I’ve rearranged this post - since most people who visit are probably just coming to see the updates. I’ve moved the table to the top. Effective July ‘04 Home Theater magazine implemented a new process for measuring contrast ratio. Finally there is an independent process that should give the potential buyer a better idea as to the brands true contrast ratio. This table holds the results. To learn more about the process read the section below the table. Category leader(s) shown in red.


Brightness Black Level Contrast ANSI Date
Brand/ Model 100-IRE wht 0-IRE Blk Ratio Contrast MFG Reviewed
Tested (ft-L) (ft-L) (xxxx:1) (xxx:1) CR (mmm,yr)
-------------- ----------- --------- -------- -------- ---- --------
Plasma
Mitsubishi PD6130 12.51 0.129 97 173 1200 AUG 2004
Vidikron VP-50 17.8 0.115 155 268 ? OCT 2004 (http://hometheatermag.com/plasmadisplays/1004vidikron/)
Panasonic TH42PX25 21.95 (40) 0.027 813 780 3000 NOV 2004 (http://hometheatermag.com/plasmadisplays/1104panasonic/)
LG RU-42PX11 27.8 (69) 0.069 403 385 5000 FEB 2005 (http://hometheatermag.com/plasmadisplays/205lg/)
Panasonic TH42PD25 18.64 (31) 0.023 810 998 4000 FEB 2005 (http://hometheatermag.com/plasmadisplays/205panasonic/)
Vinc P42HD 20.22 (57) 0.088 230 464 3000 FEB 2005 (http://hometheatermag.com/plasmadisplays/205vinc/)
LG 42PX4D 25.32 (72) 0.027 938 1327 5000 AUG 2005 (http://hometheatermag.com/lcds/805lg/)
Panasonic TH42PX500U 21.03 (55) 0.027 779 364 3000 DEC 2005 (http://hometheatermag.com/plasmadisplays/1205panasonic/)
Pioneer PDP-5060HD 20.16 (61) 0.033 611 1135 4000 JAN 2006 (http://hometheatermag.com/plasmadisplays/106pioneer/)
Philips 42PF9630A 24.2 (53) 0.015 1613 1108 10000 JAN 2006 (http://hometheatermag.com/plasmadisplays/106philips/)
Hitachi 42HDT52 19.22 0.043 447 360 FEB 2006 (http://hometheatermag.com/completesystems/206polk/index4.html)
Fujitsu P50XTA51UB 18.03 (56) 0.035 515 786 OCT 2006 (http://hometheatermag.com/plasmadisplays/1006fujitsu/)
Panasonic TH42PX60U 27.77 (49) 0.012 2314 740 10000 NOV 2006
Pioneer PRO FHD1 17.75 (44) 0.032 550 930 3000 website (http://hometheatermag.com/plasmadisplays/0207pioneer/index.html)
Pioneer PRO-940HD 18.55 (53) 0.039 476 851 MAR 2007
Hitachi P42H401 21.74 (57) 0.084 259 414 AUG 2007
Panasonic TH42PZ700U 20.2 (58) 0.015 1347 1182 5000 SEP 2007
Panasonic TH50PZ700U 16.47 (44) 0.017 969 1022 5000 SEP 2007
Pioneer PRO-110FD 16.08 (43) 0.004 4020(10645)3239 KURO DEC 2007

Note: LCD testing usually shows best contrast ratio with lowest backlight setting.
If reported the max. brightness or deepest black w/resulting (CR) is shown in brackets.

Brightness Black Level Contrast ANSI Date
Brand/ Model 100-IRE wht 0-IRE Blk Ratio Contrast MFG Reviewed
Tested (ft-L) (ft-L) (xxxx:1) (xxx:1) CR (mmm,yr)
-------------- ----------- --------- -------- -------- ---- --------
LCD (Flat Panel)
Hitachi 32HDL51 19.78 (88) 0.037 535 546 900 APR 2005 (http://hometheatermag.com/lcds/405hitachi/)
Samsung LTP468W 63.05 0.161 392 378 800 APR 2005 (http://hometheatermag.com/lcds/405samsung/)
Viewsonic N3200w 161.00 0.294 548 511 600 APR 2005 (http://hometheatermag.com/lcds/405viewsonic/)
Sharp LC-45GX6U 24.75(124) 0.049(.258) 505 (480) 442 800 JUN 2005 (http://hometheatermag.com/lcds/605sharp/)
LG 42LP1D 142.0 0.147 966 473 1200 AUG 2005 (http://hometheatermag.com/lcds/805lg/)
NuVision NVX32HDU 85.77(118) 0.142(.197) 604 (600) 586 1000 SEP 2005 (http://hometheatermag.com/lcds/905nuvision/)
Philips 32PF9996 99.08 0.185 536 505 800 OCT 2005 (http://hometheatermag.com/lcds/1005philips/)
JVC LT-32X776 70.62(104) 0.122(.186) 579 (558) 493 800 OCT 2005 (http://hometheatermag.com/lcds/1005jvc/)
Proton LX-32B1C2 72.06(161) 0.104(.234) 693 (687) 686 550 OCT 2005
Nikada LA1371 175.4 0.296 593 548 500 OCT 2005
Mitsubishi LT-4260 58.16(151) 0.103(.292) 565 (517) 547 800 OCT 2005
Sony KDL-V40XBR1 55.6 (134) 0.047(.126) 1183(1062) 976 1300 JAN 2006
Maxent MX-26X3 128.5 0.169 760 725 800 JAN 2006
Toshiba 32HLX95 103.0 (41) 0.160(.064) 644 (633) 535 800 MAR 2006
WestinghouseLVM-42w2 57.19(158) 0.064(.192) 894 (825) 712 1000 MAY 2006
Sony KDL-40S2000 79.14(141) 0.066(.119) 1199(1187) 1078 1300 JUN 2006
Hitachi 32HLX61 23.99 (94) 0.036(.171) 666 (548) 517 JUN 2006
Sharp LC37D40U 17.76(122) 0.014(.118) 1269(1034) 768 1200 JUN 2006
Sharp LC57D90U 58.43 (99) 0.045(.079) 1298(1252) 1065 1500 AUG 2006
(28) (.022) 1289
Samsung LN-3251D 41.15(138) 0.032(.120) 1286(1152) 1098 4000 SEP 2006
JVC LT-37X987 45.17(101) 0.062(.143) 729(709) 669 800 NOV 2006
Philips 42PF9831D 165.3 0.089 1857 689 4500 DEC 2006
Vizio GV42L 160.9 (38) 0.208(.052) 774(730) 797 JAN 2007
Toshiba 42LX196 124.5 (34) 0.178(.049) 699(685) 625 JAN 2007
Sony KDL46XBR2 127.5 (23) 0.091(.018) 1401(1303) 1217 JAN 2007
HP SLC3760N 89.7(151) 0.084(.143) 1068(1057) 967 MAR 2007
WinBook 40D1 125.2 (61) 0.082(.042) 1527(1457) 1226 APR 2007
Mitsubishi LT37132 134.9 (31) 0.115(.027) 1173(1150) 1077 MAY 2007
NEC Multeous M40 45.58(125) 0.041(.128) 1112(981) 922 MAY 2007
Olevia 747i 134.5 (73) 0.139(.076) 968(965) 825 MAY 2007
Sharp LC-52D92U 14.09(93) 0.006(.046) 2348(2022) 1968 JUL 2007
Samsung LN-T5265F 19.63(106) 0.017(.105) 1155(1005) 1142 OCT 2007
JVC LT-47X788 113.7 0.01 11370 1753 NOV 2007
Philips 47PFL9732D 14.5(55) 0.023(.121) 630 493 DEC 2007
Sony KDL-46XBR4 120 (22/47) 0.007(.005)17143(4to9k)1687 DEC 2007
Toshiba 52LX177 21.08(120) 0.011(.091) 1916(1320) 980 DEC 2007
Samsung LN-T5271F 63.14(20) 0.012(.010) 5262(1776) 1413 DEC 2007

Brightness Black Level Contrast ANSI Date
Brand/ Model 100-IRE wht 0-IRE Blk Ratio Contrast MFG Reviewed
Tested (ft-L) (ft-L) (xxxx:1) (xxx:1) CR (mmm,yr)
-------------- ----------- --------- -------- -------- ---- --------
DLP Projectors
InFocus 7205 22.37 0.022 1017 355 2200 JUL 2004
Marantz VP-12S3 8.514 0.003 2838 503 3800 JUL 2004
Sharp XV-Z12000 6.85 0.002 3425 424 5500 JUL 2004
Yamaha DPX-1100 9.351 0.002 4676 510 4000** MAR 2005
Samsung SP-H700AE 20.83 0.014 1488 370 2800 MAY 2005
InFocus 7210 25.62 0.018 1423 419 2800 JUL 2005
Sharp XV-Z12000 MII 7.263(22) 0.002(.016) 3632(1381) 525 MAR 2006
Yamaha DPX-1300 6.8 (18) 0.002(.009) 3400(2000) 527 JUL 2006
Sharp XV-Z20000 9.117(29) 0.001(.015) 9117(1933) 606 MAR 2007

Midrange Projectors
Panasonic PTAE700LCD 13.5 0.019 711 161 2000 MAR 2005
Sony VPL HS51 LCD 7.701 0.003 2567 162 6000 MAR 2005
Hitachi PJ-TX100 LCD 4.905(20) 0.006(.029) 818(696) 118 OCT 2005
Viewsonic Cine5000 21.7 (16) 0.017(.013) 1276(1265) 231 MAR 2007


Budget projectors
BenQ PB6200 DLP 14.48 0.021 690 283 2000 SEP 2004
Epson PL Home 10+ 11.6 0.015 773 249 800 SEP 2004
Optoma H30 DLP 9.426 0.011 857 240 2000 SEP 2004
Mitsubishi HC900 DLP 8.401 0.005 1680 417 4000***MAR 2005
Optoma H27 DLP 8.785(11) 0.005(.007) 1757(1594) 455 OCT 2005
Epson PL Cinema 550 11.78 (28) 0.003(.06) 3927(481) 168 MAR 2006
Optoma HD72 S+DLP 9.855 (7) 0.012(.009) 821(770) 395 MAY 2006

1080p LCOS/SXRD/LCD projectors
Sony VPL-VW100 15.18 0.001 15180 251 RUBY APR 2006
JVC DLA-HD1 15.56 0.001 15560 250 JUN 2007
Mitsubishi HC5000LCD 11.2(15.5) 0.002(.003) 5600(5160) 206 JUN 2007
Sony VPL-VW50 SXRD 17.05(17.8)0.002(.009) 8525(1968) 231 PEARL JUN 2007
Epson Powerlite1080p 6.852(9) 0.001(.002) 6852(4497) 211 JUL 2007

Sony VPL-VW60 BLK PEARL

Brightness Black Level Contrast ANSI Date
Brand/ Model 100-IRE wht 0-IRE Blk Ratio Contrast MFG Reviewed
Tested (ft-L) (ft-L) (xxxx:1) (xxx:1) CR (mmm,yr)
-------------- ----------- --------- -------- -------- ---- --------
LCOS/SXRD Rear Projection
JVC HD61Z575 D-ILA 165.8 0.215 771 136 ? DEC 2004
JVC HD70G886 D-ILA 130.1 0.14 929 110 ? SEP 2005
Sony KDS-R60XBR1 93.31 0.007 13330 297 10000 NOV 2005 (http://hometheatermag.com/rearprojectiontvs/1105sony/)
Sony KDS-R50XBR1 142.3 0.021 6776 294 FEB 2006 (http://hometheatermag.com/rearprojectiontvs/206faceoff/index1.html)
JVC HD56FH96 D-ILA 91.04 0.058 1570 208 FEB 2006 (http://hometheatermag.com/rearprojectiontvs/206faceoff/index1.html)
Sony KDS-60A2000 63.19(83) 0.008 7899 270 OCT 2006 (http://hometheatermag.com/rearprojectiontvs/1006sonykds60a2000/)
Sony KDSR60XBR2 130.9 0.009 14544 364 FEB 2007

Rear Projection
Sharp DLP 56DR650 83.69 0.080 1046 286 OCT 2005
SamsungDLP HLR5668W 137.2 0.018 7622 374 FEB 2006 (http://hometheatermag.com/rearprojectiontvs/206faceoff/index1.html)
HP DLP MH6580n 104.4 0.015 6960 410 FEB 2006 (http://hometheatermag.com/rearprojectiontvs/206faceoff/index1.html)
Mitsubishi DLP52628 133.6 0.030 4453 416 FEB 2006 (http://hometheatermag.com/rearprojectiontvs/206faceoff/index1.html)
Toshiba 62MX195 100.1 0.025 4004 307 FEB 2006
Samsung HLS5679W 107 0.090 1189 226 JAN 2007
JVC HD-61FN97 98.09 0.016 6131 182 FEB 2007
Mitsubishi WD65731 98.72 0.127 777 385 FEB 2007
Olevia 565H 135.8 0.035 3880 338 FEB 2007
Samsung HLS6188W 169.7 0.033 5142 561 FEB 2007
Toshiba 62MX196 65.09 0.022 2959 401 FEB 2007
NuVision 52LED DLP 125 0.085 1471 423 APR 2007

Cathode Ray Tube (CRT)
Toshiba 34HF84 15.32 <0.001* >15320 135 JAN 2005 ( http://hometheatermag.com/crtdisplays/105toshiba/)

* actually may have been lower - but that was the bottom of their test
equipment scale.
** white boost fully on, iris fully on, 2000 white boost on, iris Off
*** 4000:1 (with Cine Focus close) 2000:1 (with Cine Focus open)

The Contrast Ratio in the above table is a calculated value.
It is determined as follows: 100-IRE white / 0-IRE Black.
How they determined the ANSI Contrast value is described below.

Here is an excerpt (partially edited) from the original article that describes what they are doing.

"A New HT Measurement: Contrast Ratio
Believe it or not, contrast ratio is not the only thing you need to know about purchasing a TV. Like any measurement it shouldn't be taken alone; in many cases it shouldn't be taken at all. The human eye can, at best, discern a contrast ratio of a few hundred to one. So, even a display with a claimed 800:1 contrast ratio has more than enough contrast to appease the eye. There's more to it than that, but that's a discussion for another article.
So what's with a contrast ratio of 2000:1 or even 3000:1? Well it's misleading at best and a lie at worst. Unlike most product stats, there's no regulation for how a company measures a display's contrast ratio. One company could measure a full-white field versus a 0-IRE signal, while another company could use a 100-IRE window versus no signal at all. The American National Standards Institute (ANSI) has defined one relatively reasonable method that many manufacturers use, but it is by no means mandatory.
From this point on in our display reviews we're going to cut through all that and offer you a series of measurements consisting of contrast ratio, lowest and highest light level, and ANSI contrast all in one section. Don't worry, we'll still include the color temperature and color gamut graphs we've given you all along.

Decoding the New Measurements
For our reviews, we will define contrast ratio as the ratio of light level produced by a 100-IRE full field divided by the light level produced by a 0-IRE full field. The number inside the black box is the total light output (in foot-lamberts) that the display produces when supplied with a 0-IRE signal from a DVD. This represents the darkest the display will get when it's active and fed a signal, it should be as close to zero as possible. With projectors we use a Da-Lite Da-Mat 1.0 gain screen that measures 87 by 49 inches (93 inches diagonal). If a display has good DC restoration, the measured black level will be similar to the black level the display produces when there's other video material on the screen.
The number in the white box is the display's total light output when displaying a full-white (100-IRE) field in the same mode that produces the listed black level. (The mode chosen will be the one with the best contrast ratio), Now I'll be the first to tell you that a full-white field almost never happens in actual source material (for that matter neither does a full-black field) for this though it's what we're going to use. Why? Because we have to use a full-black field (a black window just wouldn't work), and this is its true opposite.
Doing the measurement this way is like testing an amp's maximum power output: it may not reflect the most real world scenario, but it does five an overall view of the device's output.
Using a full-white field will affect plasmas the most, as they are built to have less light output on a full field compared with a 100-IRE window. If there's a difference between the windowed measurement and the full screen, we'll list it in the measurement text.
ANSI measures the contrast ratio using a 16-box checkerboard pattern on the screen, so we'll do that, too. This is done by dividing the average level of the eight white boxes by the average level of the eight black boxes.
If a display has different settings that affect contrast ratio, we will test all of them and use the best one for both contrast-ratio measurement tests. Most importantly all contrast-ratio measurements will be done after the display is fully calibrated to as close to D6500 as it can be.
To conduct these tests, we're using a Konica/Minolta LS-100 light meter, the most accurate tool for the job. We're still be using our Photo Research PR-650 spectoradiometer for color temperature and color measurement."

Note: The mode discussed above, at least in relation these early results on projectors, was described to be high power lamp mode.
What's even more interesting in examining the differences in the numbers in the above table (for DLP HD2 -only) - is that all these projectors use the identical HD2+ chip!

Their final comments:
(this commentary was made at the process inception when only the three DLP HD+ models had been tested)
"So which one would I pick? Well that's easy. I'll take the Marantz's optics, case, and color points; the Sharp's black level and adjustable iris, and the InFocus overall light output and price."

See it's easy to purchase your TV based on the specs! :D

And for the metric folks:
Luminance - refers to the amount of visually effective light emitted by an extended source. Typically expressed in nits, footlamberts (fL) or candelas per square meter (cd/m2).
One fL = 3.43 cd/m2 or 3.43 nits
One cd/m2 (1 nit) = 0.292fL

Other pertinent data
Here’s some tests performed by primediabusiness associated with Peter Putman a well respected independent consultant. You can learn more by going here (http://videosystems.primediabusiness.com/ar/video_six_degrees_illumination/)
Note the Brightness measurements are in nits as opposed to those shown in the above table.


Black Brightness Brightness 4:3 ANSI 4:3 Peak 16:9 ANSI 16:9 Peak
Brand/ Model Level 4:3 white 16:9 white Contrast Contrast Contrast Contrast
Tested (nits) (nits) (nits) (xxx:1) (xxx:1) (xxx:1) (xxx:1)
-------------- ------ ----------- --------- -------- -------- -------- --------
LCD’s
Barco Solaris LC40M 1.25 244.65 190 204
LG L4200A 2.6 219.9 61 73
NEC LCD 4000 * 319.88 330.85 129 152 136 169

Plasma’s
Hitachi CMP4201U 0.39 83.28 67.21 257 327 284 413
NEC PS42VP4 0.34 72.92 59.98 386 457 332 389
NEC PX-61MX2A 0.56 52.56 200 358
Panasonic TH42PWD6 0.2 69.13 55.13 488 604 508 649
Philips 32FD9954 71.85 54.34 159 188 134 157
Pioneer PDP-504MX 0.54 57.5 146 208
Sony PFM-42X1 0.18 95.13 648 2007
Vizio P4 - 78.32 271 291 198 224

* mentioned to be much higher than NEC plasma, but numbers not provided in article.

And here’s some more data courtesy of PC Magazine this article (http://www.pcmag.com/article2/0,1759,1728897,00.asp) attachment.

Avg. Avg
Black White Average
Brand/ Level Level Contrast
Model Tested (nits) (nits) (xxx:1)
------------------- ------ -------- -------
LCD’s
Samsung LT-P468W 0.57 237 422
Sharp Aquos LC-37G4U 0.74 322 437
Syntax Olevia LT30HV 0.77 415 541

Plasma’s
Fujitsu P55XHA30WS 0.39 109 278
Panasonic TH-50PX25U/P 0.15 87 583
Pioneer Elite Pro 1110HD 0.45 122 269

Rear Projection
JVC HD-61Z575 HD-ILA 2.20 225 104
Samsung HL-P5063W 1.03 174 174
Sony GW KDF-60XBR950 1.24 165 137

Originally posted by semigolfer
The human eye can, at best, discern a contrast raio of a few hundred to one. So, even a display with a claimed 800:1 contrast ratio has more than enough contrast to appease the eye. There's more to it than that, but that's a discussion for another article.
So what's with a contrast ratio of 2000:1 or even 3000:1? Well it's misleading at best and a lie at worst. I read this too. Either this person understands things fairly well and just happened to write something very misleading but would save the real part for "a discussion for another article" or they took a little bit of information and didn't know how to apply it. The human eye can discern a contrast ratio of some small amount at one time. It can discern huge contrast ratios (I believe a million to one or more) over multiple images and this is mostly what on/off contrast ratio is about. Movies have both bright scenes and dark scenes, not just mixed scenes. The amount that a person can see at any one time with about equal amounts of black and white is more relevant to the ANSI CR test. Both tests are important as the ANSI CR test pretty much defines the washout effect and the on/off test defines the floor.

And people keep quoting numbers about what a person can see at one time with equal mixes as if this is all the CR we need. In real world video people are allowed to look wherever they want and I've done some testing with an image that was 3-5% white, then projector black, then room reflection black on my screen and then screen masking black (I've just described what dominates each area as everyone of them is a mixture of multiple sources of light). I can make each of these out even though the white to projector black was about 1600:1 (with a large margin of error), the white ellipse to room reflection black was probably 5k:1 or more and then it went way up to the screen masking and velvet around the screen. This was on a screen about 8' wide from about 13' back in a light controlled room. I would definitely have been able to tell if that 1600:1 area had been improved to something like 20k:1.

Plasmas putting out less ft-lamberts for full field white than windowed white is very similar to CRTs.

--Darin

Originally posted by darinp2
The human eye can discern a contrast ratio of some small amount at one time. It can discern huge contrast ratios (I believe a million to one or more) over multiple images and this is mostly what on/off contrast ratio is about.
--Darin
Here are the scientifically proven numbers
------------------------------------------------------------
The dynamic range of the human eye is .0001 to 10,000 or in other words 100,000,000 to 1 (7 decades).

The contrast ability of the human eye in daylight conditions is 1% or in other words 100:1

The contrast ability of the human eye in dark/dim conditions is 10% or in other word 10:1
-------------------------------------------------------------

In other words, the eye has a very large dynamic range but for a given luminence the contrast sensitivity (ability to discern luminence values) is very small!!!! and it changes with luminence. The lower the luminence the lower the sensitivity.


Cheers

Originally posted by xrox
Here are the scientifically proven numbers
...
The contrast ability of the human eye in daylight conditions is 1% or in other words 100:1.

In other words, the eye has a very large dynamic range but for a given luminence the contrast sensitivity (ability to discern luminence values) is very small!!!! and it changes with luminence. The lower the luminence the lower the sensitivity. I think it depends on what you use as "discern luminance values". A person doesn't have to be able to tell you what an exact luminance value is in order to see an improvement. For instance, if you put up a small white ellipse (like I did in my example) on a dark background of either 1000:1 instantaneous CR or 20k:1 instantaneous CR I guarantee you I could tell you which was which after leaving the room and coming back in. In the first case I would not mistake the background for true black and in the second case it would be much closer or I would mistake it for true black. That doesn't mean I could tell you what the background luminance is with a small margin of error, but I can sure tell you that it isn't black in one case. Going from 100:1 to 20k:1 instantaneous in images like the one I described would be a night and day difference in a dark room. So, using these 1% and 10% numbers to justify lower design thresholds in video display equipment is just incorrect, IMO.

In other words, a video display with a design that can go way beyond these 1% and 10% thresholds in images like the one I described will show differences to the human eye.

As I mentioned, in video people are not forced to look at one spot. They can choose where to look on the screen and this definitely applies to front projection where many sit at close to 1.5x in dark rooms. For bright rooms like where many people watch plasmas it is a different story.

--Darin

Originally posted by xrox
In other words, the eye has a very large dynamic range but for a given luminence the contrast sensitivity (ability to discern luminence values) is very small!!!! and it changes with luminence. The lower the luminence the lower the sensitivity.



It's also dependent on spatial frequency of the pattern. I posted a chart and some references here (http://www.avsforum.com/avs-vb/showthread.php?postid=3880686#post3880686). Darin was going to read the big thick paper and tell us all about it :)

Boy - you guys are over the post like cats on spilt milk. I agree with both of your responses. They add considerable information to this discussion. As the nuances of contrast ratio make it NOT the most straightforward spec to understand. It was even interesting to note on their proposed procedure that Home Theater Magazines measured and then calculated contrast ratio (100-IRE)/(0-IRE) rank ordered these projectors differently than the ANSI Contrast Ratio.

Although I'm not an expert in this matter I've done a considerable amount of reading on this subject trying to understand the value of this spec, when it comes to buying a TV. A couple of interesting things I've read include:
"The goal of any televion technology is to deliver a picture that looks as film-like as possible,"" Film has a contrast ratio closer to 10,000-to-1, but 4,000-to-1 is close enough to fool the human eye."

All of this has led me to believe that when I do get around to buying a TV, it will be my eyes that make my decision. But my eyes will view the competing brands, in different ambient lighting conditions, while viewing movies shot in different lighting conditions, - as many of the experts here in the AVS forum have long proposed.

What I was most interested in and the reason I posted was that - finally a trade publication was going to finally clarify how they were going to measure contrast ratio in the future, and why they were making these choices. Thus finally an "independent" source, who does a lot of reviews, would provide what appears to be most objective results than I had seen to date regarding CR.;)
And I thought the initial numbers were quite interesting considering the manufacturers started their designs with the same chip.

Originally posted by wjchan
It's also dependent on spatial frequency of the pattern. I posted a chart and some references here (http://www.avsforum.com/avs-vb/showthread.php?postid=3880686#post3880686). Darin was going to read the big thick paper and tell us all about it :)
Just what you posted is good stuff. The part that I think relates most to what I am saying is:

One complicating factor is that, dependent upon light scattering into the fovea, the eye can have partial “local adaptation” to features in an image to increase sensitivity.

The Contrast Sensitivity Function (CSF) of the HVS is a measure of the ability to discriminate small luminance changes in a sine wave pattern target on a uniform background.
This "complicating factor" definitely relates to watching a large image in a dark room. And this CSF is a measure of the ability to discriminate small luminance changes. Anybody who designs a display device who uses this as an excuse to have performance that will be noticably substandard when the luminance changes aren't small is misapplying the results.

From the things I've read contrast ratio is one of the most misunderstood things. I relate the general opinion within much of the projection community (including professional writers) that on/off contrast ratio doesn't matter to the old flat earth theory. They are just plain wrong and they've used information about instantaneous contrast ratio without understanding things. Greg Rogers from WideScreenReview is the only writer I can think of who I know really gets this stuff. I don't claim to understand it all, but there is a big jump from some of these guys who disclaim any kind of on/off contrast ratio as relevant and having a clue.

At least in this case they still published some numbers. I still find it funny how they take a black level reading of 0.002 with a device that is only accurate to either plus or minus 1 or 2 in the last digit and then claim 4 digits of precision in their CR number, though.

--Darin

darinp2,

on/off (dark room) contrast which most plasma displays quote (ie 3000:1 or whatever) is a direct measure of dynamic range and should be taken as such.

So for the average consumer an on/off contrast ratio of thousands:1 tells you two important things,

1) - the display has a fairly large dynamic range (ie large number of distinguishable/measurable luminance values are possible)
2) - the display has a fairly low black level (unless it is ungodly bright:))

would you agree?

cheers

Originally posted by xrox
on/off (dark room) contrast which most plasma displays quote (ie 3000:1 or whatever) is a direct measure of dynamic range and should be taken as such.

So for the average consumer an on/off contrast ratio of thousands:1 tells you two important things,

1) - the display has a fairly large dynamic range (ie large number of distinguishable/measurable luminance values are possible)
2) - the display has a fairly low black level (unless it is ungodly bright:))

would you agree?
Yes, I would say that's pretty much true as long as the number you get is a real number and not just a marketing number. The on/off CR does only really tell you the endpoints and it is theoretically possible for a display to have a large on/off CR and only two levels total, but the on/off CR gives you the ratio of those endpoints and can be used with white levels (like ANSI lumens) to determine both points. In practice many of the displays we compare to each other have the same number of distinguishable luminance values (might be limited by 8 bits), but the higher on/off CR projector just has a bigger ratio for the endpoints.

With plasmas they could probably just give you the white level and the black level instead of the CR (although I don't know all the controls in plasmas and how they interact). The difference with front projectors where I spend most of my energy is that the user has a lot of choices about how to use that CR with different screens sizes, screen materials, filters, etc. That is, if the front projector owner doesn't mind getting some ridiculously low white level (like 2 ft-lamberts) they can get a much lower black level with the same CR. I'm guessing that isn't true with plasmas.

--Darin

Well my Aug 2004 issue of Home Theater Magazine arrived in the mail today - the first plasma CR data is now available. It's for a Mitsubishi PD-6130 61" plasma. It's been added to the table in the original post at the top of this thread.

Originally posted by semigolfer
Well my Aug 2004 issue of Home Theater Magazine arrived in the mail today - the first plasma CR data is now available. It's for a Mitsubishi PD-6130 61" plasma. It's been added to the table in the original post at the top of this thread.
Wow, that is amazing. An ANSI CR higher than the on/off CR. This really wouldn't happen with projectors and I find it amazing that it would happen with a plasma, even with the ability to have higher ft-lamberts for white parts with less than full-screen white. It looks to me like the washout effect from white squares in the ANSI test to the black squares is almost negligable compared to the black level in the on/off test. Then the reason for the higher ANSI would be that the white squares are much higher than 12.51 ft-lamberts during the ANSI CR test. That is the only explanation I can think of unless there is something really funky in their control and they can go darker in the ANSI test than in the on/off test. That 97:1 actual vs 1200:1 spec sure is amazing too.

--Darin

Sept 2004 HTMag has hit the mailbox.

Unfortunately no flat screens tested - just Budget Projectors - the post at the top of the thread has been updated. Maybe next month!

"Wow, that is amazing. An ANSI CR higher than the on/off CR."

So amazing, it tells you the test is inherently flawed.

" Using a full-white field will affect plasmas the most, as they are built ot have less light output on a full field compared with a 100-IRE window. If there's a difference between the windowed measurement and the full screen, we'll list it in the measurement text."

Right, so the Mitsubishi was "dialing back" the white field in all likelihood. Now, what were the black level and white level for the "window test". That would tell us something a bit more useful. We rarely ask our displays to light up full white -- nearly full blue or green are much more likely.

mark

Oct 2004 Home Theater Magazine arrives today.

They tested another plasma using their new contrast ratio measurement procedure, unfortunately this plasma also not very mainstream.

Results at the top of the thread.

Nov 2004 Home Theater magazine just arrived.

Model tested, using their new contrast testing procedures: Panasonic TH-42PX25

Results added at top of thread - just like everyone has been saying Panny just knows how to make black better than the rest of the pack - but it's nice to see more independent confirmation.

Looking to purchase a flat screen - the same issue lists 400 models (LCD & plasma).

Are those numbers for real? If so, I want a Sharp XV-Z12000. How come I don't hear more about it in the $3500 FP forum? Does it have other problems?

The ANSI on the Panasonic is nothing short of stunning. The Marantz S3 and Sharp 12K are both terrific ANSI performers and to see a plasma so resoundingly beat them says a lot.

Mark

Dec 2004 Home Theater magazine arrived yesterday.

And they added an LCOS projection set to the list of models tested (using their new contrast ratio procedures) - the JVC HD-61Z575.

Hopefully a direct view LCD is coming up soon for testing, but I wonder if the results will tilt toward those found for this LCOS unit?
(i.e. high brightness (but probably not this high) - so-so blacks (but probably better than LCOS), a relatively high computed contrast ratio, but a mediocre ANSI result)

Originally posted by rogo
"Wow, that is amazing. An ANSI CR higher than the on/off CR."
So amazing, it tells you the test is inherently flawed.

" Using a full-white field will affect plasmas the most, as they are built ot have less light output on a full field compared with a 100-IRE window. If there's a difference between the windowed measurement and the full screen, we'll list it in the measurement text."

Right, so the Mitsubishi was "dialing back" the white field in all likelihood. Now, what were the black level and white level for the "window test". That would tell us something a bit more useful. We rarely ask our displays to light up full white -- nearly full blue or green are much more likely.

mark

Sorry for the late response here (only three months late :D) - but recently I reread the article, and the author did state that:
"On a 100-IRE window, the PD-6130 was capable of 27.55 ft-L (it's normal for a plasma to have more light output on a window versus a full field)" and in the body of the article (as opposed to the HT Lab Measures section where I extract most of the above data) he also noted that the screen brightness was quite high, measuring 51.4 ft-L "at full contrast" on a 100-IRE window." And again in the body of the article - "The PD-6130 black level measured a mere 0.06 ft-L on a 0-IRE window, low enough to provide an excellent viewing experience with the room lights turned down or off."

Another nice bonanza - three budget plasmas (VInc, LG, and Panasonic) were entered into the chart this month as they were reviewed in the Feb 2005 issue, which arrived in my mail today..

I also forgot to mention, I contacted the magazine about two months ago, asking that they consider testing some flat panel LCD models. They were kind enough to respond that they have four models under test, and they are scheduled for the April 05 issue. So we should get something in early March.

Interesting, very interesting.

The April '05 issue of HT Mag arrived with today's mail and they tested three LCD panels - the first ones they've done with their next contrast testing protocals.
The results are in the table at the top of the thread.

I should probably also include these notes from the review of the Hitachi - since the results were somewhat unusual :
"The best contrast ratio was achieved in the night setting and the back light set for 0%. The brightest image was achieved in the day setting with the backlight set at 100%, which produced 87.5 ft-L with a 100 IRE window."

The June 2005 issue of Home Theater Magazine has a review of Sharp's AQUOS LC-45GX6U.

The contrast results have been entered into the table at the top of this thread.
Also should note that 33 step (i.e. adjustable) backlight which really helped it's black level performance.

Bump for some great info and a question.

Is there any way to relate the light output on a 0 IRE window to a shade of grey?

ss

Originally posted by xrox
Here are the scientifically proven numbers
------------------------------------------------------------
The dynamic range of the human eye is .0001 to 10,000 or in other words 100,000,000 to 1 (7 decades).

The contrast ability of the human eye in daylight conditions is 1% or in other words 100:1

The contrast ability of the human eye in dark/dim conditions is 10% or in other word 10:1
-------------------------------------------------------------

In other words, the eye has a very large dynamic range but for a given luminence the contrast sensitivity (ability to discern luminence values) is very small!!!! and it changes with luminence. The lower the luminence the lower the sensitivity.


Cheers

Perceived contrast can be affected by how the brain processes relative contrast. See attached optical illusion example.

The A and B squares (not the letters themselves) are the exact same color (shade of gray).

Originally posted by subysouth
Bump for some great info and a question.

Is there any way to relate the light output on a 0 IRE window to a shade of grey?

ss

If the setting is for black at 0 IRE (some systems are set to do black at 7.5 IRE), then 0 IRE = black = 0 on the greyscale spectrum.

Originally posted by rogo
If the setting is for black at 0 IRE (some systems are set to do black at 7.5 IRE), then 0 IRE = black = 0 on the greyscale spectrum.

Yea I guess what I am asking doesnt make sense maybe.

I am guessing the signal is telling the display to show no light correct? But any light is in theory gonna move the black of the darkened room to some shade of grey I would think.

Are these two seperate things?

ss

Originally posted by Rgb
Perceived contrast can be affected by how the brain processes relative contrast. See attached optical illusion example.

The A and B squares (not the letters themselves) are the exact same color (shade of gray).

I think that total contrast range contributes quite a bit to the wow factor we apply to a device, even if it cant be pulled off simultaneously on the screen.

I am still gonna err on the side of trying for the deepest and give some credence to full measured CR(not this dynamic CR fraprap) as well as ANSI CR.

That Panny ED looks strong on paper as it does in real life apparently.

If I could just get my mind around what color grey those measured 0 IRE figures are.

ss

Not that is a very big deal, but if I am reading the chart correctly:

Budget projectors
BenQ PB6200 DLP 14.48 0.021 690 283 2000 SEP 2004
Epson PL Home 10+ 11.6 0.015 773 249 1200 SEP 2004
Optoma H30 DLP 9.426 0.011 857 240 2000 SEP 2004
Mitsubishi HC900 DLP 8.401 0.005 1680 417 2000 MAR 2005


It is saying that the published manufactures CR for the Epson Home 10+ is 1200..............this is incorrect. It has 1200 lumens and 800:1 CR.

Am I reading it wrong, if so, I apologize
I also have a very hard time believing that the Optima and BenQ DLP's have lower CR's than the Epson I own..........

Sorry if I am misinterpreting.......

Originally posted by subysouth

I am guessing the signal is telling the display to show no light correct? But any light is in theory gonna move the black of the darkened room to some shade of grey I would think.

Are these two seperate things?


Yes, they are. The 0 IRE signal is still "black". But as you know, on exactly zero flat panel displays is black actually, well, black. But that illuminated state is still the "black state" at 0 IRE.

Mark

Originally posted by rogo
Yes, they are. The 0 IRE signal is still "black". But as you know, on exactly zero flat panel displays is black actually, well, black. But that illuminated state is still the "black state" at 0 IRE.

Mark

Can you think of a way they could translate the light output at 0 IRE to actual IRE. Something like, 0 IRE on the Panny 42 = 8.5 IRE actual.

ss

A question for the experts here. Does presence (or lack) of color affect the perception of black? I notice that when I tweak up the color level slightly (from a low, sort of pastel, color level) on my 42PA25U it seems to simultaneously improve contrast, but I am not sure if the display is doing that or if it is just subjective. Would CR perception be any different between a black and white image vs the same image with color added?

Originally posted by dsmith901
A question for the experts here. Does presence (or lack) of color affect the perception of black? I notice that when I tweak up the color level slightly (from a low, sort of pastel, color level) on my 42PA25U it seems to simultaneously improve contrast, but I am not sure if the display is doing that or if it is just subjective. Would CR perception be any different between a black and white image vs the same image with color added?

I'm no expert but I absolutely think that color makes a big difference. Could be becuase I'm color blind though :)

P.S. - and no that doesn't mean I see in black and white (yes many people actually think that........LOL)

Cheers

Originally posted by subysouth
Can you think of a way they could translate the light output at 0 IRE to actual IRE. Something like, 0 IRE on the Panny 42 = 8.5 IRE actual.


It doesn't really work that way.

Here's the Video Essentials definition of IRE:

IRE (Institute of Radio Engineers) 1 Volt Peak - Peak Video is divided up into 140 IRE units. This is done to make numbers for luminance levels easier to communicate. The amplitude of the video signal from blanking (zero volts) to peak white is 0.714286 volts or 100 IRE units. Synchronization signals extend from blanking to -0.285714 volts or -40 IRE units.

* Picture Black Level: 7.5 IRE (above zero volts)
* Picture White Level: 100.0 IRE (above zero volts)
* Blanking Level : 0.0 IRE (zero volt level)
* Burst Pedestal : 0.0 IRE (zero volt level)
* Synchronization : -40.0 IRE (below zero volts)

Black was raised above 0 IRE in the early days of color television to get around some transmitter problems. The practice has been with us ever since.

The SMPTE component video system does use the IRE system to define video levels. It is based on a 700 mVolt system rather than the 714 mVolt composite video.

What your asking, I think, is where is the Panasonic's black on the IRE scale compared to some reference display that showed a zero luminance -- 0.0 nit -- black. And I suppose it's a valid question, but not a very interesting one.

If you want to know how good the black is, ask how many nits the display is putting out when "showing" black. The Panasonic at 0.2 is pretty outstanding. Many plasmas and LCDs are at 0.6-0.8, which is, well, not.

But the Panasonic still has a greyscale value above it's black for each of the grey steps. And it will do something different for each IRE from 0 to 100. You might not see all those differences because it can't show that many simultaneous shades, but it's not as if the tail just gets clipped off the bottom.

Mark

Originally posted by rogo

What your asking, I think, is where is the Panasonic's black on the IRE scale compared to some reference display that showed a zero luminance -- 0.0 nit -- black.

If you want to know how good the black is, ask how many nits the display is putting out when "showing" black. The Panasonic at 0.2 is pretty outstanding. Many plasmas and LCDs are at 0.6-0.8, which is, well, not.

But the Panasonic still has a greyscale value above it's black for each of the grey steps. And it will do something different for each IRE from 0 to 100. You might not see all those differences because it can't show that many simultaneous shades, but it's not as if the tail just gets clipped off the bottom.

Mark

Got it, so its a relative system.

Thats more what I was asking is what light is being put out when its producing its deepest black so I guess your suggestion of a nits reading is what I'm looking for. Deep blacks are important to me. I prefer to lose contrast range off the top to have solid blacks I can see. I actaully had my eyes checked during a physical exam for a military academy and they said I had low-light sensitivity. On the flip side, I dont like brightly lit scenes, envirions at all.

So my idea was if I(or anyone) could define a number that was black level that was sufficiently grey to be annoying, you could use that to exclude displays.

You seem to have some knowledge of this yourself as you are saying .6-.8 is not so good.

And I suppose it's a valid question, but not a very interesting one.

And try to keep the rather odd editorializing to a minimum. The question was clearly interesting.;) K? Tanx

ss

Originally posted by lifesun
Not that is a very big deal, but if I am reading the chart correctly:

It is saying that the published manufactures CR for the Epson Home 10+ is 1200..............this is incorrect. It has 1200 lumens and 800:1 CR.


Thanks lifesun;
It's been corrected - Home Theater Magazine typically doesn't include this info in their review, so I quickly look it up - just out of curiosity. First I try the manufacturer's website - if it's not there I generally use the first google return. Either way the 1200 I put in there was in error - thanks for taking the time to point it out.

Originally posted by semigolfer
Thanks lifesun;
It's been corrected - Home Theater Magazine typically doesn't include this info in their review, so I quickly look it up - just out of curiosity. First I try the manufacturer's website - if it's not there I generally use the first google return. Either way the 1200 I put in there was in error - thanks for taking the time to point it out.

Great work btw and its telling how the manufacturers CR mostly has little to do with reality.

ss

suby, I'm not saying your concern isn't interesting, dude. I'm saying the particular question you asked wasn't the interesting one. :)

If I didn't give a damn about the topic, would I have taken so much time to help you out there? :D

Originally posted by rogo
suby, I'm not saying your concern isn't interesting, dude. I'm saying the particular question you asked wasn't the interesting one. :)

If I didn't give a damn about the topic, would I have taken so much time to help you out there? :D

I know my brother. You so seldom say anything OT, I had to jump on it. My response was meant all in good fun. :) And I do appreciate, as always, all the info.

ss

Peace, my good man. :)

This month (August 2005) HT Mag tests the 42" LCD and plasma units.
Impressive CR results particularly the plasma - see table at the top of the thread (i.e. the very first post.

This month (August 2005) HT Mag tests the 42" LCD and plasma units.
Impressive CR results particularly the plasma - see table above.

Table?

ss

Table?


Thanks for pointing out your confusion - changed the wording - first post of the thread.

Thanks for pointing out your confusion - changed the wording - first post of the thread.

Doh - ADD.

The numbers on the LG model are quite good - phenomenal even.

ss

What year is the LG plasma from? Its not listed on the LG website.

ss

Both the LCD and plasma are stunning numbers. I want to read the test reports.

The LG and ViewSonic LCDs have "flame thrower" white output -- time for some sunglasses! The black level of the Hitachi impresses me -- and it looks quite good in the store.

Paul

Bump to keep this stuff on the forefront. Any new updates for August?

Sometimes I just update this page, and don't post a reply, which would essentially bump the post.

This month (Sept 2005) (magazine arrived today in the mail) however HTMag reviewed a flat screen I've never seen or heard of. It's a NuVision NVX32HDU. MSRP $2499.
It kind of pushes the concept of a flat panel as it's dimensions are: 22.15x36.9x8.66".
They gave it pretty high scores on their review, even though it lost a significant number of points due to lack of connections, and connectivity.

Learn more at: http://www.nuvision.com/

Follow-up note: the dimensions above were from the magazine. I download the specs for their 37" model and it appears to be only 4.75" deep w/o stand, and 9.8" w/stand so I'm assuming the magazine reported the dimensions w/stand - thus it is not pushing out the concept of a flat panel.

My Oct 2005 issue of Home Theater mag arrived in today's mail. Wow, it's not sept yet.
Anyways Christmas must be coming, and they've been busy little bees, as they reviewed 7 TV's/monitors/etc. this month. Now not all where flat panels, but most were.

Their highest rated (not talking contrast ratio here) model of the bunch was the:
Mitsubishi LT-4260, some of the highest reviewed numbers I've seen for an LCD from them.

Nov 2005 HTMag issue arrived today. Normally I don't post the new data unless it involves a flat panel - BUT.....

since this technology has been mentioned here as the next big leap in picture quality I thought I would mention that they tested the Sony 60" - KDS-R60XBR1 LCOS (i.e. SXRD) with a MSRP of $4999.

On-off contrast ratio of 13330:1 (and no I didn't forget the decimal point) - as always data posted to the first post of this thread. The size BTW is 39.75 x 66 x 20.25 - weight 112 lbs.

ps. the issue also has pictures and short descriptions of 25 plasma perfect speakers (by Nancy Klosek), and some reviews flat screen type HTIB 5.1 systems with wireless surrounds (by Adrienne Maxwell)

Nov 2005 HTMag issue arrived today. Normally I don't post the new data unless it involves a flat panel - BUT.....

On-off contrast ratio of 13330:1

First semigolfer, thanks for keeping us updated and I am stunned with how quick you get the mag, I'm about a week behind.

I dont mean to seem to be looking at the glass half empty but I am suspicious of this tech. The reviewer notes that if you look for it you can see the active iris system at work and its this active iris that nets that phenomenal standard CR. Meaning it takes some visual ping-pong to get good blacks on the set. The set has a mediocre grey scale track and he also noticed some uniformity issues. Colors are oversaturated too, but he said a true production model might have that sorted.

If you are Sony you should be blushing after the comments by this same reviewer though so I guess I am just curious where we are setting that performance bar against the backdrop of really just higher res and bigger screens. 7 years ago a CRT with that performance would have been panned. And its got a bulb. Just noting the irony of where we have been and where we're going.

He also suggests the unit does 1080i 3:2 pulldown. If it does that would be great but also a rare beast at this point. Can anyone verify?

ss

I am stunned with how quick you get the mag, I'm about a week behind. ss

Hmmm and it was about a week late this month.

I dont mean to seem to be looking at the glass half empty but I am suspicious of this tech. The reviewer notes that if you look for it you can see the active iris system at work and its this active iris that nets that phenomenal standard CR. Meaning it takes some visual ping-pong to get good blacks on the set. The set has a mediocre grey scale track and he also noticed some uniformity issues. Colors are oversaturated too, but he said a true production model might have that sorted.ss


I got the same impression reading his review - it scored some the highest numbers I've seen in their grading system - but there was quite a few low points, from mediocre tuner, color linearity issues, shiny graininess to the image, jagged edges, and a little noise in the shadowy areas.
I guess because of the sets high resolution, and high fill factor (I've seen 93-97% for SXRD) coupled with deep black and contrast comparable to CRT's must give it a close to film look which can be stunning compared to other technologies. But the closest I've come to this technology is reading Harknesses reviews.

If you are Sony you should be blushing after the comments by this same reviewer though so I guess I am just curious where we are setting that performance bar against the backdrop of really just higher res and bigger screens. 7 years ago a CRT with that performance would have been panned. And its got a bulb. Just noting the irony of where we have been and where we're going.
He also suggests the unit does 1080i 3:2 pulldown. If it does that would be great but also a rare beast at this point. Can anyone verify?ss

Since I like film in general - I'll probably have a deer caught in headlights look the first time I see it as well.

Maybe thats it. Maybe it is so large and believable that the downsides fade. I cant wait to see one in person. Film is the most important to me too.

I forgot he also complained about the screen. I am guessing its a type of lenticular screen as he compared it to most other RPTV screens he had seen. I wonder if its removable.

Still curious about 3:2 pulldown on 1080i....

Did you also understand that it throughputs but wont scale 720p?

ss

Jan 2006 issue arrived today. Flat Panel Bonanza
Two Plasmas: Pioneer 5060, Philips 42PF9630
Two LCD's: Sony V40XBR, Maxent 26X3
- very 1st post of thread has the data -

Hmmm is that a new black level record I see there? :eek:

ss

Yeah, the 42PF9630 seems to have a better contrast ratio than 8th gen panny? If anyone has any further insight give me a pm. I am interested in trading my a10 at bestbuy for the 42pf9630 or Maxent 50mx3

Can you think of a way they could translate the light output at 0 IRE to actual IRE. Something like, 0 IRE on the Panny 42 = 8.5 IRE actual.

ssI would think this would require the establishment of a reference display standard, as in "0 IRE on the Panny 42 = 8.5 IRE actual (as 8.5 IRE would be displayed on the <insert IRE reference display here>.

What I'd like to see more of is the black level measurement in foot-Lamberts that was used in the ANSI contrast ratio measurements, as well as the ANSI CR itself. With a few notable exceptions, this value is usually higher than the full-field black measurement.

Although I continue to update to the list - I've got to post every three months or so - so it doesn't archive.
Latest model tested a Westinghouse LVM-42w2.
Got a very nice review, if it hadn't lost points for Build Quality and Features (no tuners, nor cablecard) - it may have been one of the top models in their system as it did very well on Value and exeptionally well on Performance particularly for accepting 1080p through DVI and component video.

I was wondering if you could point me to the Westinghouse LVM-42w2 review? The review date says May 2006, which seems a bit in the future, and also, I searched Home Theater Magazine and Google and couldn't find it...

I was wondering if you could point me to the Westinghouse LVM-42w2 review? The review date says May 2006, which seems a bit in the future, and also, I searched Home Theater Magazine and Google and couldn't find it...

Not that far in the future. I just received the May issue today.

Well then, I eagerly await the online review!

3 LCD flat panels reviewed this month (June 2006) in HTM.

Sony KDL-40S2000
Hitachi 32HLX61
Sharp LC37D40U

Some table notes added because of selectable backlights, not sure if this will confuse everybody, but the number NOT in the bracket is the brightness that was measured to produce the highest contrast ratio. If you subscribe and read the reviews you would find that this was achieved in Sharps case with the lowest setting of the adjustable backlight. The TV is capable of producing much brighter images and the max they measured, is showed inside the brackets - not sure how record this in the future????

AND the lowest black level reported on a LCD....... or plasma for that point since they started this process.

I was somewhat surprised by Geoffrey Morrison's comments on the face off:
- I wouldn't want any of these displays.
- The Sony made everything looks like Cybill Shepherd in Moonlighting as in softer than hot margarine.
- The Sharp despite the great contrast ratio, black level and detail had processing anomalies that made some things just look weird.
- The Hitachi even though it moronically crushes black, something you can't defeat, at least it had the least amount of motion blur and the widest viewing angle so I would pick this one.

3 LCD flat panels reviewed this month (June 2006) in HTM.

Sony KDL-40S2000
Hitachi 32HLX61
Sharp LC37D40U

Some table notes added because of selectable backlights, ...
AND the lowest black level reported on a LCD....... or plasma for that point.

Wow. Has LCD's black levels finally cought up to plasma or is there some trickery going on. I'm going to have to check that Sharp out.

I'd like to thank semigolfer for the excellent work collecting data. It looks like a prediction I made less than a month ago has become half-true already - much faster than I thought. Here is the post made on April 28th 2006:

"I predict that within five years from this day (April 28th, 2011) at least one selling model of LCD display will have both superior black level and contrast ratio compared to the best plasma on the market, when adjusted via user-accessible controls (including any service menu present)." (http://www.avsforum.com/avs-vb/showthread.php?p=7574459)

Now for the contrast ratio...

I was somewhat surprised by Geoffrey Morrison's comments on the face off:
- I wouldn't want any of these displays.
- The Sony made everything looks like Cybill Shepherd in Moonlighting as in softer than hot margarine.
- The Sharp despite the great contrast ratio, black level and detail had processing anomalies that made some things just look weird.
- The Hitachi even though it moronically crushes black, something you can't defeat, at least it had the least amount of motion blur and the widest viewing angle so I would pick this one.


No kidding. I just read the article and all the reviews were dreadful.

No kidding. I just read the article and all the reviews were dreadful.

This guy is telling the truth. I think this feedback below would also help. Use this as a reference. But use your eyes for the final decision. Super In Plan Switching is the technology you need to look into. I owned a Hitachi 37HDL52 myself. It is not cheap. I got it for $1849. If you Google it, you can find some good deals. One more, this TV does not have HDMI. It has DVI, which is the same on the video side. BUt you need to use a component cable for audio input.

"Forget about the advertisement. What I have learned through my own research and years of experience in the semi-conductor industry has boiled down to this:
the difference in LCD panel making methodology is by far the most significant cause to the LCD picture performance. Sharp's LCD making method is TN, Sony
(the much hyped Braviia line) and Samsung use S-PVA. Hitachi, LG/Philips, Syntax, Dell, Westinhouse and others use S-IPS. Based on the testing results, S-IPS (Super In-Plane Switching) is simply the best overall choice. TN has absolutely been outclassed by S-IPS in viewing angels, response time and black

levels. S-IPS is the technology originally developed by Hitachi and later licensed to LG/Philips, Zenith, Syntax and some others. Having spent years in the

semi-conductor industry and traveled to Asia more often to Boston, I understand the S-IPS pretty well. I thought is is just a matter of finding one set and

make a quick side-by-side comparison. I looked at both LG and Syntax LCDs. Both have slightly better viewing angles than the Sharp LCD and noticeably better handling of fast motion objects esp. around the image edges at a high speed. I could still detect easily the blurry moving edges of high-speed objects. on Sharp LC-37D40U. But my frustration started with the black levels on LG and Syntax, which are disappointing to say the least and did not fare better than the Sharp LC-37D40U. The Syntax one is horrible in terms of displaying the low gray scales (conventionally called black levels). I could not see the subtle details in any dark objects such the textures of dark clothing and shadows of trees. All I see is this fuzzy darkness. LG one is, on the other hand, so much over the top. The pictures are so hard and feel like they are "screaming" at you. It does not have the color naturalism and the human skin tone is way off (the red push on human faces is unbearable). I finally turned to a friend in the TV business, who told me to look at the S-IPS originator, Hitachi. After doing some reading on their site, I was very much into Hitachi's Director Series' 32HDLX51 due to its AS-IPS (Advanced Super In-Plane Switching), cream of the crop, but I need a bigger size than 32". I decided to go for the 37HDL52. It took some time to locate one in a local shop (but I bought on-line, why not?) as it is not sold in any national chain stores like Best Buy or Circuit City. 37HDL52 beats this Sharp 37 incher hands down in all aspects of picture performance the way S-IPS should top TN. But what really helps me close the deal is its ability to display black levels. I challenge any serious buyer to find a LCD which can render better black levels than this Hitachi line of LCD. I have not found one. On a side note, I believe Hitachi made a stupid mistake to license its technology to these companies who have failed to fully capitalize the full strength of S-IPS. Last but not least, what bothers me quite a bit are the key specs from Sharp such as 1200:1 contrast ratio and176 degree viewing angles. Well, they are not even close to my naked eyes. It, at best, has a
120 degree viewing angle esp. on color shift."

Great thread - thanks for keeping the first post up to date...

This guy is telling the truth. I think this feedback below would also help. Use this as a reference. But use your eyes for the final decision. Super In Plan Switching is the technology you need to look into. I owned a Hitachi 37HDL52 myself. It is not cheap. I got it for $1849. If you Google it, you can find some good deals. One more, this TV does not have HDMI. It has DVI, which is the same on the video side. BUt you need to use a component cable for audio input.

"Forget about the advertisement. What I have learned through my own research and years of experience in the semi-conductor industry has boiled down to this:
the difference in LCD panel making methodology is by far the most significant cause to the LCD picture performance. Sharp's LCD making method is TN, Sony
(the much hyped Braviia line) and Samsung use S-PVA. Hitachi, LG/Philips, Syntax, Dell, Westinhouse and others use S-IPS. Based on the testing results, S-IPS (Super In-Plane Switching) is simply the best overall choice. TN has absolutely been outclassed by S-IPS in viewing angels, response time and black

levels. S-IPS is the technology originally developed by Hitachi and later licensed to LG/Philips, Zenith, Syntax and some others. Having spent years in the

semi-conductor industry and traveled to Asia more often to Boston, I understand the S-IPS pretty well. I thought is is just a matter of finding one set and

make a quick side-by-side comparison. I looked at both LG and Syntax LCDs. Both have slightly better viewing angles than the Sharp LCD and noticeably better handling of fast motion objects esp. around the image edges at a high speed. I could still detect easily the blurry moving edges of high-speed objects. on Sharp LC-37D40U. But my frustration started with the black levels on LG and Syntax, which are disappointing to say the least and did not fare better than the Sharp LC-37D40U. The Syntax one is horrible in terms of displaying the low gray scales (conventionally called black levels). I could not see the subtle details in any dark objects such the textures of dark clothing and shadows of trees. All I see is this fuzzy darkness. LG one is, on the other hand, so much over the top. The pictures are so hard and feel like they are "screaming" at you. It does not have the color naturalism and the human skin tone is way off (the red push on human faces is unbearable). I finally turned to a friend in the TV business, who told me to look at the S-IPS originator, Hitachi. After doing some reading on their site, I was very much into Hitachi's Director Series' 32HDLX51 due to its AS-IPS (Advanced Super In-Plane Switching), cream of the crop, but I need a bigger size than 32". I decided to go for the 37HDL52. It took some time to locate one in a local shop (but I bought on-line, why not?) as it is not sold in any national chain stores like Best Buy or Circuit City. 37HDL52 beats this Sharp 37 incher hands down in all aspects of picture performance the way S-IPS should top TN. But what really helps me close the deal is its ability to display black levels. I challenge any serious buyer to find a LCD which can render better black levels than this Hitachi line of LCD. I have not found one. On a side note, I believe Hitachi made a stupid mistake to license its technology to these companies who have failed to fully capitalize the full strength of S-IPS. Last but not least, what bothers me quite a bit are the key specs from Sharp such as 1200:1 contrast ratio and176 degree viewing angles. Well, they are not even close to my naked eyes. It, at best, has a
120 degree viewing angle esp. on color shift."


I am interested in this Hitachi 37HDL52. Does it come with HDMI? If not, I need to buy a DV-HDMI cable.

I am interested in this Hitachi 37HDL52. Does it come with HDMI? If not, I need to buy a DV-HDMI cable.

FYI, new models were recently announced for a Fall release and they sound like actual improvements over the current ones. There was a thread about this recently but the mods merged it with the Hitachi plasma thread for some reason. :rolleyes:


Google something like Hitachi 120hz lcd and you will find info on them.

I am interested in this Hitachi 37HDL52. Does it come with HDMI? If not, I need to buy a DV-HDMI cable.


Mine does not have HDMI. But I have learned that Hitachi made a transition sometime this year from DVI to HDMI. As a result of that, they add an "A" to the end of the model. So you need to get hold of 37HDL52A.

FYI, new models were recently announced for a Fall release and they sound like actual improvements over the current ones. There was a thread about this recently but the mods merged it with the Hitachi plasma thread for some reason. :rolleyes:


Google something like Hitachi 120hz lcd and you will find info on them.


I have pre-ordered one from a dealer for my vocation home. But the story has kept changing. The latest news is that Hitachi will not introduce the 120Hz LCD this year. According to the source I have, the reasons are 1) The competitive price has come down so much on plasma. LCD will have a hard time making a strong value proposition. 2)The new LCD has no means to reach the necessary Economy of Scale in order to have an attractive price point.

Just a post so that the thread doesn't archive - where it can't be updated.
Nov 2006 issue data.

This is such a great thread. Good job!

ANSI Contrast is what really matters right? It's the contrast at any given moment if I understand properly.

The plain Contrast listed is actually the Dynamic Contrast?

Interesting that the CRT has such low ANSI contrast. I've heard that it's because of internal reflections caused by the glass...any info about this?

ANSI Contrast is what really matters right? {It's the contrast at any given moment if I understand properly.}
Not exactly. It is what it is, it's the measurement of a checkerboard pattern - that is supposed to somewhat simulate a more real world performance level of the display. What a person actually watches is seldom all white or all black screens - they see pixels of all variations of color and intensity. ANSI kind of gives you an extreme version of how the display shows this blending of light. And whether the black levels are somewhat negated by the sets light output. Here is an article (http://www.hometheaterhifi.com/volume_13_2/feature-article-contrast-ratio-5-2006-part-1.html) although written for projectors gives a nice description of Contrast Ratio - be sure to at least the first two pages.

Personally I find ALL of the numbers provided as being useful. I like to know the Black level, because I found from experience that sets with the deepest blacks are what my eye show the most looking out the window picture. For me the black level number (coupled to the sets gamma correction capablilty - not discussed here) is my starting point for selecting a TV.
Although I find the Brightness numbers less meaningful, because I only watch TV in light controlled rooms, - but they help explain some high contrast numbers that my eye tells me aren't quite as "real" looking. However this information could be helpful to help me narrow my choices if I didn't have light controlled rooms, although I would want a low black level I might also look for a brighter TV such as an LCD. As for the contrast number my thoughts are below. And the ANSI number kind of sets a lower base number of what the contrast is likely to be (read the article referenced above). It should be noted, as somewhat explained in the description below the table in the post, - plasma TV's limit full field white output - I think I read somewhere that it limits total power use and protects the electronics, but this explains why several plasma TV's have higher ANSI numbers than their std contrast ratios.

The plain Contrast listed is actually the Dynamic Contrast?Again this is a case of it is what it is. The plain contrast ratio shown is found by dividing the Brightness Level by the Black Level that they recorded. Is it Dynamic Contrast well yes it can be. Because it is a 100 IRE field divided by a 0 IRE field measurement - so it's full on versus full off (if full off is defined as a signal level of 0 intensity). But is it the Dynamic Contrast reported by the manufacturers - the answer is no. What it is, is a consistent measurement of full on/off measurement of light output. They do it the same, time after time. Why isn't the same as reported by the manufacturer - who knows? The number someone could easily play with is the blackness level. When someone says full off - do they actually mean 0 IRE or do they not even apply a signal at all. No signal at all- eliminates the light entirely - thus lowering the black level and inflating the contrast ratio number. Who knows what other games could be played with filters etc.. At least Home Theater Mag has made the process consistent thus comparisons between different manufactures are more likely to be closer to the truth.

Interesting that the CRT has such low ANSI contrast. I've heard that it's because of internal reflections caused by the glass...any info about this?That's what I've read to date. The deep black level of the CRT gets washed out by it's own light.

That article was a big help as was your reply(ies). Very interesting.

So I take it that the contrast ratio is the full on/off static contrast ratio which is quite meaningful. And so...the Hitachi plasma performs rather poorely compared to the Panasonic plasma. Although it's interesting to see that the Panasonic has better numbers than the Fujitsu as well...

It also looks like the LCDs are as dark as the plasmas? That seems hard to believe...

The Sony SXRDs also seem to outperform the HD-ILAs number wise.

Just a post so that the thread doesn't archive - where it can't be updated.
Nov 2006 issue data.


yeah I wonder if there will ever be a update this year for the newer models.

yeah I wonder if there will ever be a update this year for the newer models.
Just so you know I try to update after each issue - as you will find under the issue date (i.e. Date Reviewed). As for the current issue (Jan 2007) the venerable Sony 46XBR2 has just been reviewed - which BTW recieved a very good review regardless of what they deemed poor scaling of non high def signals..

Wow some of the LCDs are just about as dark as the Panasonic plasmas? But when they are that dark they lose contrast ratio? I'm confused...how does that work? Perhaps it can't have as bright a white with the darkest possible blac?

Wow some of the LCDs are just about as dark as the Panasonic plasmas? But when they are that dark they lose contrast ratio? I'm confused...how does that work? Perhaps it can't have as bright a white with the darkest possible blac?
Hopefully the way I've reported the results isn't too confusing. For LCD's the numbers in the main column for brightness, and black level are those that yield the max contrast ratio. Although not the only factor, one of the main factors for LCD's achieving much improved contrast ratio has been the incorporation of adjustable backlights. This has driven the black levels downward. If you look at the leading LCD flat panel in black level the Sharp LC37D40U with a reading of .014 ft-L, at this backlight level it only has a brightness of 17.76, which is lower than many plasmas. This same set attains a maximum brightness level of 122 shown in brackets when the backlight is fully turned up, but the black level ALSO rises to 0.118 ft-L and the contrast level falls off to 1034 from it's maximum contrast measurement of 1269.
Since LCD's started incorporating adjustable backlights I started showing these multiple readings. Occasionally such as the case of the Sharp LC57D90U I've actually reported three readings because in this case the lowest black level did not yield the maximum contrast (although it was very close), so you will see numbers showing the maximum contrast, the maximum brightness and the lowest black level.

semi, thanks again for your thread. I come back here regularly for this info.

ss

Yes, Thanks Semi.

Mods, why is this not a sticky?

Here are the scientifically proven numbers
------------------------------------------------------------
The dynamic range of the human eye is .0001 to 10,000 or in other words 100,000,000 to 1 (7 decades).

The contrast ability of the human eye in daylight conditions is 1% or in other words 100:1

The contrast ability of the human eye in dark/dim conditions is 10% or in other word 10:1
-------------------------------------------------------------

In other words, the eye has a very large dynamic range but for a given luminence the contrast sensitivity (ability to discern luminence values) is very small!!!! and it changes with luminence. The lower the luminence the lower the sensitivity.


CheersThis is a very good way of describing it, Xrox.

Looks like the human eye has the ultimate Dynamic Iris :)

I hope OP continues to update this awesome apples-to-apples comparison of displays.!

The hardest thing to ever find at this forum are maningful apples to oranges comparisons of contrast ratio and black levels between different technologies.

How abou a sticky that dealt only with that and actually compared Plasma to LCD to DLP to D'ILA to SXRD?

Would the universe end with such a thread?

I'll ask semigolfer this: What has the best contrast ratio of any 1080p set currently produced?

If you can answer that semigolfer then you will be the most courageous person at the AVS Forum!

DBLASS posted this on another thread and I think it deserves to be posted here as well.
My turn....

Let's explore the history and the myths. Contrast ratio from the LCD mfgs are used to show how much light is leaking from the display. There are lots of technical reasons why LCDs leak light (polarizer angles, color filter alignment, etc). As CR numbers improve, it tells you that you are getting a more advanced LCD. Again.... LCD mfg use a darkroom so it is purely white (500 nits or so), divided by black (.5 nits, to .35 nits, to .20 nits, etc). This will produce a 1000:1, 2000:1 or whatever, contrast ratio.

Plasmas mfgs jumped on the CR bandwagon because their darkroom CR is 10,000:1 or better... because plasmas DON'T LEAK LIGHT on a black screen.

All of you out there seem to back up my personal feeling that ALL consumers buy on numbers. You want to buy the largest numbers for the least amount of money. 10,000:1 has to be better than 2000:1. Why? Cuz its a bigger number.

The myths are.... The human eye IS NOT A BLOODY LIGHT METER!!!!! It just doesn't work that way. We (humans) PERCEIVE brightness. We cannot register "500 nits" if we had to. We can only perceive .5 nits of black. We do not measure it with our eyes.

All lighting conditions will and does WILDLY AFFECT the CR on any display (as mentioned). It is a moving target. The display factories know this WHICH IS WHY THEY TEST IN AN UNREAL, YET CONTROLLED TEST PARAMETER--- A DARKROOM.

In doing so, they have set off the war between plasma guys and LCD guys chasing irrelevant numbers. CR in a spec is STUPID....... arghhhhh!!! It depends on all external light conditions.

One last thing to start the screaming again... LCD will reflect a total of 5-6% and plasma reflect more (I'm guessing here) about 12%. Since it is reflection that what washes out the display AND wildly lowers CR, these reflection numbers are key to performance of the technologies. To make things just a tad bit more confusing, the reflection off a LCD is diffused so the reflection is a large "flare ball" while the plasma is specular (shiney) so the reflection is mirror-like. Optics guys argue here but diffused reflection is worse when it come to controlling the reflection but "average consumers" HATE seeing their face on the display.

There is just no pleasing anyone.

At the risk of sounding like a broken record (see my post somewhere on the LCD forum). Contrast measurement is only ONE thing:
white+ reflection divided by black + reflection. PERIOD.

Reflection will make white a bit whiter BUT reflection REALLY makes black whiter.

They don't use reflection in the numbers because it is situation dependant: light source, angle, intensity (focused or diffused), other reflections.

Reflection is a key comparison in LCD spec because the mfgs are always trying to get the LCD to block better (darker blacks). Since real world testing is impossible to control, they do the test in a dark room. A typical test will read:

white (500 nits) + zero divided by black (.5 nits) + zero = 1000:1 CR

The human eye perveives brightness. Light meters read brightness

Since CR is a mathmatical formula, the human cannot "read" CR, only percieves it.
Lastly, real world CR for a good looking display is about 10:1 to 20:1. Yep, it really is that low.

Why is it that such things are never compared in a room that is EXACTLY halfway between being in a dark room and in a bright room?

Could it be because then no technology could have an ambient lighting excuse?

Again, you can't control the test environment enough to keep all the players happy. The US Air Force use to use a spec that specifically defined the test conditions. It was a lot easier. The funny thing was that the spec never called out "Brightness" for sunlight readable displays. It simply read "the display will have a 3:1 contrast in a 10K fCandle environment, with the light source coming a 30 degree angle of incidence".

You read right... for a cockpit display, they called for 3:1 !!! That all it takes to keep a pilot happy. Think it was easy to do? A display (with front window) would reflect without any tricks 15% of the light source. 10,000 fCandles is close (long story) to 10,000 nits. This means that the displays Black would read 1500 nits. How bright does White have to be to compute to 3:1 contrast. 4500 nits of white. With real optics, it can be done easily.

Who cares whether the players are happy?

It seems to me that it is not beyond the mind of man to produce a testing room where the ambient lighting conditions would not favor any two techonologies that are compared to one another--one that would be exactly midpoint between the two in offering advantages and disadvantages.

In such an environment REAL comparison COULD be made.

Any optical scientist could produce such a testing room. It is only here at the AVS Forum where manufacturers jealously guard what they like the most--obfuscation of REAL contrast ratio differences--so the video consumer is confused enough to where they are more able to sell bad product--that such things are impossible!

Any room could be set up. But the results could be meaningful or not. The point to all of this is that contrast ratio as a number, is sadly, meaningless. Considering only light meters can be used to calculate a CR, the human eye will not percieve it the same, person to person.

There may be object ways to compare technology but CR may not be the most meaningful. A test setup, creating a real life environment, would should that both LCD and plasma suck, differently and for different reasons. But by the way, the 70 year old CRT REALLY SUCKED but we all put up with it. Oh, and did I mention DLP, they suck and projector CRT, they really suck and LCOS, yep... sucky too.

THEY ALL REFLECT AND IN YOUR LIVINGROOM, THEY WILL ALL HAVE LOW CONTRAST RATIOS.

Just assume that real world contrast ratios is fools game. Pick a technology for things like response speed, color gamut and things will go a lot easier.

It will also be much more convenient for the Video display manufacturers.

The real fact is that contrast ratio has more bearing on Picture Quality than any other parameter.

Avoiding trying to quantify it is as silly as Car and Driver or Road and Track attempting to compare cars without ever mentioning horsepower.

Is contrast ratio everything? No. Is horsepower everything? No.

Are contrast ratio and horsepower unmeasureable and MEANINGLESS when it comes to comparing Video displays or cars?

What FOOL would believe that?

Contrast is very important!!! No one has debated this. What has been debated is the validity of the numbers reported.

I do not remember reading that CRs are unmeasurable or meaningless. What several people have tried to point out is that the current method for determining this ratio has little application to real world viewing. You may spend all of your viewing time in a completely dark room, if so then the ratios may be right on the mark for you. And in the long run, all these wonderful numbers/ratios seem to fade once light is introduced into the equation.

No one has avoided answering your questions. You were given an example that compared two displays - one plasma, one lcd. It seemed that you did not like the answer, you questioned the environment, testers etc. you then moved to another thread and asked the same questions......received the same answers.

Nmlobo: If you were going to devise a method for determining contrast ratio that WOULD have application to real world viewing what do you think a fair test would be?

Without a doubt most people at night would probably view movies in a darker environment but not pitch black by any means.

In the daytime most people would probably view TV in a bright environment but certainly not with the screen washed out by glare or the sun--that's not to say that some lights might hit the screen or that some sets might have reflections of some kind--but I would imagine that most people would make a small effort to watch TV in the day with bright ambient light but not washing out light.

So I really believe that you would need two different measurements. I'd call them daytime TV viewing and nightime cinema viewing.

Other than the review cited you never get such numbers.

Not to mention the fact that two different video review magazines will spout markedly different numbers on the same set.

Some will say nothing is stadardized when it comes to measurement.

Does anyone think that is just by accident?

What would the world be like if the government mandated a single standard for measurement in a darker setting and a lighter setting and all sets had to post those values?

I think it would be a better world. I know that world will never happen.

What I don't understand is why some posters here are glad it will never happen and are glad that no one knows any numbers at all. Where does that mindset come from?

Where does glee at not having anything in relation to Video displays quantified on a standardized basis come from?

I'm not saying that is your attitude--I am saying I have run across that attitude here at the forum and I think it is video Display manufacturing companies generated.

You know it's a real shame with all the contributors here who might could give some numbers in the ballpark that if you really want to know a set's contrast ratio you have to buy the measuring equipment--set up the environment and test the display yourself.

It kind of makes you wonder if this place is more an enthusiast's forum or a manufacturers sales generator.

Nmlobo: If you were going to devise a method for determining contrast ratio that WOULD have application to real world viewing what do you think a fair test would be? . . . IMHO it would be impractical to develop a test that would meet every unique condition. I also understand that manufacturers would not want to conduct multiple tests, testing costs $$. If I was going to develop a test, I would suggest a standard testing environment where the testing room would have a set level of light, certainly not a completely dark room. I would also define a standardized placement of the light sources (i.e. lamps, windows, etc). Each test article would be calibrated for best black levels and gray scale. This would verify that blacks and whites are not crushing detail. Measurement equipment would be at a fixed distance from the test article, avg./recommended viewing distance. I would measure both full-on/full-off and ANSI CRs. I would report the CR achieved, distance, lighting, and gray scale. I would include a statement that when viewed under different lighting conditions, sets would achieve different results.

Other than the review cited you never get such numbers.

Not to mention the fact that two different video review magazines will spout markedly different numbers on the same set. I looked at CNet’s review of the Pioneer PRO-FHD1, the same set reviewed by Home Theater. CNet did not measure CR but only reported the manufacturer’s numbers. I believe this is a standard practice.

What would the world be like if the government mandated a single standard for measurement in a darker setting and a lighter setting and all sets had to post those values? Don’t we have enough gov’t involvement already?
You know it's a real shame with all the contributors here who might could give some numbers in the ballpark that if you really want to know a set's contrast ratio you have to buy the measuring equipment--set up the environment and test the display yourself.

It kind of makes you wonder if this place is more an enthusiast's forum or a manufacturers sales generator. IMHO the average consumer does not care about numbers/ratios, other than for using them as a guide – most understand the higher the ratio the better. I believe the average consumer could care less if the real CR on their set is 700:1 or 7000:1. They don’t necessarily know or care that a number/ratio was obtained in a sterile environment and that it may or may not be representative of their home environment or an indicator of their viewing pleasure. They could care less if the CR reported for a neighbor’s set is 200 or 2000 points higher/lower than the one reported for their set. What they care about is how the display will look once they get it home. They use the number as a guide, not bragging rights.

IMHO This is how we should use the numbers, as a guide. It is not a number but the display in your home environment that provides your viewing pleasure. This is why some of us like plasma, some lcd, some crt, some projection sets. If we relied on only numbers I doubt there would be a clear winner with any technology.

Artwood- Interesting you should use a car analogy. Consider Contrast like mileage ratings. As we all know, cars are rated for mileage (MPG). Now, you might think that 26 MPG is better than 24 MPG, but what would you say if you owned that car and never got anywhere close to that mileage? Mileage ratings are done in a lab on a dynomameter. As they say, actual mileage depends on driving style, raod conditions, how the car is loaded and equiped, etc.

CR is a base-line spec. It actually is meaningless in an environment because the actual CR in a lit room is very low, BUT if you compare the SAME CR test rating, ie, darkroom white/black screen, then all displays tested the same way will tell you something (though not everything). In this case a 3000:1 CR display will always hold a wider CR than a 1500:1 display no matter the environment. This means that if the tests were the same, a higher CR will hold a higher CR in the real world. The better display may wind up at 20:1 in a lit room while the lower CR display will be 15:1.

The better display may wind up at 20:1 in a lit room while the lower CR display will be 15:1.

DBlass, you make some very good points but what explains a higher rated CR LCD losing out to a lower rated CR plasma in lower lighting conditions?
Are you saying one CR rating is not as true as the other?
I think this is what is confusing to people.

Plasma CRs can be very high but to some extent irrelevant (see below). It does explain that given a set environment all CR of all displays drop when you compare "real" numbers (real environment) versus "lab" numbers.

I am saying that only under very specific conditions is a CR rating valid and it is less valid as you cross technology lines. If plasma do not emit light on a black screen, then their CR is somewhat meaningless because it is essentially infinite. Contrast on an LCD was used as a measure of how well/bad they block light since their backlights are on all the time. When comparing one LCD to another, then if the test is the same test, say white/black in a darkroom, then the results will tell which display blocks light better.

Comparing LCD versus plasma is much less meaningful since a CR test (white/black in darkroom) for plasma really does not mean much since the answer is infinite. LCD have a lower reflection percentages since they only really reflect off the front surface while plasmas have greater internal reflections. Having said that, LCD have a diffused (dull) front surface that scatters the light of the reflection, making an uglier reflection. Plasma's smooth glass front relfects more but in a controlled manner.

Now, if one could outline a specific environment with a prescribed light source, angle of insidence, reflectivity of the walls, etc. then a test: either white/black= CR or White+reflection / Black+reflection= CR THEN the comparison would not only be interesting but very compelling as to how the technologies stack up.

The answer is NOT digital. It is analog; meaning "it depends".

Most people here are looking for the "digital"answer. Name the winner, name the loser.... it isn't going to work like that.

Contrast ratio aside, pure black levels on that chart seem to indicate that good LCDs are within shooting range of good Plasmas, and Sony's particular flavor of LCOS is the only thing that comes close to approaching the black levels & overall contrast of good ol' CRT.

The truth is CRT actually have crummy contrast in a lit room.

Could someone help me translate this:

WestinghouseLVM-42w2 57.19(158) 0.064(.192) 894 (825) 712 1000 MAY 2006

If I'm generally on track, the point is that high contrast ratios can come from OK blacks and very bright whites or conversely OK whites, and very dark blacks. Either will give the same 'ratio' and this chart is to tell just what the whites and blacks are. Where I get lost is the brackets. Is the white 57 or is it 158? Is the black .064 or .192? Any comments on this monitor's scores appreciated.
Thanks in advance.

Could someone help me translate this:

WestinghouseLVM-42w2 57.19(158) 0.064(.192) 894 (825) 712 1000 MAY 2006

If I'm generally on track, the point is that high contrast ratios can come from OK blacks and very bright whites or conversely OK whites, and very dark blacks. Either will give the same 'ratio' and this chart is to tell just what the whites and blacks are. Where I get lost is the brackets. Is the white 57 or is it 158? Is the black .064 or .192? Any comments on this monitor's scores appreciated.
Thanks in advance.

The highest contrast ratio recorded is: 894
It is achieved with a brightness (100 IRE) of 57.19, and a corresponding Black Level (0 IRE) of 0.064. Which means that the highest contrast is achieved with the backlight probably set at a very low level.

But the unit can also be very bright. With a brightness (100 IRE) of 158 ft-L, and a corresponding Black Level of 0.192 ft-L, and this yields a contrast ratio of 825. In this case the backlight level is probably at it highest level, and thus the black level suffers.

And the 712 represents the units ANSI contrast level, as explained in the write-up.

And finally one of the vendors selling the unit published a manufacturers specified contrast ratio of 1000.

The highest contrast ratio recorded is: 894
It is achieved with a brightness (100 IRE) of 57.19, and a corresponding Black Level (0 IRE) of 0.064. Which means that the highest contrast is achieved with the backlight probably set at a very low level.

But the unit can also be very bright. With a brightness (100 IRE) of 158 ft-L, and a corresponding Black Level of 0.192 ft-L, and this yields a contrast ratio of 825. In this case the backlight level is probably at it highest level, and thus the black level suffers.

And the 712 represents the units ANSI contrast level, as explained in the write-up.

And finally one of the vendors selling the unit published a manufacturers specified contrast ratio of 1000.

Thanks for the feedback, that explains it very well. I didn't realize the backlight was adjustable. So to the trained eye, would you call .064 a good black level, something that would look good in a dimmed but not dark room? This is my usual viewing environment.

Any room could be set up. But the results could be meaningful or not. The point to all of this is that contrast ratio as a number, is sadly, meaningless. Considering only light meters can be used to calculate a CR, the human eye will not percieve it the same, person to person.

There may be object ways to compare technology but CR may not be the most meaningful. A test setup, creating a real life environment, would should that both LCD and plasma suck, differently and for different reasons. But by the way, the 70 year old CRT REALLY SUCKED but we all put up with it. Oh, and did I mention DLP, they suck and projector CRT, they really suck and LCOS, yep... sucky too.

THEY ALL REFLECT AND IN YOUR LIVINGROOM, THEY WILL ALL HAVE LOW CONTRAST RATIOS.

Just assume that real world contrast ratios is fools game. Pick a technology for things like response speed, color gamut and things will go a lot easier.
The new LCDs have the darkest (very low light reflection) of any other technologies screens. For bright room watching, they can't be beat. My LCD has a manufacturers CR of 800, it's great. The numbers don't mean anything. What is more important is does it crush blacks and whites and does it get bright enough for the viewing environment while not doing that. Ones eyes provide the best comparison.

The truth is CRT actually have crummy contrast in a lit room.
But in a non bright room the CRT is king for blacks and shadow detail. CRT-RPTV being the best as there is less glare and no reflections off the screen when there is no protective screen. My CRT-RPTV can't compete with my LCD when the room is bright but the LCD can't compete with the CRT-RPTV in subdued light.

Thanks for the feedback, that explains it very well. I didn't realize the backlight was adjustable. So to the trained eye, would you call .064 a good black level, something that would look good in a dimmed but not dark room? This is my usual viewing environment.
Yes I believe your phrasology is accurate. It could be classified as good. Here's what the reviewer had to say: "If black level is your primary concern, this display bests many of the LCD's reviewed, producing 0.064 foot-lambers at the 0 backlight setting"

and elsewhere in the same review: "You can adjust the LCD's backlight in increments from 0 to 100. We measured the monitor's black level and light output at the 0, 50, 100 backlight settings; much to our surprise, the contrast ratio was excellent in all three (894:1, 824:1, and 879:1 respectively)"

But as you can see from other numbers in the table there appears to be several other brands/models that are clearly in a darker category, with the lowest two measured (Sharp D40U, and Sony XBR2 at 0.014 and 0.018) why not trek on down to the store and see if this matters to you - just make sure you adjust the backlight before viewing. (Note: Let's see - avg these two and you get, 0.016 then divide this into .064 - and you get a factor of 4 - not sure if this means 4 times blacker.)

In case they're archiving when they bring the site down Sunday - thought I'd post to keep the thread alive.
In the mean time that recent test (July '07) of the Sharp D92 set some new records on contrast (shown in table - 1st post of this thread) althought the overall review was not near the best we've seen for LCD's.

Well all good things must come to an end....and this appears to be a good time to stop reporting these numbers. Started in Aug 2004, ending Dec 2007. If you stumble onto this thread - all data collected is stored in the very first post of this thread.

Personally I've got a 50" plasma, I'm thrilled about, a 46" LCD that also makes me happy, and I'm in the market for a front projector for my new home theater room, so I probably won't be using this information to help me narrow my list of future purchase choices (since once I get a front projector there likely won't be any additional purchases for quite some time). And although I have never considered contrast levels the "be all" performance spec it is one of the things I look at first to help narrow the field. It has been an important one for me and has helped me purchase sets that I will be happy with for years. Plus generally there hasn't been much interest in this thread to go through the trouble of maintaining it.

It's also timely because HTM just announced that after testing the Pioneer Elite Kuro PRO-110FD that it is indeed the finest flat panel (plasma or LCD) they've ever seen. What did they say :eek: - twice as good as any other flat panel. Bet you can't wait to see it numbers (go to the very first post of this thread)

On the LCD side, manufacturers have also improved their numbers and performance, but they've (mostly Sony/Samsung) ALSO come up with some circuitry (let me mostly quote from the article here) that "is riding the video signal with its backlight depending on the average brightness level, artificially driving black and white levels to reach better numbers with static test patterns which create artificially high contrast ratio measurements, much like using an auto-iris on a projector"
thus the value of having these numbers will require the perspective buyer to do even more research to understand them.

In closing - good luck on your purchases!

senigolfer, I see you updated the list for the new Pio plasma, but you didnt red highlight it across the board. Its tops in almost every category. A truly amazing set.

ss

Why are you no longer updating this thread? I love this thread, it's a great reference!

I agree, great thread.

Though admittedly, I'm rather biased.

i would love to know how the new 1,000,000:1 contrast ratio on the sharp limited edition aquos models tests out. not that anyone out there even has one....