When certain types of source material are transferred to NTSC 30 frames per second video, the math doesn't add up. Film for example is shot at 24 frames per second. To equally distribute those 24 frames over 30 NTS frames, a process called 'pulldown' is initiated.

Good cadence detection will allow a DVD player to detect and reconstruct the original source material, resulting in a much more detailed image without interlacing artifacts.

Film Detail Test (2:3 Pulldown Video): out of 10 points

The most commonly encountered type of pulldown is called 2:3 pulldown, which is used when film is transferred to video. Because of this, the 2:3 pulldown test is called the "Film Detail test," and is given a larger weight in scoring than the other tests.

This test shows a scene with a racecar driving around a track. If the 2:3 pulldown is not detected by the player, a moiré pattern artifact will show up during playback. Properly made DVDs have a "flag" inserted into the data that allows DVD players to recognize this 2:3 pulldown and to correct the display accordingly. This test simulates a bad edit by using no flag, therefore relying on the decoder's ability to identify and react to the 2:3 pulldown video.

Good decoder processing will quickly identify the 2:3 pulldown and clean up the video so that the moiré pattern created by the process is corrected.

The Film Detail test

Software: 0 points
There is no pulldown detection working in the PowerDVD software if there is no flag on the DVD for it to detect.

ATI Avivo: 5 points
Without the pulldown detection checkbox enabled in the catalyst drivers, Avivo detected nothing. But when the checkbox was enabled, Avivo managed to correct the pulldown artifacts after a half of a second or so. Because of the lag, they get five out of 10 points.

Geforce Purevideo: 5 points
Similar to Avivo, there is a slight lag before Purevideo locks onto the signal. There is a difference we experienced however in that Purevideo has to re-establish a lock every time the looped scene begins again, while the Radeon seemed to only need to establish a lock during the first loop. This is not a killer difference, though; according to HQV's criteria, Purevideo will also get five out of 10 points in this test.