A Guide to Choosing the Right 19" LCD Monitor - 7 Models Reviewed
by Kristopher Kubicki on November 30, 2004 12:04 AM EST- Posted in
- Displays
Subjective Analysis (continued)
Generally, we follow up our application analysis with a table pitting two LCDs against each other using benchmarks that we derived from VESA's handbook and test patterns from DisplayMate/CheckScreen. All of our monitors are running on an X800 Pro AGP over a DVI connector unless denoted otherwise. Monitors that don't have DVI connectors are compared using a D-sub connector instead. We have simplified our table a little bit so that we can fit all of our data on the same page, but our Notes From the Lab section flags any behavior that we would typically note on the table. The table ranks each benchmark on a scale of 1 to 5, 5 being the best rating and 1, the least desirable.Here is generally how we rate a category:
5 - Outstanding; we have not seen anything to date that could rival our impression of this monitor's performance.
4 - Good, but room for improvement. There are units on the market that perform better.
3: Average; this monitor performs well enough to maintain the status quo, but does not excel.
2: Improvement needed; this monitor performs poorly in performance of this category.
1: Unacceptable; this product does not pass even basic performance requirements.
| DisplayMate / CheckScreen / VESA FPDM 2.0 | ||||||||
| BenQ FP931 | NuTech L921G | Planar PE191M | Samsung 193P | Samsung 910V | Sony SDM-S94 | ViewSonic Q190MB | Dell 2001FP | |
| Intensity Range Check | 4.5 | 4 | 4 | 5 | 4 | 4 | 4 | 5 |
| Black Level Adjustment | 4.5 | 5 | 4.5 | 5 | 4.5 | 5 | 5 | 4.5 |
| Defocusing, Blooming, Halos | 5 | 5 | 5 | 5 | 4 | 5 | 5 | 5 |
| Screen Uniformity and Color Purity | 4.5 | 5 | - | 5 | 4 | 4 | 5 | 4.5 |
| Dark Screen Glare Test | 3 | 4 | 4 | 4.5 | 4 | 4 | 4 | 4.5 |
| Primary Colors | 3 | 4 | 4 | 4 | 3 | 3.5 | 4 | 4 |
| Color Scales | 3 | 4 | 4 | 4 | 4 | 4 | 4 | 4 |
| 16 Color Intensity Levels | 3 | 4.5 | 4.5 | 4.5 | 3.5 | 4.5 | 4.5 | 4.5 |
| Screen Regulation | 5 | 5 | 5 | 5 | 5 | 5 | 5 | 5 |
| Streaking, Ghosting | 5 | 5 | 5 | 5 | 2 | 4 | 5 | 5 |
Notes From the Lab
Above, you can see the Planar PE191M only scored N/A on its screen uniformity test. While the monitor is quite good, the cluster of defective pixels really strained our eyes and we found it fairly difficult to get an accurate reading of the screen uniformity. When we look for screen uniformity errors, we usually look on the corners and sides where light may be seeping through from the backlight around a poorly insulated edge. Since the dead pixels were in the corner, we had trouble when attempting to inspect the corner accurately.You'll notice that our LCDs grab 4.5s pretty much across the board with the exception of the BenQ. Spotting 6-bit LCDs are fairly easy for people who do a lot of graphics work. The image below displays 256 different shades of blue across the top; it represents the 256 hues of blue that are found in an 8-bit sub pixel. No amount of dithering can render this square correctly on a 6-bit LCD.
As expected, the lack of a digital cable on the Samsung 910V hurt it in more ways than one. For most of our other LCDs, we knew that we were getting the correct signal with the test pattern rendered the way in which the LCD manufacturer had intended it to render. After testing a few monitors on a digital connection, we almost immediately recognized small amounts of interference on the 910V, and we did the best that we could to record our observations on the other tests.
We were not real surprised to see the Samsung 193P pull ahead in a lot of the subjective testing. It's a $700 monitor, and it should perform better in many situations. Samsung seems to have a knack for making good panels. However, if we had to pick a second best performer (other than the Dell 2001FP), we were a little surprised to see that the ViewSonic Q190MB and the NuTech L921G had posted generally sound performance numbers. Color was on balance, screen uniformity was there, and glare was fine.
Facebook
Twitter
RSS
97 Comments
View All Comments
xann - Monday, January 14, 2008 - link
it is like:640*480 1 :1
800*600 1,25 :1,25
1024*768 1,6 :1,6
1152*864 1,8 :1,8
1280*1024 2 :2,13
1600*1200 2,5 :2,5
1440*900 2,25 :1,87
1680*1050 2,625 :2,1875
1920*1200 3 :2,5
2560*1600 4 :3,33333
wide screen resolutions are not one on one.
i look for 19" LCD with res 1600*1200
xann - Monday, January 14, 2008 - link
it is like:640*480 1 :1
800*600 1,25 :1,25
1024*768 1,6 :1,6
1152*864 1,8 :1,8
1280*1024 2 :2,13
1600*1200 2,5 :2,5
1440*900 2,25 :1,87
1680*1050 2,625 :2,1875
1920*1200 3 :2,5
2560*1600 4 :3,33333
Cay - Friday, December 17, 2004 - link
One of the main reasons I think PVA is in a completely different league regarding response time compared to TN/IPS, is this guide by X-Bit Labs.http://www.xbitlabs.com/articles/other/display/lcd...
Cay - Friday, December 17, 2004 - link
The tips at the start were great though, and pointed out several important things I'd have overlooked otherwise.Cay - Friday, December 17, 2004 - link
This review wasn't very useful to me, frankly. It contains too much subjective talk, and barely any data. I miss the following:.) For each panel, clearly state the technology used (TN, M/PVA, IPS), as well as if it's 8 or 6 bit. This gives more meaningful information than all the manufacturer specs.
.) Screen photographs of select, demanding images taken under the same conditions. Here, I want to see how each image looks like on EACH screen, not just 3 samples total.
Printer reviews do this well. You can dedicate one HTML page for each test image (eg 1 for the Max Payne scene, 1 for the Morrowind pic), and put the results from different monitors right below each other. This would allow me to directly compare how still images look like. At the end of the image block, you can put your comments.
With well-chosen test images, this might give a pretty complete idea of how good a monitor's still image/color quality/contrast is.
.) A meaningful measurement of response time.
When you claim to "not notice any difference" between a PVA and a TN panel, I have to conclude that the Anandtech crew is much less sensitive to response time than most people. That's great for you, but bad for me - I'm not getting any response time information.
Build a device to measure all gray-black or grey-grey response times. Tom's Hardware and X-Bit have some you can use for inspiration.
Just including this measurement does not mean that you have to emphasize its importance. You can put it in a section together with your subjective response time interpretation.
610 - Tuesday, December 14, 2004 - link
Is the NuTech L921G currently available in the US?The article says that the review unit was not store-bought, but doesn't specify how it was obtained. I've found only one site anywhere that claims to sell it, and it's not listed in any of the usual price-comparison sites (like PriceScan, for example).
KristopherKubicki - Monday, December 13, 2004 - link
ecove:I believe all of the monitors you mentioned use the same AUO panel. Performance will be very similar.
Hope that helps,
Kristopher
ecove - Thursday, December 9, 2004 - link
I've noticed in researching 19" LCDs that there are a number of models from various manufacturers (eg Princeton LCD19D, CTX S962A/G, Advueu ADV190DT) that all share what appear to be identical specs to the Nu and Viewsonic models reviewed here. Is it a safe assumption that these monitors all share the same panel and have essentially the same performance?stephenbrooks - Tuesday, December 7, 2004 - link
That sample blue image on the penultimate page really made me worry about the LCD I'm using to view it here because it looked terribly dithered! Fortunately I realised the image you showed there was actually a 256-colour GIF (?!) of the image you should have used.R3MF - Tuesday, December 7, 2004 - link
@ #69 -agreed, i would like to see what the Dell 1905FP is like too.........
@ #87 -
yes, the Dell 1905FP is rated at 20ms, and while it may not be best for games like UT2k4, but what about games like Mafia and Rome: TW where twitch gaming is irrelevant?
REMF