Talk:Contrast ratio
This is the talk page for discussing improvements to the Contrast ratio article. This is not a forum for general discussion of the article's subject. |
Article policies
|
Find sources: Google (books · news · scholar · free images · WP refs) · FENS · JSTOR · TWL |
![]() | This article is rated Start-class on Wikipedia's content assessment scale. It is of interest to the following WikiProjects: | |||||||||||||
|
2008
[edit]contrast ratio is also a cinematography term that is not (yet) covered well within the article .
"Many display devices favor the use of the full on/full off method of measurement, as it cancels out the effect of the room and results in an ideal ratio. Equal proportions of light reflect from the display to the room and back in both "black" and "white" measurements, as long as the room stays the same. This will inflate the light levels of both measurements proportionally, leaving the black/white luminance ratio unaffected."
This seems to be magic ... a constant delta in enumerator and denominator of the ratio will surely produce a reduced quotient. Also it is probably not the desire of the display device which method of measurement it favours. This statement is just wrong.
- I'll explain what I meant. Suppose you're doing a measurement in a completely non-reflective room and you're looking at a 1000 lumen screen at 100% white. The projector produces, say, 10 lumens for that screen at 100% black input signal. You'll get a 100:1 measured contrast ratio for that setting. Now if the room were to reflect some of the light back, it would do it with equal ratios for both the black and the white signals - for example, 10%. This would produce 1100 lumens for white signal and 11 lumens for the black signal, and still a 100:1 measured contrast ratio. The reflectivity of the room would then be cancelled out. Do you have another wording in mind that would demonstrate this concept in a more understandable fashion? Santtus (talk) 22:38, 19 May 2008 (UTC)
is this the ANSI contrast? Actually, the darkest color is black, and that means zero luminosity (should be not hard to achive)... the contrast then would have an infinite contrast ratio. thus i guess the contrast ratio isn't that good in telling you about the quality of a display - or am i wrong? thanks, --Abdull 13:34, 26 Apr 2005 (UTC)
It's quite difficult to achieve zero luminosity using any display technology. For example, in DLP projectors some light will reflect from between the mirrors, and some will reflect from the very mirrors themselves. Reflection from aluminum is not completely mirror-like (specular reflection), there will be some diffuse reflection as well, meaning that mirrors that are turned away in a [[DMD] device and that should look totally black, will scatter a little light into the screen - maybe 0,05% of the amount if the mirrors would be turned the other way (figures not checked). Santtus 22:05, 24 July 2005 (UTC)
Actually CRTs achieve a theoretically infinite contrast ratio and practically achieve enough of one to make such quantification pointless. (comment 23:53, 8 January 2006 Zbobet2012)
- However, in practice CRT's can't achieve infinite contrast ratios. Four problems: The reflectivity of the surface is not zero, the CRT response is not linear to the ends, and drifts (so you may not be able to uze the dark end of the scale), the circuit response is not linear to the ends, and drifts (so you may not be able to uze the dark end of the scale), and the noise of the input video signal is not zero, so in practice all CRT's have adjustable contrast and brightness, and a noticable contrast ratio variation between units, when you line them up on the bench. —Preceding unsigned comment added by 218.214.18.240 (talk) 00:56, 8 June 2009 (UTC)
- Checked that fact, you're right. I found something lot more interesting tho. I'll re-arrange the article later on.. lots of stuff that needs to be added. I'll work on that. Added some references already.
- The static contrast ratio of the eye is just about 0,5-2%, so that a display with 200:1 contrast ratio would suffice, if the average light level in the image would stay the same. The overall light level in the image usually changes, so greater ratios are required. The eye adjusts to various light levels to the extent that about 10^7:1 ratios can be achieved, but mass produced display devices are incapable of such contrast ratios. This is explained to better depth in here [1].
- A presence of single candle in an average room will render 30000:1 (or infinite, for that matter) and 500:1 displays indistinguishable from each other. In a typical brightly lit office only contrast ratios of 10:1.. 30:1 can be achieved. In this situation, the determining factors are the ambient light level, reflectivity of the display surface and the maximum luminocity the device is capable of. The slight variations in the lowest blackest levels would be totally lost in the presence of ambient light 10 or 100 times more than that of the device. Santtus 11:58, 11 January 2006 (UTC)
Removed the unsourced template
[edit]I added sources to perhaps one half of all the statements that could use one. I'll try adding some more references soon. Santtus 12:06, 11 January 2006 (UTC) Oops, sorry I'm late from addding references. Btw one thing.. I guess that someone mixed having ability to resolve brightness levels 1% apart (1:100) and having 100:1 resolution. Santtus 21:15, 5 March 2006 (UTC)
New reference format
[edit]I've updated the first couple notes to the new format, which can match up the numbers more accurately. Shawnc 23:19, 20 April 2006 (UTC)
This sentence
[edit]"This may not sound too bad though, as the static contrast ratio of a human eye is somewhere between 100:1 - 1000:1 so the details in those highlights might not be resolvable anyway." This can probably use a citation and be re-worded a bit, in particular "sound too bad" and "might not be resolvable". Shawnc 00:06, 21 April 2006 (UTC)
- That 100:1 - 1000:1 thing was from an article, although I summed up the controversy surrounding the issue. I'll try to look at it soon enough. Santtus 15:12, 26 April 2006 (UTC)
- You may want to edit the article on eye as well, because it states only the 100:1 figure. GreenGrass 14:58, 7 May 2006 (UTC)
- Was in the middle of editing, but I couldn't finish my writing before getting a headache ;) I'll finish it later. I checked from various sources that 100:1 seems to be fine - it's the most often quoted value. Found some info on contrast sensitivity too, but it measures minimum noticeable contrast, not maximum. Now I just got one more question. How do you define maximum contrast a person can see? I'll try to find an answer for that.. Santtus 13:18, 9 May 2006 (UTC)
- Well, as the performance of an eye is apparently limited by the performance of its lens, it should be easy to measure its static contrast ratio. You can fill 50% of your FoV with a white object, and 50% with a black object (with sufficiently high ratio between their diffuse reflectivity). Then you place a small text on the black object with a brightness just a bit more than the background. Afterwards, you gradually increase the brightness of the text until it becomes resolvable (in layman's words: "readable"). Than you measure the contrast ratio between the luminosity of the text and the luminosity of the white object, multiply it by two, and there you have it. (I took the last number from my head, it should be the minimum contrast necessary for the eye to detect edges of letters, which is probably a bit subjective issue). GreenGrass 19:55, 15 May 2006 (UTC)
- Was in the middle of editing, but I couldn't finish my writing before getting a headache ;) I'll finish it later. I checked from various sources that 100:1 seems to be fine - it's the most often quoted value. Found some info on contrast sensitivity too, but it measures minimum noticeable contrast, not maximum. Now I just got one more question. How do you define maximum contrast a person can see? I'll try to find an answer for that.. Santtus 13:18, 9 May 2006 (UTC)
- You may want to edit the article on eye as well, because it states only the 100:1 figure. GreenGrass 14:58, 7 May 2006 (UTC)
Good article
[edit]Good job on the article. I would start working on correcting grammar & etting some pictures.
Extraordinary contrast ratio
[edit]It appears that the phrase "extraordinary contrast ratio" doesn't mean anything special, except that a someone feels that the contrast ratio is great. I've found no examples of professional use of that phrase, so I'll revert the changes unless I'm otherwise convinced. Santtus 09:12, 13 June 2006 (UTC)
Local contrast
[edit]The article should mention the difference between local and "global" contrast (both spatial). Local contrast means contrast near differences between luminances. For example, if there is a white box on a black background, local contrast is the contrast close to the border between the white and the black areas. Global contrast is a bit like dynamic contrast, except both bright and dark areas may exist simultaenously - contrast is measured where the differing luminances don't "contaminate" each other. CRT monitors may have infinite global contrast because of the ability to turn off the electron beam completely, but they have worse local contrast than LCD displays, because due to internal reflections bright areas have halos which contaminate the dark areas. ANSI contrast measures local contrast unless the checkerboard pattern is very sparse. Global contrast should still be differentiated from dynamic contrast, because for example even if you adjust the backlight of an LCD monitor to achieve a high dynamic contrast, it won't help you if you have bright and dark areas at the same time. The great thing (for CRTs!) is that out eyes have low local contrast and high global contrast. This is also exploited for LCD monitors that are backlit by LEDs. Since there are fewer LEDs than pixels, the local contrast is low, but the global contrast can be great. A local HiFi magazine measured each three contrasts for the televisions they reviewed. This online article also highlights the differences between local and global contrast (in the article they're called contrast and dynamic range respectively): http://www.displaymate.com/ShootOut_Part_1.htm As you can see, the CRT has a global contrast of ~18,000:1 compared to the LCD's ~600:1! But a 9x9 checkerboard pattern demolishes CRT's local contrast, reducing it to ~100:1, while the LCD retains ~600:1 —The preceding unsigned comment was added by 84.231.81.152 (talk • contribs) 23:46, 3 December 2006 (UTC).
Methods of measurement
[edit]This paragraph:
Many manufacturers of display devices favor the use of the full on/full off method of measurement, as it will effectively cancel the effect of the room completely, giving as high ratios as possible. Equal proportion of light will reflect from the display to the room and back in both measurements, as long as the room stays the same. This will inflate the light levels of both the "black" and the "white" measurements in the same proportion, unaffecting the black/white luminance ratio.
is simply not true! If Lw, Lb and Lr are the white, black and room luminosity, respectively, then Lw/Lb might be very different from (Lw+Lr)/(Lb+Lr). The reason manufacturer prefer this contrast measurement is that it might give them higher contrast, than say the ANSI contrast. See for example the CRT (global) contrast vs. LCD (global) contrast in the previous comment.
I would suggest the following phrasing:
Many manufacturers of display devices favor the use of the full on/full off method of measurement, as it will give higher contrast ration over the ANSI contrast measurement method.
previous comment by user:yuvalaviel
I wrote that paragraph with assuming otherwise unilluminated room, with Lr = 0. Now that could be mentioned.
Actually it has not been established what is the most usual level of illumination in the testing room. If there is absolutely no other light than that what reflects from the screen, then we can establish that a fraction f (0<f<1) of the light reflects from the screen to the room and back. Then Lw/Lr = (Lw * (1+f)) / (Lr * (1+f)).
Judging from overly inflated contrast ratio figures generally presented in the marketing literature, it doesn't seem believable that the manufacturers would have generally performed their tests in any significantly illuminated environments.
Perhaps someone can shed light on this situation? 62.220.237.65 23:04, 29 March 2007 (UTC)
Contrast Ratio as marketing spec
[edit]I'd like to see some mention of contrast ratio as a marketing term for LCD computer monitors. There seems to be quite a bit of emphasis on contrast ratio. Is it really significant and measured the same across manufacturers? --24.249.108.133 22:19, 13 April 2007 (UTC)
Agreed, I'd like to see this as well. The external link http://www.practical-home-theater-guide.com/contrast-ratio.html seems like a good source to cite for how the ratios are inflated and how salespeople can misrepresent their meaning.
With the development of LED backlit LCD displays, LCDs will have the potential to get much better black levels than previously possible (because the LEDs that are responsible for backlighting a given section of the screen can be dimmed or turned off completely when that section needs to be darker than is possible with the LEDs at full brightness). I know this is possible intuitively, but I am no LCD engineer and I don't know what this technology is/will be called. It's dynamic something, but dynamic contrast ratio is already taken. Any ideas if or how this concept should be added? --71.228.196.178 01:04, 15 November 2007 (UTC)
- it's called "local dimming". the first hit on a google search is http://hometheatermag.com/gearworks/108gear2/ Kapland (talk) 11:12, 26 November 2009 (UTC)
The bit that is there at the moment seems very similar to http://ap.viewsonic.com/me/dcr/ 124.171.10.207 (talk) 08:10, 30 August 2008 (UTC)
Observation form CRT monitors
[edit]hmm, I have let the mention of "infinite contrast CRTs" to stay, but I'd wish to find a source to mention how it actually is. The monitors I've owned have shown visibly grey "black" level. For a reason or another, the contrast ratio was not being advertised in the time when LCD monitors where unavailable. When did they start measuring and advertising contrast ratios anyway? (forgot to sign) Santtus 22:02, 5 September 2007 (UTC)
- The contrast ratio is nearly the same on all CRTs and it also depends on the settings, if you make the picture dark enough (but not toot dark), you can get very high contrast ratios. It also depends on room lighting, it is best for teh contrast if you use the monitor in a complete dark room. --MrBurns (talk) 13:45, 10 December 2008 (UTC)
Blown out?
[edit]"The trick for the display is to determine how much of the highlights may be unnoticeably blown out in a given image under the given ambient lighting conditions". Could someone please clarify this sentence? I don't understand what is meant by "blown out" in the article's context. Is there a less technical term for it, perhaps? —Preceding unsigned comment added by 98.221.120.172 (talk) 20:42, 18 October 2008 (UTC)
- I'm pretty sure the author is referring to "blown out highlights", also called clipping (photography). --68.0.124.33 (talk) 05:08, 14 December 2008 (UTC)
Comparison
[edit]Could I ask someone for a better comparison with regard to what is a "good"/"desired" ratio, as to what is a "low-end"? For example; a link to a table of products showing their current ratios? -- Mjquin_id (talk) 19:37, 21 August 2009 (UTC)
Different Ways of Measuring Contrast Ratio
[edit]I'm here as a result of some confusion while comparing 2 monitors at NewEgg.com. One monitor has a Contrast Ratio of "2000:1 ASCR" while the other has "DC 2000:1(1000:1)"
What does this mean, and how can the wiki be edited to explain the difference ? —Preceding unsigned comment added by Jonny Quick (talk • contribs) 06:35, 4 September 2010 (UTC)
Copyvio
[edit]I added the copyvio template thinking it would just be a simple tag, but I see that I opened up a can of worms and right now and too tired to really go through the process. I noticed quite a bunch of text was copied word for word from a website that copyvio notwithstanding is not very reliable. B137 (talk) 04:02, 21 April 2015 (UTC)
poor status of article
[edit]1 "contrast ratio" mainly is a wide-spread pleonasm.
2 The International Lighting Vocabulary CIE S017/E:2011 states:
contrast 1. in the perceptual sense: assessment of the difference in appearance of 2 or more parts of a field seen simultaneously or successively (hence: brightness contrast, lightness contrast, colour contrast, simultaneous contrast, successive contrast, etc.)
2. in the physical sense: quantity intended to correlate with the perceived brightness contrast, usually defined by one of a number of formulae which involve the luminances of the stimuli considered: for example by the proportional variation in contrast near the luminance threshold, or by the ratio of luminances for much higher luminances
3 Contrast is often specified by the ratio of luminance levels, so this ratio should not be called "contrast ratio", because it rather is a luminance ratio.
4 There is no such thing as "luminosity" in display engineering and visual perception.
5 Contrast between two optical states is not limited to white (100% RGB) and black (0%). Why should this "cancel out the effect of the room and result in an ideal ratio"?
6 "Brightness, as it is most often used in marketing literature, refers to the emitted luminous intensity on screen, measured in candela per square metre (cd/m2). The higher the number, the brighter the screen."
Marketing publications are not the right guide to solid Wikipedia articles. "Brightness" is not the emitted "luminous intensity", "luminous intensity" is not quantified in candela per square metre (cd/m2),
7 The concerns listed above are not the only ones that make this article at least doubtful. Complete revision is recommended in order to avoid continued confusion of the readers.
panjasan (talk) 06:58, 11 August 2015 (UTC)
Start a discussion about improving the Contrast ratio page
Talk pages are where people discuss how to make content on Wikipedia the best that it can be. You can use this page to start a discussion with others about how to improve the "Contrast ratio" page.