Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Post
  • Reply
ToastyX
Mar 15, 2004
N
yaaarrr!

LakesGuzzler posted:

Annoyingly, a significant reason I bought it was for colour accuracy and most of the time I care about that under OSX. Only these Dell monitors have some serious problems under OSX (which I'm certainly experiencing). Pinstripes.. just what I always wanted...

I can reduce the pinstripes a little (or at least move them from one shade to another) by screwing around with the gamma settings, but on the whole I'd currently recommend against this monitor for Macs - including Hackintoshes.
I've seen some people complain about that. Mac OS X seems to do some kind of dithering since I can calibrate without getting banding. My guess is the Dell also does some kind of dithering, which is clashing with the dithering being done in Mac OS X. I haven't tested the Dell to know for sure, but I've seen something similar happen with the Samsung F2380.

If you're using a Mac, I recommend the NEC EA231WMi instead. It costs a little more, but it also has an IPS panel and is similar to the Dell, except without the pinstriping problem.

Adbot
ADBOT LOVES YOU

ToastyX
Mar 15, 2004
N
yaaarrr!

Magic Underwear posted:

Nothing to do with pinstriping, but I think the quote is pretty clear.
That doesn't justify your ridiculous response. There's nothing stupid about using a cheaper IPS monitor for color accuracy, especially if you need something close to sRGB.

In fact, I don't know where you got the idea that the U2410 is better for color accuracy. If you need something close to sRGB, it's actually worse because the excessively wide gamut oversaturates everything except in a small number of color-managed programs, and the sRGB mode is actually worse than a cheaper IPS monitor and is also less functional since the color adjustments are disabled. It's supposed to be factory calibrated, but people have gotten monitors that don't match because it doesn't have uniformity compensation, so brightness and color variations across the panel affect the factory calibration, which defeats the purpose of factory calibration in the first place.

If you really care about color accuracy and need to work with wider gamuts, then you'd want a higher-end monitor like the NEC PA241W, which costs more. Otherwise, you'd be better off with something like the HP ZR24W if you want 1920x1200.

LakesGuzzler posted:

Edit: FYI this sounds interesting/promising http://tech.groups.yahoo.com/group/psychtoolbox/message/10369
If I eventually get it working properly I will retract my non-recommendation for Macs, not that it should be taken as gospel anyway.
That's very interesting, but for some reason they've singled out NVIDIA when I know for sure there's dithering with ATI Macs as well.

I would still recommend the NEC EA231WMi over the Dell U2311H for Mac users simply because I've verified myself that the NEC EA231WMi works on both NVIDIA and ATI Macs without any trouble.

ToastyX fucked around with this message at 12:34 on Dec 14, 2010

ToastyX
Mar 15, 2004
N
yaaarrr!

Number_6 posted:

If anyone knows where I can buy an inexpensive 22", 1680x1050 monitor, with at least 75 hz refresh and overall good image quality, and no discernible input lag, please post it here.
The 2209WA is exactly that, if you can still find one.



DrDork posted:

22" 1680x1050 monitors aren't hard to find. 75Hz though, is...uhh...well it'd be an odd bird, at the very least. Virtually no LCD monitors actually support 75Hz in any meaningful way.
I wouldn't say virtually none. It's true that many LCD monitors don't support higher refresh rates properly, but there are some that can. The problem is hardly anyone tests for higher refresh rates, and many people don't know how to test higher refresh rates properly because you have to add a custom resolution and deal with custom timing parameters while avoiding bandwidth limitations, and sometimes the video card driver can get in the way, so there isn't much data out there. I can only report on the ones I've seen myself. It used to be rare for non-TN monitors, but now there are a few that can do higher refresh rates without skipping frames:

22" Dell 2209WA (IPS, 1680x1050) - 76 Hz over DVI, best response times I've seen in an IPS panel, but seems to be discontinued
23" NEC EA231WMi (IPS, 1920x1080) - 83 Hz over DisplayPort, 74 Hz over DVI, but it doesn't have overdrive (response time compensation), so there's a little bit more motion blur than the Dell 2209WA
23" Samsung F2380 (PVA, 1920x1080) - 76 Hz over DVI, but the response times are visibly worse than the NEC

Someone mentioned they were able to get the Dell U2311H to do 75 Hz, but I haven't tested that one myself.

All those monitors have no significant lag either, which I consider more important than response times.

As for TN monitors, I don't usually test those, so I don't have a big list:

Samsung 226BW (A panel) - 77 Hz, confirmed myself
Dell E2209W - 75 Hz, I saw someone verify with photo evidence

If you're going to get a TN monitor for higher refresh rates, you might as well go for one of the 120 Hz monitors instead.



DrDork posted:

As was born out in the previous thread, response times are basically dirty lies and should (at most) be used as such:
Is the response time >10ms? If so, you'll probably notice ghosting/latency when gaming.
Is the response time <10ms? If so, you may or may not notice ghosting/latency when gaming regardless of what the number is.
Even that's not a good rule of thumb because the NEC EA231WMi is rated at 14 ms while the Samsung F2380 is rated at 8 ms, but the Samsung clearly has more ghosting than the NEC.

DrDork posted:

That is, the difference between 2ms, 5ms, and 8ms amount to measuring techniques and should be ignored. A 2ms panel coupled with a lovely processing chip will perform worse than an 8ms panel coupled with a nice and fast chip. Sadly, there's no way to know which is which without looking at individual reviews.
While the numbers themselves may be bullshit, there's some meaning behind the numbers:

TN:
5 ms = TN without overdrive
2 ms = TN with overdrive
1 ms = TN with bullshit

IPS:
6-8 ms = IPS with overdrive
12-16 ms = IPS without overdrive

VA:
They just make poo poo up.

People say TN panels are the fastest, but TN panels are only fast with certain gray-to-gray transitions. Manufacturers started using gray-to-gray so they could claim lower response times on TN panels. TN panels without overdrive are actually quite slow with dark objects moving against light backgrounds, with some transitions in the 25-30 ms range. Overdrive is designed to help speed up those transitions.

IPS is the opposite. They are faster than TN with dark objects moving against light backgrounds but are slower with light objects moving against dark backgrounds. IPS without overdrive never exceeds 25 ms, and IPS with overdrive brings those transitions down to under one frame, making them about as fast as TN panels.

VA panels are horrible with gray-to-gray transitions, with some transitions exceeding 50 ms. The only way to counteract that is to buffer frames to determine how much overdrive to apply, which is why VA panels with good overdrive have lag.

Basically, if you care about response times, TN with overdrive and IPS with overdrive are the fastest. They are about the same on average with TN being slightly better with gray-to-gray transitions and IPS being slightly better with other transitions.



Edit: I should also mention none of that takes the sample and hold effect into account, which is a much bigger problem that hasn't been properly addressed with LCD monitors. Each frame persists in your eyes for 10-15 ms, and when following motion, the image that persists in your eyes mixes with the image that's still on the screen, causing a blurring effect. CRT monitors don't have that problem because the image is only on the screen for a split second, but LCD monitors keep the image on the screen until the next refresh. Having a higher refresh rate reduces the holding period, but what's really needed is higher refresh rates + overdrive + proper backlight scanning, which no monitor seems to do. That is why even the fastest LCD monitors still have some motion blur compared to a CRT monitor.

ToastyX fucked around with this message at 23:19 on Dec 18, 2010

ToastyX
Mar 15, 2004
N
yaaarrr!

stonewallpatton posted:

Yes. I went through Mac OSX guide to setting up a dual monitor. Adjusted all the necessary preferences (brightness, contrast, gamma, etc.).

Maybe I am expecting too much out of a cheaper monitor in comparison to the iMAC display.


Honestly, in that picture, unless the camera is messing with the white balance, the iMac has way too much blue and a little too much red as well, while the ASUS looks very close to what it should be. The ASUS is slightly yellower on the right side in that picture because it's at an angle, and TN panels get yellower at horizontal angles. The left side is almost spot on. I think it's the iMac that needs adjustment more than the ASUS, but the iMac has no color controls. You'll have to use Mac OS X's built-in color calibrator, which I find a bit limited. The better option would be to buy calibration hardware, which will get them both closer to each other.

The other option is to use the custom RGB color controls on the ASUS (if it has such a thing) to reduce the red and green while leaving a little more red than green. That should get it closer to the iMac. Also, the ASUS seems to be set to a lower brightness than the iMac, so either raise the brightness on the ASUS or lower it on the iMac.

advion posted:

I would say that's about as good as you're going to get the cheaper display to match up to the better quality IPS in the iMac.
That's a pretty ignorant thing to say, especially considering it's the iMac that looks off in that picture. IPS panels don't magically have better colors. That's why calibration is needed regardless of the type of display. It should be possible to get them closer than the picture shows.

ToastyX fucked around with this message at 15:37 on Dec 28, 2010

ToastyX
Mar 15, 2004
N
yaaarrr!

DrDork posted:

Considering that the PA241W is using some sort of P-IPS, while it will be an LG panel, it's unlikely to be the same S-IPS used in the U2410 and ZR24W.
All three are H-IPS. P-IPS is just a marketing term for "Professional" H-IPS, and S-IPS is often used as a broad term for all IPS panels, even though H-IPS is a bit different from older S-IPS panels.

The same applies to e-IPS, which is just H-IPS with cheaper energy-efficient standard gamut backlighting. Recently though, LG has started releasing 6-bit e-IPS panels with temporal dithering, which includes the Dell U2311H. This is why Mac users have had issues with pinstriping on the Dell U2311H because Mac OS X also tells the video card to dither, causing the two dithering patterns to clash. For Windows users, it's not a big deal since the dithering is very good at emulating 8-bit color.

The Dell 2209WA and NEC EA231WMi are 8-bit e-IPS panels, but the 2209WA isn't available anymore, and the EA231WMi was recently discontinued in favor of the EA232WMi, which reportedly has a 6-bit e-IPS panel. If you want a cheap 8-bit IPS panel, I recommend getting the NEC EA231WMi as soon as you can before they disappear, but the NEC doesn't have overdrive, so it has slightly slower pixel response times than the Dell when dealing with darker colors.

ToastyX
Mar 15, 2004
N
yaaarrr!

Alexander Nevermind posted:

Anyone familar with this Samsung monitor: http://www.newegg.com/Product/Produ...-_-24001463-L0C

And how it compares to the U2311Hs? The specs look identical but alot cheaper in price. What's the good word?

Monitor will be used mostly for gaming, FPS and RTS the majority, maybe an MMO.
For gaming, definitely get the Dell. The Samsung is not ideal for gaming due to the ghosting. It's actually the worst possible monitor you can get if you care about response times. Even the NEC EA231WMi which is rated at 14 ms is better than the Samsung.

The Samsung's main strong point is the high static contrast ratio, which makes it ideal for movies in a dark room because of the deeper black, but the viewing angles are worse, and that affects colors more than the contrast does.

The Samsung has a 6-bit c-PVA panel, so Mac users should be aware of dithering issues with that monitor as well. Due to the different dithering patterns, the Samsung clashes worse with ATI-based Macs, while the Dell clashes worse with NVIDIA-based Macs. There's no workaround for ATI-based Macs, but there's a workaround for NVIDIA-based Macs here (solution 5): http://psychtoolbox.org/wikka.php?wakka=BitsPlusPlusPlatformVariations

For Windows users, most people will not notice the difference between 6-bit with temporal dithering and 8-bit if the dithering is done well. You'd have to be looking at gradients to notice the difference, and even then it may be hard to notice. It has almost no impact on how colors look. In my experience, viewing angles, color gamut, and gamma calibration have the biggest impact on colors.

ToastyX fucked around with this message at 07:03 on Mar 4, 2011

ToastyX
Mar 15, 2004
N
yaaarrr!
I'm surprised the DS-305W is still available, or maybe they shipped another batch recently. It's similar to the HP ZR30W in that it's a wide-gamut H-IPS monitor without a scaler. It's probably the best bang for the buck in 30" monitors.

It's basically a rebranded Hazro HZ30W, so you can check out this review: http://www.tftcentral.co.uk/reviews/hazro_hz30w.htm



movax posted:

Think it drops the scaler as well, which isn't a huge deal when it's just being used for PC usage (obviously what the ZR24W is intended for as it doesn't have the many inputs like you said).
The ZR24W has a scaler, but the scaler can't preserve the aspect ratio at 1080p for some odd reason, so it's always stretched to 1920x1200. It can handle other resolutions though. That's only a problem if you plan to hook up a game console or external Blu-ray player and don't want the picture stretched.

ToastyX
Mar 15, 2004
N
yaaarrr!

Nonpython posted:

Does it have gameably low input lag?
It has no lag other than the pixel response times. The same applies to the HP ZR30W and the HP ZR24W, so they're all less than one frame. The Dell U2711 has one frame of lag plus pixel response times.

ToastyX
Mar 15, 2004
N
yaaarrr!

Stew Man Chew posted:



oh what the gently caress! :aaaaa:

Is this normal?
That's the IPS glow. It's normal at extreme angles like that.

The point is the viewing angles on TN and VA panels are bad enough to distort the colors even when you're sitting in front of the monitor, while the colors on IPS panels remain relatively stable. However, light starts to leak through at an angle, and it's noticeable at extreme angles like that. The color of the glow can either be orange/bronze/purple or gray/white, depending on the particular panel being used. There is a way to filter out that leakage, but LG isn't using that filter in any of their current panels for some reason. Only a few monitors had it.

ToastyX
Mar 15, 2004
N
yaaarrr!

Kilometers Davis posted:

I just got my U2410 in. I'm not totally sold on it. The response time makes even browsing feel off and the anti-glare is very distracting from the image itself. Has anyone else felt this way and gotten used to it? I'm coming from a glossy FHD2400 so it's pretty much a totally different monitor.

e: I played a bit of Crysis 2 and it's extremely hard for me to aim. I hope I get used to this so I don't have to eat a restocking fee.
Try game mode, which reduces the total lag to under one frame, but that messes with the colors.

As for the anti-glare coating, the only real alternatives with IPS panels are the 27" Apple Cinema Display and the previous 24" Apple Cinema Display (if you can find one), both of which have less than one frame of lag, but they require a DisplayPort video card.

Otherwise, if you want a glossy monitor with the best responsiveness for gaming, get the 120 Hz ASUS VG236H.

ToastyX
Mar 15, 2004
N
yaaarrr!

GreatGreen posted:

Do shiny screens tend to have deeper black levels or am I completely making that up?
Glossy screens reflect light instead of diffusing it, so black can appear deeper in a lit room, but in a dark room, it makes no difference.

ToastyX
Mar 15, 2004
N
yaaarrr!

DrDork posted:

No, the U2311H, U2410, U2711, and U3011 all have lag under 20ms. You're probably thinking of the 2408wfp, which did have lag in the 40-50ms range. The older you go, the worse the lag usually gets.
For reference:

The U2311H has no significant lag (< 3 ms) other than pixel response times.
The U2410 has one frame of lag (17-20 ms) except in game mode where it has no significant lag other than pixel response times.
The U2711 and U3011 have one frame of lag, and game mode has no effect.

DrDork posted:

If you're not some pro gamer who feels he lives and dies by single frames, it's unlikely you will ever notice the (very small) performance difference between a TN and the U2311H in game mode.
Game mode has no effect on the U2311H. It's not needed anyway since there's no lag to reduce.

ToastyX
Mar 15, 2004
N
yaaarrr!

Goodpart posted:

The other problem is that a lot of monitors are very inconsistent. Some panels have varying lag, from anywhere between 0ms to about 25ms. That means that at any given point you could be bang on the mark, or up to 2 frames off. Very frustrating.
It's not possible for lag to vary like that. Most LCD monitors lag within 3 ms of an exact number of frames with no fluctuation. It's the testing methods that are flawed.

Goodpart posted:

I did some quick searching and found a benchmarking for the Viewsonic, which showed anywhere between 0-2 frames of lag. So I suspect that's out.
I wish people would stop using that lovely Flash-based FlatpanelsDK timer with clone mode. That timer is not guaranteed to update at any particular interval, and clone mode is not guaranteed to be synchronized. LCD monitors also refresh from top to bottom like a CRT, but you can't see where the LCD is refreshing with a single timer. The sample and hold nature of LCD monitors also makes the numbers jumble together, making it hard to see the new number coming in. All of those problems combined produce wildly fluctuating results. That ViewSonic would probably show no more than 3 ms of lag when tested properly.



Rexides posted:

What about response time? I am reading things like 2 and 5ms for TN monitors, and they seem unnoticeably small. Plus, the U2412M with 8ms on gaming mode doesn't seem that much behind. By the way, does anyone know what the response time for this monitor is on normal mode, or it doesn't really matter?
Game mode doesn't do anything special on most monitors. The U2410 is the only exception I know about, where it gets rid of the one frame lag. The U2412M doesn't have any significant lag, and response times shouldn't be an issue.

ToastyX
Mar 15, 2004
N
yaaarrr!

DrDork posted:

I'd be more suspect if their methodology hadn't been the same for quite awhile, and (up to now) pretty much always provided predictable, sensible results that meshed well with other sites and with actual user experiences. Put another way, if their methods were poo poo I'd have expected someone would have noticed and called them out on it 50+ monitor reviews ago.
I keep having to explain to people that the most common lag testing methods are not accurate enough to determine an exact amount of lag, but people are so used to inaccurate results that it's an uphill battle to fight all the misinformation.

Most monitors lag within 3 ms of an exact number of frames, and many monitors don't buffer any frames at all, so results like this are not unusual. People should focus more on how many frames are being buffered by the monitor to determine the lag rather than rely on dubious ms values from unreliable testing methods.


sethsez posted:

That's low to the point where I have to question the result.
The funny thing is that's the most accurate result in that list.

All of the monitors in that list under 12 ms (and even some above that) actually have no more than 3 ms of lag, not including pixel response times. I know for sure that the Dell 2209WA is less than 3 ms, but they have it listed as 9 ms.

All of the 22" and 23" Dells and the U2412M should be no more than 3 ms. The U2410, U2711, U3011 are all one frame (17-20 ms), except the U2410 has a game mode that gets rid of the one frame lag. The game mode on the other models doesn't do anything special.


HalloKitty posted:

I'm sure they were amazed as you and probably did it a few times.

Robolizard! posted:

I'm sure sure they've tested for any difference in latency between the two.

movax posted:

I'm sure they checked that result several times
You guys put way too much faith in these reviews.

I bet if you use a DVI splitter to compare the U2312HM vs. the U2311H, every shot would show that both monitors are virtually identical. There is no 10 ms difference.


HalloKitty posted:

I'm not really sure how you can balls up a test that is surely just a picture being taken with the TFT and CRT visible, just measuring the timestamp differences.
Oh, let me count the ways:

1. People keep using timers that aren't guaranteed to update at any particular interval, especially that FlatpanelsDK Flash-based timer or that Lagom Javascript timer.

2. Clone mode is not guaranteed to be synchronized. It IS possible to do an accurate test using clone mode but not with standard stopwatch programs.

3. LCD monitors refresh from top to bottom like a CRT, but there's no way to see where the LCD is refreshing when using a single timer in one part of the screen.

4. The sample and hold nature of LCD monitors causes the numbers to jumble together, making it harder to see the new number coming in.

All of those problems combined produce wildly fluctuating results, which are then averaged to produce an unreliable estimate. If the test was accurate to begin with, there would be no fluctuation at all.

Prad.de recently came up with a very accurate testing method using an oscilloscope. They determined that the U2412M only has 0.9 ms of lag, which TFT Central lists as 9.4 ms. Prad also determined that the LG IPS231P only has 1 ms of lag, which TFT Central lists as 6.9 ms. The LG actually has slower pixel response times than the U2412M, so TFT Central's numbers don't make sense even if you factor that in. Prad also determined that the HP ZR22W only has 0.7 ms of lag, which shows how common these results really are. Unfortunately, Prad only recently started using the oscilloscope method, so many of their older reviews are not accurate. They also made a lag testing tool that could be used with clone mode, but ironically, most of their reviews didn't even use it properly, and it actually exacerbated the problem of numbers jumbling together.

The most accurate test that doesn't require special equipment is this bars program: http://hardforum.com/showthread.php?t=1423433

That program will let you see where each monitor is refreshing, and nothing jumbles together because each bar is a separate frame. When using that program with clone mode, it will synchronize with one of the monitors, and the other monitor might tear if both monitors aren't perfectly synchronized. If you wait for the line of tearing to roll off the screen before taking a picture, the result will be very close to accurate.

I got someone to use that program to test the HP ZR24W against a CRT. Here is a good shot: http://toastyx.net/zr24w-lag.jpg

The rightmost bar is the current frame, and each frame takes about 16 ms to refresh at 60 Hz. That shot is good because it shows the CRT is only about 5 ms into refreshing the new frame, and the ZR24W is already starting to show at least half of that, which means the lag can't be more than 3 ms. TFT Central has it listed as 10 ms. The fading in is caused by pixel response times, which is the only real delay on that monitor.

ToastyX
Mar 15, 2004
N
yaaarrr!

movax posted:

Where do they detail this, I'm curious to what they did. If it's simple enough, I want to try it on my U3011.
I'm not sure how they do it, but they first mentioned using an oscilloscope in this huge article about lag: http://www.prad.de/en/monitore/specials/inputlag/inputlag.html

ToastyX
Mar 15, 2004
N
yaaarrr!
Apparently, the HP ZR2440w supports 1:1 pixel mapping. It's the replacement for the HP ZR24w and probably has the same panel has the U2412M.

http://h10010.www1.hp.com/wwpc/us/en/sm/WF05a/382087-382087-64283-72270-3884471-5163690.html

quote:

1:1 scaling supports full HD 1080p letterboxing.

ToastyX
Mar 15, 2004
N
yaaarrr!

Alereon posted:

The monitor itself will probably look fine, it's so cheap because it doesn't support HDCP copy protection, meaning you can't use it to watch Blurays, hook it up to a TV box, or otherwise use it for watching protected content.
I'm pretty sure that's not true. What monitor doesn't have HDCP support these days?

ToastyX
Mar 15, 2004
N
yaaarrr!

Alereon posted:

The Detailed Specs specifically says it doesn't support HDCP. They would be able to easily disable it via the firmware and not have to pay the per-monitor royalties, so that may explain its absence. I suppose it COULD be an error but I don't see anything contradicting those specs.
NewEgg specs aren't always accurate. I checked the user reviews and a couple of them mention using it with the PS3, which requires HDCP support when using HDMI.

ToastyX
Mar 15, 2004
N
yaaarrr!
The U2412M is not 9.4 ms. That's a rough estimation based on averages from an unreliable testing method. The U2412M has no real lag other than pixel response times.

Lag doesn't fluctuate, and lag is usually within 3 ms of an exact number of frames. The U2412M and U2312HM both have 0 frames of lag.

ToastyX
Mar 15, 2004
N
yaaarrr!

dud root posted:

Overall I'm going to buck the trend and say I'm disappointed with it- games where you pan left to right such as TF2 or KOTOR quite simply have what appears to be a low frame rate. I'm using a GTX580 & have transitioned from a 1600x1200 CRT @ 85Hz. The difference in game smoothness is night/day.
That's every LCD though. The problem is a combination of 60 Hz plus sample and hold (lack of flicker).

The main point is the U2412M is not bad for gaming when compared to other LCDs, but it will not match a CRT. The closest you can get to a CRT in an LCD is a 120 Hz LCD.

ToastyX
Mar 15, 2004
N
yaaarrr!
The internal TMDS on the FX 5000 series GPUs could not handle the full DVI bandwidth. Only cards with an external TMDS chip could handle 1920x1200.

ToastyX
Mar 15, 2004
N
yaaarrr!

movax posted:

Updated with a blurb on the Catleap, can you guys give it a once-over to make sure I didn't miss anything?

It's kinda lost in there, I'll PM a mod to change our title again to draw more attention to it. I'm also curious to hear about experiences with 3rd party stands for the Catleaps; Monoprice stands proved...subpar to hold up my U3011.

Also, does the stock stand pivot? I want to get one for my parents, but if I love it enough, I might replace one of my 2209WAs with it.

e: also since I don't even know, quick breakdown of Achieva vs. Shimian vs. Crossover would probably be good too...

The main difference between the three is the stand.

Catleap - This monitor has the worst stand, and you have to take the monitor apart to remove the neck of the stand and the useless plastic arch that covers the VESA holes. It wobbles, and some of them lean. It also sits higher than most monitors, like around iMac level, and there's no height adjustment.

Achieva - This monitor sits lower, and there's still no height adjustment. You can easily remove the stand without taking the monitor apart, and the VESA holes are easily accessible.

Crossover - The LED-P version has the best stand with all the adjustments: height, tilt, swivel, and rotate (pivot).

In terms of performance, they're all the same except the earlier Catleaps made in February can do 85-100 Hz while the current ones can't.

All of the monitors are glossy. The Achieva and Catleap both come in glass and non-glass versions, but several people have reported getting dust behind the glass on the glass models. There's really no point in getting the glass versions since both versions are glossy. That's just one more layer for them to screw up.

Animal posted:

I'm sure it's not glossy. It's matte. There's a big difference between glossy and matte so I'm pretty sure. I won't take a picture, I'm going to bed. You'll just have to take my word for it.
If you have a matte version, then you must take a picture for the sake of mankind! It's not supposed to be matte, and nobody else has reported getting a matte version.

spanko posted:

The difference between this and my ZR24W is pretty huge. I guess it isn't glossy though and I'm disappointed in that. I thought they were all glossy. It is WAY clearer, and there definitely isn't any AG coating on it.
That sounds like glossy to me.

ToastyX
Mar 15, 2004
N
yaaarrr!
The U2412M doesn't have 9 ms of lag. It's the same as the U2312HM.

ToastyX
Mar 15, 2004
N
yaaarrr!
They don't know how to test lag properly. I thought we cleared this up several months ago: http://forums.somethingawful.com/showthread.php?threadid=3372494&userid=0&perpage=40&pagenumber=53#post396262199

U2412M: http://www.prad.de/en/monitore/review/2011/review-dell-u2412m-part13.html#Latency
The latency is determined as sum of the signal delay time and half of the average frame. Key for gamers, we measure extremely short 0.9 milliseconds for the signal delay.

U2312HM: http://www.prad.de/en/monitore/review/2011/review-dell-u2312hm-part13.html#Lag
The lag is measured as the total of the signal delay time and half of the average frame change time. We measured the signal lag, which is important for gamers, at am extremely short 1.1 milliseconds on the U2312MH.

ToastyX fucked around with this message at 11:06 on May 7, 2012

ToastyX
Mar 15, 2004
N
yaaarrr!
If you want to include pixel response times, the numbers still don't make sense. The U2312HM does not have a total response time of 0.6 ms under any circumstances.

U2312HM: When half of the average frame change time of 8.2 milliseconds is added, the average overall lag amounts to 9.3 milliseconds

U2412M: Another 6.5 milliseconds later, the target brightness is reached. Being at 7.4 milliseconds, the mean total latency is very short

The difference between the two is about 2 ms when measured by Prad.de, with the U2412M actually being faster than the U2312HM.

Realistically, there isn't a significant difference between the two.

ToastyX
Mar 15, 2004
N
yaaarrr!
PLS is just Samsung's version of IPS. You can literally get four Korean IPS monitors for the price of that Samsung.

ToastyX
Mar 15, 2004
N
yaaarrr!
Don't bother with the IPSMS. It has one frame of lag, and it doesn't seem to have a proper backlight brightness control, which means you can't lower the brightness without hurting contrast and possibly crushing blacks. I would avoid that model unless you plan to run it at full brightness and don't care about the lag.

The single-input versions don't have those problems, but I'm not sure if you can use them with game consoles. I've read reports that some of the single-input monitors can accept 720p with pixel doubling, but my 2B Catleap can't. The newer Catleaps and the other models have a different timing controller, so maybe they have that capability, but I don't have one to verify.

ToastyX
Mar 15, 2004
N
yaaarrr!

DrDork posted:

Yeah, but your monitor is only 1920x1080, while the Korean ones are 2560x1440. Big difference. You're free to speculate as to why NVidia caps a lot of their cards' refresh rates at high resolutions, whereas ATI does not, but for the moment it's certainly something to consider if you're lusting after the concept of a 120Hz 2560x1440 screen.
They both cap the pixel clock. It's a long story:

AMD's driver caps the pixel clock to 330 MHz with dual-link DVI, which is only enough to get 81-85 Hz at 2560x1440, depending on the timing parameters used.

NVIDIA's driver originally allowed up to 400 MHz with 500-series (Fermi) and older cards, which was enough to get 97-102 Hz at 2560x1440.

Then the 600-series (Kepler) cards came out, and the driver allowed those to go over 500 MHz, which was more than enough for 120 Hz, but SLI was still limited to 400 MHz.

We tried to get both AMD and NVIDIA to remove those limits, but it was fruitless.

June:

I modified AMD's driver to allow higher pixel clocks, which allowed pretty much any AMD/ATI card made in the last few years to get at least 110 Hz.

Then for some reason, NVIDIA capped the pixel clock to 330 MHz starting with the 304.48 beta driver. 301.42 was the last version that allowed higher pixel clocks.

August:

The article was written at this point, but it did not mention that you had to use a modified driver to overcome the limits with AMD/ATI cards, nor did it mention that the older NVIDIA drivers allowed higher pixel clocks.

September:

I got an NVIDIA card and modified the driver to remove the 330 MHz limit as well as the 400 MHz limit for SLI. Then I found a way around the 400 MHz limit for older cards.

Now pretty much any AMD/ATI and NVIDIA card made in the last few years can get at least 110 Hz with these monitors by using the modified drivers.

All of this information was posted on 120hz.net. I wrote a quick start guide here: http://120hz.net/showthread.php?683-Overclocking-Quick-Start-Guide

The only problem is modifying AMD's driver breaks HDCP support and video acceleration (DXVA), but video acceleration can be disabled in Flash and most video players. NVIDIA's driver doesn't have that problem.

The only way we'll ever see 2560x1440 @ 120 Hz monitors from big name brands is if someone develops the hardware to do it over DisplayPort because they're not going to support using modified drivers. Ironically, the monitors that have the DisplayPort timing controller are the ones that can't overclock.

If you want to know more about the various limitations with dual-link DVI, I'll quote what I posted here: http://overlordforum.com/viewtopic.php?f=8&t=65

ToastyX posted:

Some people were claiming 2560x1440 @ 100-120 Hz is not possible with DVI because they didn't understand where the limitations were coming from. They didn't even test one of these monitors to verify what they were saying.

Let's start with the official DVI standard. The DVI standard states that single link implementations must be limited to 165 MHz or less, and anything greater than 165 MHz must be split across two links. Unfortunately, many people have misinterpreted that to mean each link is limited to 165 MHz for a total of 330 MHz with two links, but that's not true. In fact, the DVI standard specifically states, "there is no maximum for the dual link implementations."

That leaves four possible sources for limitations:

1. Cable
2. Monitor
3. Video card
4. Video drivers

Let's start with the cable. Most dual link cables are designed to handle at least 330 MHz, which effectively means 330,000,000 pixels per second. TMDS uses 8b/10b encoding, which means 24-bit color is transmitted using 30 bits per pixel. That's where 7.92 Gbps and 9.9 Gbps comes from. 7.92 Gbps (330,000,000 * 24 = 7,920,000,000) is the amount of data transferred, and 9.9 Gbps (330,000,000 * 30 = 9,900,000,000) is the amount of bandwidth used to transfer that data. That is the MINIMUM. Most cables can handle more than that. If you reach the limits of what the cable can handle, you will start seeing artifacts like flickering lines or colored pixels. It will not lose frames.

The amount of bandwidth used includes blanking, which includes the period between each line and each frame that isn't shown on the screen. 2560x1440 @ 60 Hz is typically sent as 2720x1481 @ ~59.95055 Hz to make the pixel clock 241.50 MHz, so 2720 * 1481 * ~59.95055 = 241,500,000 * 30 = 7,245,000,000 = 7.245 Gbps

Now on to the transmission of the data from the video card to the monitor. The video card sends a constant stream of data. At higher refresh rates, the data is sent faster. It's not possible for frames to get lost. If the monitor or the cable can't handle the faster data rate, the monitor won't be able to process the data and you won't get a usable picture. Some monitors do process the data and skip frames at higher refresh rates, but I've already proven that's not the case with the single-input models (see pictures in the original thread).

As for the monitor itself, the main board with the DVI input has a DVI receiver that is designed to handle at least 450 MHz, which is enough to get 109-114 Hz, depending on the timing parameters used. That's not quite enough for 120 Hz, but the receiver can usually handle more than designed. The receiver takes the DVI input and sends the data to the timing controller board using quad channel LVDS output, which can easily handle the data rate. The timing controller then sends the data to the panel. The LVDS timing controller made by LG is the one that can handle high refresh rates. The monitors that can't overclock have a DisplayPort timing controller made by Parade.

Now on to the video card. Most video cards made in the last few years have TMDS transmitters that can handle at least 225 MHz per link. That also gives us 450 MHz with two links, but most TMDS transmitters can handle a little more than designed, and some newer ones can handle 300+ MHz per link. 120 Hz requires around 468 to 497.76 MHz, depending on the timing parameters used.

The biggest obstacle has been the video drivers. Even though the hardware is capable, both AMD/ATI and NVIDIA have placed limits in their drivers:

  • AMD/ATI's driver limits dual-link DVI to 330 MHz, which is only enough to get 81-85 Hz, depeneding on the timing parameters used.
  • NVIDIA's driver originally allowed 500+ MHz with 600-series cards, but SLI and older cards were limited to 400 MHz, which is still enough to get 97-102 Hz, depending on the timing parameters used. After 301.42, they implemented a 330 MHz limit for all cards.

By now, I have managed to remove most of the limits from both drivers. Patchers are available here:


See the Quick Start Guide.

ToastyX
Mar 15, 2004
N
yaaarrr!

chiyosdad posted:

So I bought one of those Korean IPS monitors everyone's been raving about (a FIRST FSM-270YG). I'm trying to hook it up to my laptop (using a hdmi-dvi connector, because the monitor only has dvi ports), but it's not working (the monitor is blank and the light is red instead of blue). I tried pushing the projector button to either mirror the monitor or extend my desktop, but this has no effect. Both the monitor and the cable are definitely working because the monitor shows up under devices, it just doesn't display anything. I am wondering if my graphics card is incompatible with the monitor? I have a Radeon HD 6630M. Thanks in advance.
It might be possible to use HDMI with my driver patcher: http://www.monitortests.com/forum/Thread-AMD-ATI-Pixel-Clock-Patcher

The problem is 2560x1440 @ 60 Hz requires a 241.50 MHz pixel clock, which is possible with dual-link DVI, but HDMI is single link. It's technically possible to push the native resolution through a single link, but the driver limits single-link connections to 165 MHz. The patcher modifies the driver to overcome that limitation.

Keep in mind the monitor was never intended to run that way, so it might not work. Also be aware of the video playback issues described under known issues.

The only other option is to use a DisplayPort to dual-link DVI converter if your laptop has DisplayPort (or mini-DisplayPort). Monoprice.com has the cheapest one I know: http://www.monoprice.com/products/product.asp?p_id=6904

I don't know of any HDMI to dual-link DVI converters, but that wouldn't work anyway because you would still run into the same driver limitations.

ToastyX
Mar 15, 2004
N
yaaarrr!

Kilazar posted:

I am looking at a PC refresh for my wife and I. And was thinking of buying the recomended Asus VG236H. But I don't see any that actually advertise that these are 120mhz refresh.
...
Now the VG236HE does show 120. Is this perhaps a typo in the suggestion post?
http://www.newegg.com/Product/Product.aspx?Item=N82E16824236184&Tpk=VG236H
They are both 120 Hz. The VG236HE is just the VG236H without the 3D Vision kit.



Kilazar posted:

I am also looking at http://www.newegg.com/Product/Product.aspx?Item=N82E16824236305
Unfortunately there is no feedback on it. Does anyone have an opinion on this particular unit?
That's just a standard 60 Hz TN monitor. I don't see the point of getting that one.

ToastyX
Mar 15, 2004
N
yaaarrr!
It's not the backlight. It's delamination of the panel, a common issue with older LG IPS panels.

ToastyX
Mar 15, 2004
N
yaaarrr!

veedubfreak posted:

Well I've asked pretty much every where else, might as well ask here also.

I have a GTX 690 and 3 30" monitors.
1 is an old HP Lp3065 and the card sees it as native 2560x1600
1 is a new Dell u3011 and the card sees it as native 2560x1600
1 is an old Dell 3007 and the loving card sees it as native 1280x800 yet will still display to it properly at 2560x1600.

Due to this weird glitch with what the card thinks is the native resolution, I can not get the nvidia surround to span to all 3 monitors. Has anyone run into this sort of thing before? Is there some sort of internal scaler that the monitor uses that confuses the driver? Is there anyway to make the monitor show up as native 2560x16000?
It says the native resolution is 1280x800 because that's what the monitor reports as the primary resolution. They probably did that to make it default to 1280x800 so it would display properly when connected with a single-link DVI cable. That shouldn't be a problem by itself, but NVIDIA Surround can be stupid about monitors that don't match exactly. Some people had trouble with monitors where the sync polarities didn't match, which shouldn't even matter.

You can use my Custom Resolution Utility to override that information. 1280x800 is probably listed as the first detailed resolution for the 3007. You want 2560x1600 to be the first. You can select a resolution and use the arrow buttons to move them around.

Even better, remove all the resolutions except 2560x1600 and make sure the extension block is disabled so there's no confusion about which resolutions to use. Do this for all the monitors to make them match. To make this easier, you can just import this file for each monitor: http://www.toastyx.net/2560x1600.bin

After you're done editing, click OK to save the changes, and reboot. Then you should be able to enable NVIDIA Surround properly.

ToastyX
Mar 15, 2004
N
yaaarrr!
You can run a 120 Hz monitor at 100 Hz. 1920x1080 @ 100 Hz is around the same amount of pixels per second as 1680x1050 @ 120 Hz. LightBoost works at 100 Hz as well.

Adbot
ADBOT LOVES YOU

ToastyX
Mar 15, 2004
N
yaaarrr!

Ak Gara posted:

That reminds me, when I use lightboost I loose my audio over GPU > hdmi > AVR > speakers. :confused:
How are you using LightBoost with HDMI?

  • 1
  • 2
  • 3
  • 4
  • 5
  • Post
  • Reply