Today's Fastest Retail LCD

An anonymous reader writes "ViewSonic has recently released a very exciting product, a nineteen inch LCD display with a 3ms response time. This is the fastest LCD panel currently available to consumers, and it is clearly aimed at gamers and movie watchers. Dubbed the VX924, the display is part of ViewSonic's X series which tries to comnbine performance with style. The word on the street is that Samsung will have a 4ms display available this year, but this may be the only 3ms."

Comments

Response Measurment

[Bios_Hakr]

I seem to recall some controversy about how response is measured. Some numbers are reported as the time it takes to go from black to white and back to black. Some are reporting just from black to white or white to black. And some are reporting the time it takes to go from one gradient of gray to another gradient.

Buyer beware.

Re:Response Measurment

[dsginter]

I seem to recall some controversy about how response is measured.

There's really no need for the controversy when the stinking refresh rate is well above the pixel response time. Everyone is babbling about how they have great pixel response but then they go and run the monitor at 75Hz (=13ms). When I can run a 3ms monitor at 300Hz, then I will be impressed.

Re:Response Measurment

[Mprx]

A lot of people say very small changes in lag don't make a difference, but I think they are confusing it with reaction time. It's true that you're not physically able to react to visual stimulus in less than about 150ms (reaction to sound is slightly faster, reaction to touch faster still), but we are certainly capable of detecting differences in response times at much finer granularity than that.

This is something I have actually tested. I set up a white square on a black background, moveable using the arrow keys. It was shown on a CRT running at 60Hz, synced to vertical refresh so each frame was exactly 1/60s. I implemented an ABX test, with one of control methods A and B being lagged by 1 frame, and the other not lagged (chosen at random). X randomly matched either A or B, and I could switch between them at will so long as the square was not moving. I tested myself with 15 trials, with no feedback on score until all the trials were complete.

It turned out to be exceedingly difficult to tell the difference, but it was possible. I successfully matched 12 of the 15 trials, and as guessing would produce 50% success we can use the binomial theorem to calculate the probability that I was just lucky.

p = (15 nCr 12) * 0.5^12 * 0.5^3
= 0.014

A statistically significant (p>0.05) result, so it is overwhelmingly likely that 16.7ms change in latency is perceptible. It's subtle, but the "feel" is noticeably different. However, I think the oldness and softness of my keyboard made things much harder, and I'm certain I would not have been able to tell the difference after drinking even a small amount of alcohol. Room for further testing here!

Seriously

[HolyCrapSCOsux]

I have a 19" LCD that I use everyday. Is it THAT noticeable if I have 7-10 ms instead of 3?

Re:Seriously

[pla]

I have a 19" LCD that I use everyday. Is it THAT noticeable if I have 7-10 ms instead of 3?

No, with the condition that the stated time actually measures the real response time (ie, the worst case from any state to any other state). Humans cannot resolve different colors or brightnesses that change faster than roughly 15ms (most people don't even notice changes under 25-30ms, but for some reason, geeks as a group tend to notice flicker far more than the general population).

As my main display, I currently use a 19in DVI panel with a "mere" 12ms response time (note that the "DVI" part of that makes a HUGE difference - Most of the artifacts people blame on poor response time actually come from doing an unnecessary D2A2D conversion). And it looks simply beautiful, even for action movies... No muddiness or ghosting whatsoever.

That said, I don't think any manufacturers measure their response time as a worst-case. So currently, the only real test of how well it will look playing movies or games - Try one out. Go into Best Buy or CC or even Wallyworld, pick out a few models you like based on appearance, then go home and buy your favorite for half the price online.

Pure Commercialism

[xaosflux]

Nothing to see here, not even an A to FR!

Could we at least get a coupon?

Product Announcements Section

[grondak]

All the best product announcements come out on /. Man, if I want to know about a nerdly phone or an LCD monitor that /matters,/ I'm going to be sure to click through to the cool product announcements.

Can we please create a Product Announcements Section and let me turn it off?

That would be the nerdliest way to deal with this stuff: organize it right out of my existance.

Re:Great but....

[Surt]

Unfortunately for you and me, that will be long after they finish fixing the response time issue, because there are a lot more gamers out there than digital photography buffs. Gaming should be fixed at 2ms response times, so they are getting close now. When every LCD maker is pumping out 2ms panels, then we can expect to see them start competing mainly on brightness, then color depth and finally color rendition.

Color depth?

[eagl]

What's the color depth? 6 or 8 bit? I don't care how fast an LCD is... If it shows even a HINT of color banding then it's worthless to me, worse than the crappy used packard bell 15" monitor I have hanging off of my server.

Unfortunately, not many manufacturers are listing color depth in the specs, focusing instead on non-standard claims of response time. There ought to be at least 4 standard measurements - overall brightness, color depth, resolution, and black-white-black response time. Instead, we get resolution, *maybe* a claim of supporting x million colors which could mean anything since they all interpolate to improve image quality anyhow, and a bogus response time number.

The worst part is that so-called enthusiast and gamer hardware review sites let them get away with this. If the color depth isn't printed on the box, the review sites don't even bother to get and report the number. So they're comparing 6 and 8 bit LCDs against each other and not reporting an important difference between the two, or giving great review ratings on monitors without bothering to mention that the monitor only supports a 6 bit color depth so you're guaranteed to get color banding in many situations.

Ok, we admit it... They're ALL fast now. So how about some info on actual image quality?

Re:it's a tradeoff

[Chairboy]

No, I counted them, they're all there.

-- pause for laughter --

BTW, the square root of 16,777,217 is 4,096. What does that tell you about screen resolutions needed to see all 16m colors at once?

Best regards,

7th grade math.

something new to lie about

[Anonymous Coward]

In the days of CRT monitors the only thing that mattered was the size ... and so manufacturers lied about it, and eventually there were lawsuits, and now we have the crazy "19 inch (17.5 inch viewable)" way to describe how big the screen is on a CRT. Thankfully this didn't infect LCD advertizing copy-writers, so when they say "19 inch", it really is that big.

But they have to fudge something. There's no storage in a monitor, so they can't fall back on the old and trusted "100 GB" which is based on 10^9 bytes in a gigabyte, and is the pre-formatted size. Only a few LCD monitors have built-in speakers, so usually they don't have the option to use TMPO watts @ 1KHz rather than RMS across 20Hz-20Khz. So being creative types, they've found that "contrast ratio", and "response time" aren't specified very well, giving plenty of room to put impressive numbers in big type next to the picture in the ad.