Screen comparative test : CRT vs LCD - BeHardware
>> Miscellaneous >> Monitors

Written by Vincent Alzieu

Published on August 24, 2004


Page 1

Screen comparative test : CRT vs LCD

Screen comparative test : CRT vs LCD
More and more screen manufacturers are leaving the CRT market. Mitsubishi, LG and Samsung are the last three manufacturers. Even Sony, the Trinitron tube inventor, left the market. Iiyama, Viewsonic, Acer and a few other manufacturers still have a couple of CRT products but admit that they are leaving progressively the market and that they want to put forward their LCD screens.
Some of the manufacturers forecast a dark future for the CRT screens. They think that CRT screens will completely disappear in the next four years. According to these manufacturers everyone will use LCD screens. LCD screens are less bulky, more aesthetical, more practical, but also more expensive. So do they really represent a real progress? This polemical question needs to be clearly exposed.

There are dozens of screen comparative tests online and in magazines. However most of them are dedicated to LCD screens and some of them (rarer) to CRT screens. Now is the time to put face to face these screens and by looking at the challenger’s records this duel will be very interesting.

The challengers choice

On the left side of the ring, the BenQ FP767-12 with the 12 ms AU Optronics panel. This 17” equipped with a TN + Film panel is one of the best LCD screen ever released. Compared to the TFT equipped with Hydis 20 ms panel, the FP767-12 is less reactive but has more homogenous brightness, more accurate colors and wider viewing angles. MVA and PVA screens have even more wider angles but their response time measured is way below the BenQ.
The FP767-12 is an excellent reactivity and color quality compromise.

On the other side of the ring, the CRT Mitsubishi Diamond Pro 930SB. The name of the screen: Diamond Pro is already a proof of quality. This range is said to be the best on the market. The picture geometry is good and the product characteristics are above average: maximum definition 1920 x 1440 pixels 73 Hz, OptiClear (remove the possible reflections), horizontal frequencies 30 – 110 Khz…

Despite different sizes announced, the manufacturers recommend using both of them with the 1280 x 1024 resolution. The 767-12 screen size is 17” and the 930SB with the right settings reach 17.7”.

So the Mitsubishi has the biggest display area.

Page 2
The bill : buying + electric consumption

The bill : buying + electric consumption
LCd screens are more expensive and the FP767-12 isn’t an exception.
The NEC screen is just a little bit cheaper. The Diamond Pro 930SB is more expensive than a standard 19” and is one of the most expensive screens of its category. However it is possible to find it cheaper online.

Buying the product is one thing, but the electrical consumption is another thing. If these screens are switched on almost all day long, using them has a price. We used a "Power Control" tool to measure their electrical consumptions. According to this measuring tool, the BenQ 767-12 needs 43 Watts. A screen switched on permanently will cost 43 watts / 1 000 x 24 hours x 365 days x 0.1057 dollar per kWh = $49.84. For five years the electrical consumption will cost $249.21. If you use moderately your screen 1 hour every two days (200 days a year for example) the annual cost will be $1.1 only.
The DPro930SB requires 91 Watts. The annual cost is $105 and $527 for five years if the screen is permanently switched on.

Here is a graphic of the overall costs for three and five years. These graphs include the buying price plus the electrical costs 7 days a week 365 days.

This isn’t really surprising: LCD screens are cheaper to use with time than CRT screens. Using an LCD or a CRT screen 18 hours, 365 days a year has the same cost. However if the screen is permanently switched on, the LCD is cheaper to use.

The results are slightly different for five years. The CRT is cheaper to use if it is switched on less than 10 hours a day.

Adding the buying price to the electrical consumption shows that the BenQ is cheaper than the Mitsubishi screen for daily long uses. If the screen isn’t constantly used, the CRT technology remains cheaper in along run.

Of course with a $100 cheaper screen results might change the results. The Mitsubishi MultiSync 95f screen is cheaper to buy but has a higher electrical consumption 125W instead of 91 W (Nec-Mitsubishi measure). If the screen is permanently switched on the electrical consumption will be $724. The overall cost including the buying price is $1009 and is superior to the DPro930SB.

Page 3
Design, bulkiness

Design, bulkiness
The FP767-12 is 15.9“ high and 14.9” wide and 7,5” thick.

Of course, the Mitsubishi screen is the bulkier (17.5” x17.60”x17.7”). This cubic shaped screen is three times bigger than the flat screen and generates a waist of space on your desk. For an optimum vision comfort, the distance between the user and the screen must be: 1.5 x the exact screen size diagonal.

The size of the monitors is approximately 17”, so the right distance is: 17 x 1.5 = 25.5 in.

25.5 + 7.5 (the BenQ thickness)= 33” from the back of the FP767-12 screen to the user. Any type of desk is wide enough to optimally use this screen.

25.5 + 17.7= 43.2” from the back of the Mitsubishi screen to the user. These time standard desks are not big enough for this screen.

It is not surprising but the LCD screen win the bulkiness award. However if your desk isn’t too close to a wall it is possible to put the back of your screen a little bit beyond the desk limit.

Page 4
Installation, when a graphic card formats a screen

Installation, settings, inputs
Both screens only have analogical inputs (VGA or D-Sub 15 pins). A DVI input could have been an asset for the FP767-12. The DVI input avoids the synchronization phase between the clock and the screen. In practice the difference is limited and 98% of these synchronizations with the graphic card signal are made without any problems. From the first time you start your computer the screen image is sharp and has the right position.

Initiating the Mitsubishi screen has been more difficult. We couldn’t get a picture on the screen. First problem, in clone mode the graphic card could see the CRT screen on the main input. And second problem the screen LED was green meaning that the screen was switched on. With windows on the black CRT screen and on the LCD screen in clone mode we realised that the resolution and frequency parameters were abnormally high. Phone call to NEC to understand the failure, one question was enough to understand our issue: Is the graphic card in your computer a NVIDIA? Yes, I’m using a GeForce FX 5950 Ultra for my tests. Answer: this is a problem specific to these series, NVIDIA’s graphic cards formats these screens! And that isn’t a joke.

The problem happens with the DPro930SB, DPro930SB-BK and FP912SB series and all the GeForce 2, 3, 4 and FX graphic cards. When the computer is switched on, the graphic card starts communicating with the screen. The monitor Eeprom, a ROM memory (read only), send to the graphic card the VESA information: maximum resolution, screen refreshing rate, colorimetric etc. Here, this exchange doesn’t work. The graphic card set a frequency not compatible with the screen, 150 Hz in 1920 x 1440 pixels, and modifies the Eeprom information. This operation shouldn’t be possible. But Mitsubishi choose to use a rewrite instead of a ROM memory to simplify their production line. The same memory is used for several screen ranges. The only difference is the information included in the memory. NEC acknowledged that the problem was connected to their memory. However this problem only happened with NVIDIA’s graphic cards. Regardless of the problem origin, the screen needs to be changed and Nec-Mitsubishi’s technical help team come to your place to replace it. The screen new versions no longer have this problem.

With the new screen, we can pursue our tests. This time, the screen switches on. The standard picture setting is 17.3” wide. A couple of adjustments are required. It took us only one minute to get a nice picture (17.7”) without any obvious geometrical defaults. The OSD isn’t aesthetical but it is efficient. Now, the last parameter to adjust is the color.

So, for the installation process, the advantage goes to the BenQ screen. However it has to be said that the Mitsubishi screen is the first one that we met to have this Eeprom format problem. Usually the only parameter to adjust with CRT screens is the picture geometry.

Page 5
Brightness measure, color, contrast

Brightness, color, contrast
The contrast ratio and brightness are two data measured according to the ISO norm. Please read this article to understand how we disagree with this norm: absurd data, faulty results, etc. ISO is apparently working on a new version of this norm. We hope that it will be release soon!

BenQ FP767-12
To stand out on real or on-line demonstration areas, flat screen are usually displaying excessive brightness. For your comfort, flat screen brightness should never reach over 110 cd/m˛. Influenced by the ISO norm and the resellers, the manufacturers commonly announce twice more, 220 cd/m˛, based on non representative measurements. Sometimes screens even have brightness over 300 cd/m˛. Such parameters are unnecessary and dangerous. Pushing the neon brightness to their limits reduce their life and the life of your monitor. Your eyes tiredness is also connected to your screen brightness. The brighter it is the faster your eyes get tired. So even if with LCD screens we don’t have to recompose the picture provided by a CRT screen, being exposed to a permanent dazzling light strains our eyes.

6500 K color temperature test

Interpreting the graph
X axis: 0 black to 100 white
Only the red curve, Delta E is of interest. Delta E is a measure of the gap between the desired color and the one measured by our colorimeter.
If Delta > 3, the desired color is noticeably different from the color displayed.
As soon as the Delta E is inferior to 3, even graphic designers agree that results are accurate.
There is, however, still a slight difference between the requested color and the color displayed.
If Delta is between 1 and 2, most of the users won´t notice a difference between the desired and displayed color.
Eye sensitivity is when Delta E =1. Below this value, the result is perfect.

The BenQ screen has better results than the average LCD screen. Standard brightness is “only” 165 cd/m˛, this result is completely acceptable. With the sRGB mode, recommended for multiple use on a computer (games, office software, Web…), the FP767-12 has excellent results. 90% of the colors displayed are accurate, 96% are very good (DeltaE < 2), and 91% are perfect. This very is better than most of the LCD screens. Unlike common ideas on LCD screens, not expensive screens for picture editing exist.

5000 K color temperature test
If most of the users adjusts their screens to 6500K, graphic designers use the 5000K color temperature. This color temperature is clearly more difficult to display.

Once more, the FP767-12 has excellent results. Knowing that even the graphic designers rarely calibrate their screens, this screen with the right settings could be used for professional picture editing. However the result is still far from being perfect.

Mitsubishi Diamond Pro 930SB
6500 K color temperature test

Mitsubishi straight things up with this test. To get real colors, CRT screens are still the bests. The quality and accuracy are impressive. With this screen the black is really black. The best black reached is 0.07 cd/m˛ instead of 0.2 cd/m˛ for the LCD.

With sRGB parameters, used for this test, some settings like brightness and contrast are not accessible. We have to use standard parameters. And there is actually a lack of brightness. The 930SB white at 79 cd/m˛ is a little bit weak. A better result would be 90 cd/m˛.

5000 K color temperature test

The result is even more impressive at 5000K. Any colors from white to black are accurately displayed. The “top of the line” CRT reputation is verified. These screens are the only one to reach to the graphic designers requirements.

CRT screens are the bests for color quality.

Page 6
Color, practical tests

Color, practical tests
A measuring tolls provides results and it is quite useful. However, sometime our eye has a slightly different perception and notice stronger defaults than the colorimeter.

Picture distortion

Visually, even if the Mitsubishi screen is given as the more accurate by our measuring tool, the BenQ picture seems better, brighter and more vivid. The other problem with the Mitsubishi screen is that the manufacturer recommends using the monitor with the 1280 x 1024 pixels resolution. With this resolution the picture ratio is 1.25 instead of 1.33 for a 14,2 in x 10,6 in screen.

With the 1280 x 1024 resolution, pictures are slightly vertically squeezed. To avoid this problem it is better to use 4/3 resolutions like the 1280 x 960. This resolution is however not always available in games, so the best one to choose will be the 1024*768 or the 1280*1024. CRT screens can change their resolution without any picture quality loss.

With the LCD the resolution to use is 1280 x 1024 pixels. Changing the flat screen native resolution results in a lower quality picture interpolation and a distortion as the other resolutions have 1.33 ratios. The sharpness is reduced, and a powerful graphic card is required for this type of resolutions.


The two screens have analogical connections to the computer. However the sharpness differs from one screen to the other.

The BenQ FP767-12. CD icon.
The same icon on the CRT screen:

The better sharpness of LCD screens isn’t a legend. Even if the Mitsubishi is a high quality product, the middle range FP797-12, has better results for the sharpness. BenQ wins once more.

Colors defaults

With the 6500K color temperature, the Mitsubishi screen tests showed that it provides almost a perfect black. The BenQ requires a little bit more time to reach below DeltaE = 2. Here are the tests results in practice:

It is difficult to realise it, but he LCD grey scale is better with the CRT than with the LCD. It is possible to notice stripes of colors with the LCD. The CRT screen color scale is progressive. This problem isn’t really obvious, and it appears mainly on the color scales. However a photographer or a doctor displaying black and white X-rays on their screen might be annoyed.

Mitsubishi wins.

Black deepness

The colorimeter results were categorical: the CRT screen provides deeper black. The visual test slightly changes this result. The black is perfect if all lightning sources pointing to the CRT screen are switched off. Under standard circumstances, with fluorescent light, day light the better result is provided by the LCD screen. The LCD panel absorbs the light and the DPro930SB partly reflect it.

One point for BenQ.

Page 7
Vision angles

Vision angles

The BenQ screen is equipped with a TN + Film technology panel. These panel vision angles are the smaller. It is possible to notice this problem on the pictures especially when you look at your screen from below. Fortunately, this situation isn’t the most encountered.

The Mitsubishi screen is perfect from any angles. So the advantage goes to the CRT.

Page 8
Games reactivity

Games reactivity
The BenQ FP767-12 is one of the fastest LCD screen. The response time announced is 12 ms. We already explained in the "Focus on LCD article" that the response time is rarely representative of the real monitor quality. But for once, the response time announced is representative of the screen performances. Only the Hydis panels (displaying a theoretical 20 ms response time) are even faster (but the color quality and vision angles are worse).

The CRT screen

Trying to palaver is useless, the CRT clearly win this test. Even if the FP767-12 is one of the best LCD screen on the market, the DPro930SB is the only one perfectly reactive, fluid and without afterglow. The BenQ screen has blurred pictures and trails of light behind following objects when the picture is moving. When looking at this screen only, the afterglow isn’t obvious but with a LCD and a CRT screen in clone mode (with the same picture) the difference is obvious. However the BenQ screen can be used to play. The picture is only less pleasant.

The upper picture shows the afterglow resulting from a quick movement. Playing games like Doom or Unreal, include spending your time jumping around and quickly looking in your back. So in this type of situation when your mouse is constantly in motion on your desk the afterglow appears. The LCD screen can’t keep up with quick movements. The successive pictures are superimposed instead of disappearing progressively.

If you are one hard gamer, the CRT remains the best choice. We can’t wait to test the new panels to see if this result might be changed.

Page 9

It is now certain that CRT screens are going to disappear. And it is also certain that hard gamers and exigent graphic designers won’t probably like the new LCD screens.

Here are in a nutshell the advantages and disadvantages of the CRT and LCD screens.

Opposing the two screen technologies showed some very surprising results. These results were not always expected and were sometime contrary to the products reputations. For example, a CRT screen isn’t necessarily more expensive than a very good LCD screen. With time, the LCD requiring less power might even be the cheapest (see our measures in page2).
For the color quality the eye and the colorimeter provides two different results. The colorimeter results don’t include any external light sources. And according to this colorimeter the CRT provides an exceptional contrast ratio and an incredible black deepness. However this result is accurate only if you work with a cache surrounding your screen. Without The LCD screen is less affected by surrounding light.

Opposing the two screens also confirmed some of the established facts. It is not a surprise LCD screens are less bulky than CRT screens. They are also easier to install and have straight away nice pictures. The CRT screens require thorough settings. However CRT screens are better for games even if a quick panel like the 12 ms AU Optronics doesn’t shows any obvious afterglow.

Now this conclusion is of course very reductive because the main points are briefly summed up and that other screens may have produced slightly different results. Later, we will make more tests with different sizes and technologies. However these first tests will allow you to make your own mind, with some pictures about the CRT and LCD technology. Will the LCD technology transition be a good thing for you? Even in our office opinions diverts…

Copyright © 1997-2014 BeHardware. All rights reserved.