HQV Benchmark 2.0 Analysis – ATi, nVidia and Intel

Overview:
This chapter focuses on various cadences which can occur in video content from a plethora of sources. Severe moire is used as an indication that cadence lock has not been achieved. It is the amount of time it t akes to achieve a cadence lock, indicated by a reduction in the moire distortion, which is the scoring mechanism for this specific test.

Notes:
Both of these tests check for colour artifacts which would be introduced by upsampling of previously downsampled video. When video is compressed the colour data is condensed and this truncated colour information can result in artifacts if the image is not correctly upscaled from the remaining information.

Scoring:
To score this correctly we have to analyse the patches which should have no colour lines in the dark areas around the colour bars. The colour bars should be smooth.

The image above shows heavy artifacting in the black areas between the bars which has been highlighted to make it easier for you to notice. This would result in a lower score. We need to carefully analyse chroma bleed.

Upsampling Errors: Interlace Chroma Problem (ICP) (Score either 5, 2 or 0)
No Visible colour lines can be observed in the dark region around the colour bars of patch. The colour bars should be smooth. – Scores 5
Very faint colour lines can be observed in the dark region around the colour bars of patch. The colour bars have some mild ‘stair' stepping. – Scores 2
Highly observable colour lines in the dark region around the colour bars of patch and within the colour bars. – Scores 0

The ATi cards all score 5 in this test and again the higher end GTX boards from nVidia match the results. The Intel solutions have some noticeable stair stepping in the results as does the GT220. The G210 gives the worst result in this test as colour lines are really noticeable.

Become a Patron!