Right. I need to know what the difference is between quoted 'Contrast ratio' and 'Dynamic contrast ratio' ratings for TVs. Obviously manufacturers will put the highest figure in their spec sheets, but which one is the one to believe? I'm currently looking at TVs from Panasonic (TH-42PZ85B - 30,000:1 native and 1,000,000:1 dynamic) and Toshiba (40ZF355D - 30,000:1 dynamic). Now, looking at the two quoted figures, the Tosh seems to be woeful in comparison. Any help, please?
Contrast ratio question........
- Thread starter admin_exported
- Start date
The answer is believe none of them - typically over-inflated, meaningless nonsense.
There is meant to be an industry-wide measuring standard coming into existence, but until then (and even after then), read a lot of reviews rather than spec sheets, and more importantly still, trust your eyes and take along your own Blu-ray/DVD discs to test challenging scenes - we recommend several discs/scenes in any average issue.
There is meant to be an industry-wide measuring standard coming into existence, but until then (and even after then), read a lot of reviews rather than spec sheets, and more importantly still, trust your eyes and take along your own Blu-ray/DVD discs to test challenging scenes - we recommend several discs/scenes in any average issue.
it used to be static ratio.. like 100-1500 and some measured in dynamic... 5000+ but now its just madness...
The native contrast is usually quoted for plasma screens. This is the "actual" contrast it an deliver on the screen in real life.
"Dynamic" contrast is something LCDs have conjured up to make their panels look good. The panel cannot deliver this contrast ratio in any one scene. LCDs adjust the backlight for dark scenes to make them look darker, and in bright scenes the backlight goes up to make them brighter. The "dynamic" contrast is the the difference between these two. However, this "dynamic" contrast ratio cannot be achieved on the screen in one shot.
So the real measurement, is "native" contrast.
"Dynamic" contrast is something LCDs have conjured up to make their panels look good. The panel cannot deliver this contrast ratio in any one scene. LCDs adjust the backlight for dark scenes to make them look darker, and in bright scenes the backlight goes up to make them brighter. The "dynamic" contrast is the the difference between these two. However, this "dynamic" contrast ratio cannot be achieved on the screen in one shot.
So the real measurement, is "native" contrast.
some plasmas have theirs measured in dynamic too.. but they are just mad and so are native ratios these days, noone should choose a tv over their ratios.
Quoting static and dynamic contrast ratio is essentially plasma manufacturers trying to highlight that the figures quoted by LCD manufacturers are not the same. It's a shame it has gone this way, but a bit of understanding helps. In a gross oversimplification think of contrast ratio as the number of steps from black to white. If you have 1000:1 then there are 1000 different little shades from black, to slightly less black, to a little less black, and on and on to greyish white, whiter white, actual white (this isn't strictly how it works but I think it makes it a lot easier to understand). If contrast is only 500:1 then there are only half the amount of different levels achieveable on-screen at any one time.
Now static contrast ratio is the more commonsense figure. It is as simple as the explanation above, at any one time what can the display show all at once on the screen. Dynamic contrast ratio is a theoretical figure... The technogubbins in the screen "dynamically" changes the brightness and gamma of the display throughout viewing. So if the processor reckons the image is predominantly dark, it lowers the light output, pushes a deeper gamma curve, and gives the viewer lots of black and black detail at the expensive of less punchy whites. Vice versa is also true. A high brightness shot (e.g. clouds, ice, lots of daylight) will involve lots of light output, really nice whites, but a raised black level (slightly more greyish blacks) and black crush (less seperation between different shades of black). The dynamic figure is essentially the manufacturer saying, well when the screen is dark we get x, when it's white we get y, lets make this overall range the achieveable figure. However going back to a single, static picture on-screen, when the image contains an equal portion of white and black (or dark and light) the dynamic range of the image is significantly lower. There is also the question of actually being able to see the shift taking place throughout viewing...
Unfortunately the subject doesn't end there. Firstly the figures are almost totally made up (funny that it should work out to exactly 20,000:1 or 35,000:1!). Secondly there are ways of taking the measurements which can give bloated figures, but would never represent what the display would achieve in a normal setup in normal day to day use. But that rant is for another thread!
Now static contrast ratio is the more commonsense figure. It is as simple as the explanation above, at any one time what can the display show all at once on the screen. Dynamic contrast ratio is a theoretical figure... The technogubbins in the screen "dynamically" changes the brightness and gamma of the display throughout viewing. So if the processor reckons the image is predominantly dark, it lowers the light output, pushes a deeper gamma curve, and gives the viewer lots of black and black detail at the expensive of less punchy whites. Vice versa is also true. A high brightness shot (e.g. clouds, ice, lots of daylight) will involve lots of light output, really nice whites, but a raised black level (slightly more greyish blacks) and black crush (less seperation between different shades of black). The dynamic figure is essentially the manufacturer saying, well when the screen is dark we get x, when it's white we get y, lets make this overall range the achieveable figure. However going back to a single, static picture on-screen, when the image contains an equal portion of white and black (or dark and light) the dynamic range of the image is significantly lower. There is also the question of actually being able to see the shift taking place throughout viewing...
Unfortunately the subject doesn't end there. Firstly the figures are almost totally made up (funny that it should work out to exactly 20,000:1 or 35,000:1!). Secondly there are ways of taking the measurements which can give bloated figures, but would never represent what the display would achieve in a normal setup in normal day to day use. But that rant is for another thread!
so in terms of this contrast ratio (1000:1) etc a projector seems very weak compared to plasmas etc does this make a difference in picture quality ?
Clare Newsome:
The answer is believe none of them - typically over-inflated, meaningless nonsense.
There is meant to be an industry-wide measuring standard coming into existence, but until then (and even after then), read a lot of reviews rather than spec sheets, and more importantly still, trust your eyes and take along your own Blu-ray/DVD discs to test challenging scenes - we recommend several discs/scenes in any average issue.
You look good clare.....[the photo] tantalising....just like the News presenters. They should make you editor in chief.If there was more women in the HI Fi scene. All the people here are blokes.
Its about time we design TVs with lifestyle aesthetics in mind.
The answer is believe none of them - typically over-inflated, meaningless nonsense.
There is meant to be an industry-wide measuring standard coming into existence, but until then (and even after then), read a lot of reviews rather than spec sheets, and more importantly still, trust your eyes and take along your own Blu-ray/DVD discs to test challenging scenes - we recommend several discs/scenes in any average issue.
You look good clare.....[the photo] tantalising....just like the News presenters. They should make you editor in chief.If there was more women in the HI Fi scene. All the people here are blokes.
Its about time we design TVs with lifestyle aesthetics in mind.
888:Clare Newsome:
The answer is believe none of them - typically over-inflated, meaningless nonsense.
There is meant to be an industry-wide measuring standard coming into existence, but until then (and even after then), read a lot of reviews rather than spec sheets, and more importantly still, trust your eyes and take along your own Blu-ray/DVD discs to test challenging scenes - we recommend several discs/scenes in any average issue.
You look good clare.....[the photo] tantalising....just like the News presenters. They should make you editor in chief.If there was more women in the HI Fi scene. All the people here are blokes.
Its about time we design TVs with lifestyle aesthetics in mind.
(Sits back to await the reply..........)
The answer is believe none of them - typically over-inflated, meaningless nonsense.
There is meant to be an industry-wide measuring standard coming into existence, but until then (and even after then), read a lot of reviews rather than spec sheets, and more importantly still, trust your eyes and take along your own Blu-ray/DVD discs to test challenging scenes - we recommend several discs/scenes in any average issue.
You look good clare.....[the photo] tantalising....just like the News presenters. They should make you editor in chief.If there was more women in the HI Fi scene. All the people here are blokes.
Its about time we design TVs with lifestyle aesthetics in mind.
(Sits back to await the reply..........)
The contrast figures given by manufactures are generally meaningless except in comparing models by the same manufacturer. Some measure them, some calculate them, they are not measured with the display colour calibrated for viewing video, they are for full black and white screens not mixed images so give no idea what simultaneous contrast in a video image will be.
The contrast quoted by display manufactures is usually full on / full off, that is the luminance of a full white screen devided by the luminance of a full black screeen. Maximum native sequencial contrast and possibly close to the maximum native simultaneous contrast.
Dyanmic Contrast implies the use of a back light and dynamic gamma control, this improves contrast. If the highest luminance part of the image is say only 50% of the maximum luminance of the display the native contrast would be 50% of the displays maximum contrast. But with dynamic contrast it can double that to the full native contrast by dropping the back light by 50% and replotting the greyscale so the luminance of the bright part of the image does not drop by 50% but the luminance of the darkest parts does. So dynamic contrast increases the number of scenes the display can use it full native contrast, but the most contrast any individual image can have is the native contrast, it can not use dynamic contrast to improve the contrast of a image containing a luminance equal to the brightest the display can go as it is unable to lower the black light. If it has multiple backlights controling different parts of the screen it gets more complex. Dynamic contrast will be calculated off backlight on highest dynamic level - brightest scene full white devided by backlight at lowest dynamic level - darkest scene full black. So will usually be higher than the contrast possible in a single image, it is a maximum sequencial contrast ratting.
However in practice images are not full white or full black they contain lots of different luminance levels. ANSI contrast measurement used to be made, they used a checkerboard of white and black squares and devided the average of the middle of the white squares by the average of the middle of the black squares. This was a worst case situation and generally taken to be a guide as to how much wash out effect the display suffered, black parts of the screen becoming less black due to the presence of white parts of the screen. It was unrepresentative of actual video performance being lower, since it gave low figures and manufactures started to vary in how they measured it, it has been abandoned by manufactures.
Unfortunately the contrast measurements are not taken with the display calibrated to D65 grey, so maybe higher than is possible with correct greyscale - colour. They are also taken in a completely black room or are calculated rather than actually measured. A tiny difference in the measurement of black level can make a huge difference to the calculated contrast ratio, and with the figures quoted by manufactures being so high you wander what they use to measure the black level.
You do not view the display in a pitch black, black painted unrefelctive room. Maybe fortunately given that the luminance of most flatpannels would give you eyestrain. But the typical living room will cause a major reduction in visible contrast as ambient lighting hitting the screen will lower its contrast.
Actual perception of contrast is determined by Gamma, image luminance and surround effect. Higher gamma will be perceived as more contrast in bright images, lower contrast in dark images. Lower gamma will increase contrast in darkscenes and decrease it in bright scenes. Video is mastered to 2.2 gamma with robustness checked for upto 2.5 gamma. Visible perception of contrast is decreased by the dark surround effect, so a display in a pitch black room needs a higher gamma to compensate. Perception of contrast is increased by display luminance so a brighter display will look as if it has higher contrast even if it has the same contrast, at least in bright scenes.
Liam @ Prog AV: It's a shame it has gone this way, but a bit of understanding helps. In a gross oversimplification think of contrast ratio as the number of steps from black to white. If you have 1000:1 then there are 1000 different little shades from black, to slightly less black, to a little less black, and on and on to greyish white, whiter white, actual white (this isn't strictly how it works but I think it makes it a lot easier to understand). If contrast is only 500:1 then there are only half the amount of different levels achieveable on-screen at any one time.
Digital video Blu-ray, DVD, and Digital TV Standard and High Definition, all use 8-Bit greyscale. There are only 256 levels, for film/video reference black is 16, reference white is 235. A display should not display below black <16 when calibrated, but the ability to show below black is useful when actually calibrating. Above reference white >235 should be displayed when calibrated to allow for over shoots and spectacular highlights that are permitted with video encoding, if does not display above white the image will still look ok. These digital levels are not equal distance steps of luminance they are logarithmic, plotting them gives a gamma curve usually 2.2-2.5, the gamma number is a power function. On a video greyscale 50% grey is actually about 20% luminance. This is because human vision is also logarithmic we perceive a 50% drop in luminance as a 20% drop in brightness.
The contrast quoted by display manufactures is usually full on / full off, that is the luminance of a full white screen devided by the luminance of a full black screeen. Maximum native sequencial contrast and possibly close to the maximum native simultaneous contrast.
Dyanmic Contrast implies the use of a back light and dynamic gamma control, this improves contrast. If the highest luminance part of the image is say only 50% of the maximum luminance of the display the native contrast would be 50% of the displays maximum contrast. But with dynamic contrast it can double that to the full native contrast by dropping the back light by 50% and replotting the greyscale so the luminance of the bright part of the image does not drop by 50% but the luminance of the darkest parts does. So dynamic contrast increases the number of scenes the display can use it full native contrast, but the most contrast any individual image can have is the native contrast, it can not use dynamic contrast to improve the contrast of a image containing a luminance equal to the brightest the display can go as it is unable to lower the black light. If it has multiple backlights controling different parts of the screen it gets more complex. Dynamic contrast will be calculated off backlight on highest dynamic level - brightest scene full white devided by backlight at lowest dynamic level - darkest scene full black. So will usually be higher than the contrast possible in a single image, it is a maximum sequencial contrast ratting.
However in practice images are not full white or full black they contain lots of different luminance levels. ANSI contrast measurement used to be made, they used a checkerboard of white and black squares and devided the average of the middle of the white squares by the average of the middle of the black squares. This was a worst case situation and generally taken to be a guide as to how much wash out effect the display suffered, black parts of the screen becoming less black due to the presence of white parts of the screen. It was unrepresentative of actual video performance being lower, since it gave low figures and manufactures started to vary in how they measured it, it has been abandoned by manufactures.
Unfortunately the contrast measurements are not taken with the display calibrated to D65 grey, so maybe higher than is possible with correct greyscale - colour. They are also taken in a completely black room or are calculated rather than actually measured. A tiny difference in the measurement of black level can make a huge difference to the calculated contrast ratio, and with the figures quoted by manufactures being so high you wander what they use to measure the black level.
You do not view the display in a pitch black, black painted unrefelctive room. Maybe fortunately given that the luminance of most flatpannels would give you eyestrain. But the typical living room will cause a major reduction in visible contrast as ambient lighting hitting the screen will lower its contrast.
Actual perception of contrast is determined by Gamma, image luminance and surround effect. Higher gamma will be perceived as more contrast in bright images, lower contrast in dark images. Lower gamma will increase contrast in darkscenes and decrease it in bright scenes. Video is mastered to 2.2 gamma with robustness checked for upto 2.5 gamma. Visible perception of contrast is decreased by the dark surround effect, so a display in a pitch black room needs a higher gamma to compensate. Perception of contrast is increased by display luminance so a brighter display will look as if it has higher contrast even if it has the same contrast, at least in bright scenes.
Liam @ Prog AV: It's a shame it has gone this way, but a bit of understanding helps. In a gross oversimplification think of contrast ratio as the number of steps from black to white. If you have 1000:1 then there are 1000 different little shades from black, to slightly less black, to a little less black, and on and on to greyish white, whiter white, actual white (this isn't strictly how it works but I think it makes it a lot easier to understand). If contrast is only 500:1 then there are only half the amount of different levels achieveable on-screen at any one time.
Digital video Blu-ray, DVD, and Digital TV Standard and High Definition, all use 8-Bit greyscale. There are only 256 levels, for film/video reference black is 16, reference white is 235. A display should not display below black <16 when calibrated, but the ability to show below black is useful when actually calibrating. Above reference white >235 should be displayed when calibrated to allow for over shoots and spectacular highlights that are permitted with video encoding, if does not display above white the image will still look ok. These digital levels are not equal distance steps of luminance they are logarithmic, plotting them gives a gamma curve usually 2.2-2.5, the gamma number is a power function. On a video greyscale 50% grey is actually about 20% luminance. This is because human vision is also logarithmic we perceive a 50% drop in luminance as a 20% drop in brightness.
aliEnRIK:And how many people watch their tvs with gamma settings way out...........
There is alot of leeway 2.2-2.5 built into the mastering, because they do not know what your viewing environment will be. Darker environment needing higher gamma to compensate for dark surround effect. Since displays track gamma curves the effect of higher contrast is only on the darkest parts of the image, To display accurately gamma of 2.2 from 5%white to 100%white only requires 728:1 contrast, to display down to 1%white needs 25,119:1, and down to black needs infinite contrast. But the bottom 5% of the greyscale is critical in giving the image solidity and creating the illusion of image depth.
As far as how many people use the wrong gamma settings it depends on the display type and viewing conditions. With CRT you generally had no control over gamma. Depending who you go by CRT had a average gamma of 2.2 or 2.35 or 2.5. Flat pannels often have a gamma of 2.2 which is fine in a living room, while a projector in a black home cinema looks better with higher gamma. The choice of gamma setting comes down to higher gamma giving more image depth - punch in bright images and more colour saturation, but a overall less bright image. Lower gamma giving a brighter image with more visible shadow detail. If you display image looks natural and consistantly gives the illusion of image depth, looks like you could throw a ball through the screen or shout at the people on it, then gamma is correct.
Display image quality can be measured by manufactures but they do not bother. I asume this is because it would be time consuming, and lack of consumer comprehension of what on earth the graphs indicated might render it a pointless exercise. So they just use large numbers to impress, and simple ideas like large contrast good, low contrast bad, which do not give a true indication of picture quality.
MTF modulation transfer function curves, which are calculated off the contrast of line pairs per degree of the viewers visual arc, so you need to know display size and viewing distance, clearly indicate the displays ability to produce a sharp image with good clarity. It can also be weighted for a overall image quality because the visual importance of the center, middle, corners of the display are different. Plotting from large details to small details, wide black and white lines to thin black and white lines. MTF at the start to middle of the curve indicates contrast in large to mid details, how much objects standout clearly from the background, how sharp the image looks. MTF at the middle to end of the curve indicates contrast in the middle to fine details how much texture of objects is visible, like skin tones to clothing texture, how detailed the image looks. MTF at the end of the curve indicates absolute display resolution, the least critical indicator of apparant sharpness. The overall perception of sharpness is determined by MTF from 3 to 12 line pairs per degree of viewers visual arc. For camera image reproduction MTF of 50% is the point taken to be out of focus, MTF 35% is the point that a area of the image is no longer in the zone of confusion - the part of the image that looks in focus due to depth perception, MTF 5% is taken as the resolution limit of the display, although a trained observer can still distinguish objects down to MTF 0.05%.
Gamma tracking, what curve is tracked 2.2-2.35-2.5 and how smoothly and accurately that curve is tracked particularly at the bottom dark end of the grey scale where the display can not faithfully follow the curve. Limited by MTF, indicates the displays contrast and apparant image depth in bright to dark scenes. Poor gamma tracking leads to images that look good in some scenes and poor in others due to altering perceptions of image depth and scenes that look less natural due to exageration or under emphasis in the contrast of various elements of the picture. Gamma curves also indicate a displays suitability to different viewing environments.
Greyscale color accuracy across the entire greyscale black through shades of grey to white, primary colour coordinates, and colour decoder accuracy. All combine to indicate if the image has correct, natural looking colours. Because the eye-brain perceives colors as ratios, inaccurate colours look more natural with the colour saturation/luminance control lowered, it reduces the size of the error, but also robs the image of colorfulness and some contrast - image depth. The eye brain is most sensitive to color errors in skin tones, accurate colour makes the people on screen look incredibly lifelike as it they are actually standing in front of you, and makes the actresses look more beautiful.
Then you have if the display is capable of displaying the 8-bit greyscale on an individual pixel bases, or relys on temporal or spatial dithering to create the shade. If the display can show changes in 8-bit greyscale pixel on a frame to frame bases or if the changes from one shade to another have lag time. How often and how long the pixels are illuminated per frame.
There is alot of leeway 2.2-2.5 built into the mastering, because they do not know what your viewing environment will be. Darker environment needing higher gamma to compensate for dark surround effect. Since displays track gamma curves the effect of higher contrast is only on the darkest parts of the image, To display accurately gamma of 2.2 from 5%white to 100%white only requires 728:1 contrast, to display down to 1%white needs 25,119:1, and down to black needs infinite contrast. But the bottom 5% of the greyscale is critical in giving the image solidity and creating the illusion of image depth.
As far as how many people use the wrong gamma settings it depends on the display type and viewing conditions. With CRT you generally had no control over gamma. Depending who you go by CRT had a average gamma of 2.2 or 2.35 or 2.5. Flat pannels often have a gamma of 2.2 which is fine in a living room, while a projector in a black home cinema looks better with higher gamma. The choice of gamma setting comes down to higher gamma giving more image depth - punch in bright images and more colour saturation, but a overall less bright image. Lower gamma giving a brighter image with more visible shadow detail. If you display image looks natural and consistantly gives the illusion of image depth, looks like you could throw a ball through the screen or shout at the people on it, then gamma is correct.
Display image quality can be measured by manufactures but they do not bother. I asume this is because it would be time consuming, and lack of consumer comprehension of what on earth the graphs indicated might render it a pointless exercise. So they just use large numbers to impress, and simple ideas like large contrast good, low contrast bad, which do not give a true indication of picture quality.
MTF modulation transfer function curves, which are calculated off the contrast of line pairs per degree of the viewers visual arc, so you need to know display size and viewing distance, clearly indicate the displays ability to produce a sharp image with good clarity. It can also be weighted for a overall image quality because the visual importance of the center, middle, corners of the display are different. Plotting from large details to small details, wide black and white lines to thin black and white lines. MTF at the start to middle of the curve indicates contrast in large to mid details, how much objects standout clearly from the background, how sharp the image looks. MTF at the middle to end of the curve indicates contrast in the middle to fine details how much texture of objects is visible, like skin tones to clothing texture, how detailed the image looks. MTF at the end of the curve indicates absolute display resolution, the least critical indicator of apparant sharpness. The overall perception of sharpness is determined by MTF from 3 to 12 line pairs per degree of viewers visual arc. For camera image reproduction MTF of 50% is the point taken to be out of focus, MTF 35% is the point that a area of the image is no longer in the zone of confusion - the part of the image that looks in focus due to depth perception, MTF 5% is taken as the resolution limit of the display, although a trained observer can still distinguish objects down to MTF 0.05%.
Gamma tracking, what curve is tracked 2.2-2.35-2.5 and how smoothly and accurately that curve is tracked particularly at the bottom dark end of the grey scale where the display can not faithfully follow the curve. Limited by MTF, indicates the displays contrast and apparant image depth in bright to dark scenes. Poor gamma tracking leads to images that look good in some scenes and poor in others due to altering perceptions of image depth and scenes that look less natural due to exageration or under emphasis in the contrast of various elements of the picture. Gamma curves also indicate a displays suitability to different viewing environments.
Greyscale color accuracy across the entire greyscale black through shades of grey to white, primary colour coordinates, and colour decoder accuracy. All combine to indicate if the image has correct, natural looking colours. Because the eye-brain perceives colors as ratios, inaccurate colours look more natural with the colour saturation/luminance control lowered, it reduces the size of the error, but also robs the image of colorfulness and some contrast - image depth. The eye brain is most sensitive to color errors in skin tones, accurate colour makes the people on screen look incredibly lifelike as it they are actually standing in front of you, and makes the actresses look more beautiful.
Then you have if the display is capable of displaying the 8-bit greyscale on an individual pixel bases, or relys on temporal or spatial dithering to create the shade. If the display can show changes in 8-bit greyscale pixel on a frame to frame bases or if the changes from one shade to another have lag time. How often and how long the pixels are illuminated per frame.
Gamma can also be controlled from the source (If it has gamma control), my LX50 dvd player has it (Which is just as well as its gamma in default is WAY out)
such a long explanation but i just wanted to know wether a projector was weaker in contrast than a tv
Similar threads
TRENDING THREADS
-
-
-
We've built a gold-standard hi-fi system that delivers sweet streams aplenty – and looks good, too- Started by What Hi-Fi?
- Replies: 1
-
-
I test OLED TVs for a living – and trust me, different sizes perform differently
- Started by What Hi-Fi?
- Replies: 2
-
Samsung Galaxy S25: launch date rumours, expected specification and more
- Started by What Hi-Fi?
- Replies: 0
-
Facebook
Twitter
Instagram