drebel
2nd December 2002, 17:06
all
I've been using all of our latest methods for the black blocks issue with good results(lumoff,lumafilter,blockbuster...).Forgive my ignorance,but besides mpeg4 that prob seems to exist even in mpeg2(for another reason?).I'm reffering to the famous IRE prob.I'm copying a brief expanation from a site:
>>>In the original black and white television system the voltage level for picture material at black was zero Volts DC or 0 IRE. (IRE and other terms used in this explanation are defined in the glossary of terms.) The level for white was 716 millivolts or 100 IRE. When color first came to NTSC it became necessary to raise the level of black to +7.5 IRE, compressing the dynamic range between black and white. By the time all three networks had gone to color in 1964 the +7.5 IRE black level was no longer necessary, but the rule wasn't changed. In 1990 there was a push in the program production industry to change it back to 0 IRE. That change actually took place in Japan but didn't happen here in the United States. Part of the push for the change came from the fact that a lot of program material was being produced in the SMPTE component video domain which uses 0 Volts for black. The majority of DVD's are mastered using 0 Volts DC for black. In reality all DVD's should be mastered using 0 Volts DC for black, but there are exceptions. The dynamic range of the component video picture is enhanced relative to the standard composite NTSC color system. The DVD player will reduce that dynamic range, moving it from 0 IRE to
+ 7.5 IRE, in providing an NTSC output which is necessary for compatibility with standard home display devices. Which position should you use? That depends on the display device itself or the video processor driving the display and/or how the DVD player is being used in the entire home entertainment system. Ideally you would want to use the 0 Volts black position to preserve the dynamic range available from the disc itself. That is only practical if the set has an individual memory for the enhanced mode. This individual memory should be able to remember settings for contrast and gray scale in addition to brightness. We've come across video processors claiming compatibility with black at 0 Volts DC but found that some of them don't actually have that capability. In short, reading manufacturer's specifications doesn't always give you an accurate picture. If the processor reports an ability to deal with black at 0 Volts and has at least 10 bit processing, chances are far better that it will function properly than if the processing is 8 bits. Video processing in many Plasma displays is less than 8 bits. As a result you'll see lots of problems in the picture, especially in the areas near black. The selection you make will most likely affect all video outputs of the DVD player. If you are routing the NTSC output for signal monitoring and the component output for actual viewing, the NTSC picture probably won't look correct. Blacks will be crushed from the DVD picture if the NTSC monitor has been set up to look correct on all other sources. It takes a well thought out system to take advantage of the 0 volt "Enhanced" capability of DVD players.<<<
That leads me to the next dummy question:
In order to avoid luma filtering ,could the codec ,let's say ,COMPARE the darkest luma value that can find during start credits(usually black) to a refference "enhanced black mode" and set is as luma 0+difference?And then lower all luma range according to this difference?
I'm not a coder,so i cant say if such a feature is already implemented in CVS,but i think it would be great to have another option in configuration panel
Hope i'm making some sense.I'm looking for a way to defeat luma probs with "near black" DVDs mastered with +7.5 IRE (there are quite a lot,and i think that's the reason why some DVDs really need lumoff for tv output and some not)
Ps : -Sorry for the misleading.I realise that you 're bugfixing like crazy wright now
-Dont kill the messanger:)
regards,
george
I've been using all of our latest methods for the black blocks issue with good results(lumoff,lumafilter,blockbuster...).Forgive my ignorance,but besides mpeg4 that prob seems to exist even in mpeg2(for another reason?).I'm reffering to the famous IRE prob.I'm copying a brief expanation from a site:
>>>In the original black and white television system the voltage level for picture material at black was zero Volts DC or 0 IRE. (IRE and other terms used in this explanation are defined in the glossary of terms.) The level for white was 716 millivolts or 100 IRE. When color first came to NTSC it became necessary to raise the level of black to +7.5 IRE, compressing the dynamic range between black and white. By the time all three networks had gone to color in 1964 the +7.5 IRE black level was no longer necessary, but the rule wasn't changed. In 1990 there was a push in the program production industry to change it back to 0 IRE. That change actually took place in Japan but didn't happen here in the United States. Part of the push for the change came from the fact that a lot of program material was being produced in the SMPTE component video domain which uses 0 Volts for black. The majority of DVD's are mastered using 0 Volts DC for black. In reality all DVD's should be mastered using 0 Volts DC for black, but there are exceptions. The dynamic range of the component video picture is enhanced relative to the standard composite NTSC color system. The DVD player will reduce that dynamic range, moving it from 0 IRE to
+ 7.5 IRE, in providing an NTSC output which is necessary for compatibility with standard home display devices. Which position should you use? That depends on the display device itself or the video processor driving the display and/or how the DVD player is being used in the entire home entertainment system. Ideally you would want to use the 0 Volts black position to preserve the dynamic range available from the disc itself. That is only practical if the set has an individual memory for the enhanced mode. This individual memory should be able to remember settings for contrast and gray scale in addition to brightness. We've come across video processors claiming compatibility with black at 0 Volts DC but found that some of them don't actually have that capability. In short, reading manufacturer's specifications doesn't always give you an accurate picture. If the processor reports an ability to deal with black at 0 Volts and has at least 10 bit processing, chances are far better that it will function properly than if the processing is 8 bits. Video processing in many Plasma displays is less than 8 bits. As a result you'll see lots of problems in the picture, especially in the areas near black. The selection you make will most likely affect all video outputs of the DVD player. If you are routing the NTSC output for signal monitoring and the component output for actual viewing, the NTSC picture probably won't look correct. Blacks will be crushed from the DVD picture if the NTSC monitor has been set up to look correct on all other sources. It takes a well thought out system to take advantage of the 0 volt "Enhanced" capability of DVD players.<<<
That leads me to the next dummy question:
In order to avoid luma filtering ,could the codec ,let's say ,COMPARE the darkest luma value that can find during start credits(usually black) to a refference "enhanced black mode" and set is as luma 0+difference?And then lower all luma range according to this difference?
I'm not a coder,so i cant say if such a feature is already implemented in CVS,but i think it would be great to have another option in configuration panel
Hope i'm making some sense.I'm looking for a way to defeat luma probs with "near black" DVDs mastered with +7.5 IRE (there are quite a lot,and i think that's the reason why some DVDs really need lumoff for tv output and some not)
Ps : -Sorry for the misleading.I realise that you 're bugfixing like crazy wright now
-Dont kill the messanger:)
regards,
george