The HQV Benchmark
The HQV Benchmark
HQV stands for Hollywood Quality Video. The HQV DVD video benchmark includes a regular DVD with different scenes that present potential visual problems to a DVD decoder. Points are awarded if the decoder handles the problems so that viewers don't see them; points are not awarded if it doesn't do the job. In looking over our benchmark results, we immediately observed that the Intel and AMD processors scored the same, irrespective of their model numbers (and in fact, the results from both processors are pretty similar).
Rather than stepping you through the entire sequence of tests and explaining them in detail, we refer you to Don Woligroski's January 9, 2007, article for Tom's Hardware entitled Avivo vs. Purevideo, Round 1 wherein he walks readers through the whole shebang. Here, we provide a brief table instead that summarizes results for each vendor's CPUs with a one-liner explanation and remarks for each one. We need only observe here that each test (or sub-test, when a particular test involves multiple versions or iterations) produces a score (maximum scores vary), where the score indicates the degree to which the decoder can handle things perfectly.
| Test Name | Max. Score | AMD | Intel | Explanation/remarks |
|---|---|---|---|---|
| De-interlacing colour | 10 | 10 | 10 | Able to show small colour bars flicker-free |
| De-interlacing jaggies 1 | 5 | 5 | 5 | Looks for jaggies rotating a line in a circle |
| De-interlacing jaggies 2 | 5 | 3 | 3 | 3 small lines moving slightly within an arc |
| De-interlacing: flag test | 10 | 10 | 10 | Actual video of Old Glory flapping away |
| Detail enhancement | 10 | 10 | 10 | Enhance detail of blurry parts in video scene |
| Noise reduction | 10 | 10 | 5 | Ability to remove compression artefacts |
| Motion adaptive noise reduction | 10 | 10 | 10 | Ability to show motion with no motion trails of smearing artefacts |
| Film detail
(2:3 pulldown) |
10 | 10 | 10 | Conversion of film to video, Intel got noted (but not dinged) for failing 2:3 pulldown detection |
| Film Cadence | 40 | 25 | 25 | Eight tests of 5 points each; pass or fail; tests for flicker or jaggies |
| Scrolling titles (horiz) | 10 | 5 | 5 | Looks for jaggies or artefacts in title text |
| Scrolling titles (vert) | 10 | 5 | 5 | Looks for jaggies or artefacts in title text |
| TOTALS | 130 | 93 | 88 | AMD edges Intel only in noise reduction |
Overall, the CPU seems to make very little difference in the output of high quality video when everything except CPU and motherboard stay the same between builds. We have to guess this reflects the common graphics card more than differences among CPUs.
- Previous page 3DMark05
- Next page The A/V Experience
- Google takes its calendar on the go
- Steve Jobs promises a "Greener" Apple
- Intel claims 2-year lead over AMD's 45nm CPUs
- IBM uses natural process for chip making
- China mobile phone subscribers at nearly 481 million in March
- Opting for RAID Level Migration
- The Best Games Never Published, Part 3
- Four OpenSource Groupware Tools
- HD DVD experiences playback problems on Xbox 360
- Apple curing laptop battery woes with software update
-
DIY Projector
-
diy mac
-
diy htpc
-
DIY NAS
-
DIY ultra high refresh rate
-
common mistakes on new build pc
-
Build a pc
-
new build amd
-
gaming PC build
-
AMD budget build
-
first time system build
-
no picture after new build pc
-
Build Your Own Silent HTPC
-
how to build your own solar
-
build your own PC pros and
-
s3 hdmi
-
hdmi