I am reading principledtechnologies pdf of the test and noted some problems with the test bed setups and information (Alan noted some issues as well):
(from page2)
This is false. Standard defintion TV or SDTV is not the same as analog broadcast TV. SDTV can be 480i or 480p (interlaced or progressive) Where analog broadcast TV has but a single format, SDTV has 12 digital formats and can be 4:3 or 16:9. Reviewers, particularly video reviewers, in addition to reviewing a product, also serve to inform the reader. It is critical that the information is correct, otherwise it puts their review into question. The fact is that the term "SDTV" is often substituted for analog TV broadcasts, which it absolutely is not.
I also question outputting 1080i to a LCD display with a native resolution of 720p. IMO, all the tests should have been done at 720p resolution.
Also, if you read the Catalyst driver setup, the settings they used are not even 1080i or 720p resolutions, they are using mapped resolutions of 1776x1000 (its an optimised resolution to map 1920x1080 down to 1776x1000 to reduce overscan on some displays) and 1152x648 another mapped resolution(Catalyst drivers have native 1280x720 resolution built into the driver) The Nvidia settings specifically use actual resolutions rather than mapped resolutions "Under TV Overscan Correction, select 0 next to "Zoom Out %" and "9. In the HDTV Overscan Compensation screen, select Underscan." These 2 setting specifically direct the driver to output the resolution you select, not map the resolution to fit the screen like the Catalyst driver settings they selected. I'm not sure that they would have any bearing on the tests they run here, but it does offer yet another variable into the mix.
Outputing 1080i to a 720p display means interpolation that occurs in the display cicuits itself. The interpolation should be taking place in the output devices if thats what you are testing, not the display so as to take that variable out of the equation.
It also looks like this thread isn't going to be taken seriously, so I'll not waste my time adding to it beyond this post. I'll contribute to another if the topic rises again and is actuall a discussion on the topic rather than a dissection of Off topic posters "witty" retorts.
(from page2)
This is false. Standard defintion TV or SDTV is not the same as analog broadcast TV. SDTV can be 480i or 480p (interlaced or progressive) Where analog broadcast TV has but a single format, SDTV has 12 digital formats and can be 4:3 or 16:9. Reviewers, particularly video reviewers, in addition to reviewing a product, also serve to inform the reader. It is critical that the information is correct, otherwise it puts their review into question. The fact is that the term "SDTV" is often substituted for analog TV broadcasts, which it absolutely is not.
I also question outputting 1080i to a LCD display with a native resolution of 720p. IMO, all the tests should have been done at 720p resolution.
Also, if you read the Catalyst driver setup, the settings they used are not even 1080i or 720p resolutions, they are using mapped resolutions of 1776x1000 (its an optimised resolution to map 1920x1080 down to 1776x1000 to reduce overscan on some displays) and 1152x648 another mapped resolution(Catalyst drivers have native 1280x720 resolution built into the driver) The Nvidia settings specifically use actual resolutions rather than mapped resolutions "Under TV Overscan Correction, select 0 next to "Zoom Out %" and "9. In the HDTV Overscan Compensation screen, select Underscan." These 2 setting specifically direct the driver to output the resolution you select, not map the resolution to fit the screen like the Catalyst driver settings they selected. I'm not sure that they would have any bearing on the tests they run here, but it does offer yet another variable into the mix.
Outputing 1080i to a 720p display means interpolation that occurs in the display cicuits itself. The interpolation should be taking place in the output devices if thats what you are testing, not the display so as to take that variable out of the equation.
It also looks like this thread isn't going to be taken seriously, so I'll not waste my time adding to it beyond this post. I'll contribute to another if the topic rises again and is actuall a discussion on the topic rather than a dissection of Off topic posters "witty" retorts.