhat lesser. For example, the 0-200 response time might be 25ms, even though the response time for 0-255 (what they put on the box) is 4ms.
Tom's Hardware has some wonderful reviews on monitors, including data on the grey-to-grey transitions. Definitely worth a look. The graphs give a nice visual to explain what is
really going on with monitor timings.
This is why people still like CRTs, even though 8ms monitors are commonplace (and the equivalent refresh rate for an 8ms monitor is technically 125Hz, which would be far, far more than enough for gaming since most video cards can't handle that kind of frame rate for rich textured games). The times listed on the box don't accurately describe the time it takes for one color to change to another color; especially at less-than-full brightness, which requires a bit of restraint.
Even the VX922 (a fine monitor, and generally considered the benchmark for quick response time monitors) is really nowhere close to the fabled 2ms response time listed on the box. While faster by far than most competitors, the VX922 still has a response time of about 10ms for a 0-150 transition (black to a mid-grey). Link to the Tom's review =
http://www.tomshardware.com/2006/03/27/the_spring_2006_lcd_collection/page35.html
You'll also note that ViewSonic sacraficed some display uniformity for that fast timing.
The bottom line is that yes, the faster the timing, the faster the monitor refresh (except when Overdrive is used poorly:
http://www.tomshardware.com/2006/03/27/the_spring_2006_lcd_collection/page3.html and that's supposed to be a 6ms monitor - Overdrive is responsible for that big hump). But you also want to be cognizent of color balance, uniformity of display (think: halos), and ergonomics and other considerations before you go out and pick up that 1337 h4x 2ms monitor. And keep in mind, if you're getting transitions under 8ms, you won't notice unless that 8ms is a lie. Which it probably is.
I hope this helps y'all. I strongly recommend checking out Tom's "Graphics and Displays" section to understand a little more about what to look for in a monitor. The Spring 2006 Collection article on 19" monitors is a good place to start, and there's plenty more if you're in the market for something else (20", wide panel, etc).
Facebook
Twitter
Instagram