HQV Benchmarks

The HQV Benchmark DVD is made by Silicon Optix and has become a standard for testing the quality of DVD decoders. The tests were made to showcase the capabilities of very high end DVD hardware, and it just happens to make a very objective series of benchmarks for testing more moderately priced DVD players. The DVD contains numerous video clips to test how well the decoder handles some common problem-areas for some of the lower quality players. The videos test things like picture detail and noise reduction, and points are scored for each test according to how well the video processor does.

ATI recently came out with some new drivers which were meant to improve the score on these tests, and today we are testing them against the latest version of NVIDIA's PureVideo decoding. It should be noted that while ATI's DVD decoder is free with an AIW card, the PureVideo decoder adds additional cost to NVIDIA products.

Many of the functions tested in HQV aren't limited in application to DVD decoding. Things like deinterlacing need to happen on the AIW when tuning TV signals as well. This way we can also get an idea of the quality of video the AIW will be capable of in general.

Color Bar/Vertical Detail

The first test on the HQV DVD is the Color Bar/Vertical Detail test. This basically tests how well the video processor tells if something is moving or not, to ensure proper deinterlacing. We found that ATI did slightly better than NVIDIA in this test, as there was some minor flickering that was evident on NVIDIA's side. ATI scored perfectly here.

  • ATI: 10
  • NVIDIA: 5
  • (highest score: 10)




The Card and The Features HQV Jaggies Tests
Comments Locked

43 Comments

View All Comments

  • bldckstark - Friday, February 10, 2006 - link

    The example shown does not change the overall score of the card. The example shown is for the readers reference to the test, and is not what the test scored from. There may be other reasons someone may not give these tests merit, but this is not one of them. You could maybe rank on the author for this, but not the tests.
  • mpeavid - Friday, February 10, 2006 - link

    The example shown does not change the overall score of the card.

    But how do we know that? Take example cadence 2224. According to the text the same item is being compared, yet different frames are clearly shown. If their methadology was more concise, their text is not.

    You have to be clear about this or it misleads your readers. Its like doing a 3D test using 2 different scenes to render. Anandtech uses all the same 3D scenes to render right?

  • rjm55 - Friday, February 10, 2006 - link

    Other sites did AIW 1900 reviews on January 31st. Why so long for AT? Did ATI pass you over on sending a sample?
  • fishbits - Friday, February 10, 2006 - link

    quote:

    If performance continues to increase at the rate that it has been, we aren't sure how game software will be able to keep up.

    By adding more polys, textures, particles, lights, shadows and shaders. You really didn't know this? Call any respected game dev house and ask them if they could possibly come up with a use for more GPU horsepower. The answer will be "Of course genius, we've got code and models we're waiting for capable hardware to run on, it's been that way for years. We'll take every bit of it we can get." Tell Anand I want you to spend this weekend benching EQ2 maxed out and tell us Monday if "we" still "aren't sure."

    Anyhow, sounds like a nice card, but I'd rather have a more dedicated gaming card and a seperate TV tuner solution.
  • Griswold - Monday, February 13, 2006 - link

    Of course they want more power so they dont have to write efficient and optimized code. Especially your EQ2 example comes to mind. There are far too few companies that come up with highly optimized code that will run top notch on current hardware and provide extra eye candy on future generations.
  • Backslider - Friday, February 10, 2006 - link

    The 7800GT used in the test must be stock. The one I purchased came overclocked and performs much better than what the benchmarks are showing.

    ATI is still too pricy at the moment, I looked up and down for an X1800XL that could come with in price range of the 7800GT that I purchased, and I couldn't find one. I wasn't going to pay $60 over when they perform so identical. The prices were approx.

    X1800XL 256 Stock $330
    7800GT 256 OC $270

    ATI get those prices down.
  • tuteja1986 - Friday, February 10, 2006 - link

    I sold my 7800GTX bought a x1900XT and i couldn't be happier :! if G71 fixes some issues like IQ and HDR with AA then i will sell my X1900XT and buy a 7900GTX :) or eles wait for R6XX and G8X.
  • Backslider - Friday, February 10, 2006 - link

    Having owned an X800xl and a 7800GT, I honestly didn't see an IQ difference. The whole HDR with AA thing, well, you must play a lot of Far Cry.

    Good luck with keeping up with the latest and greatest though, it's almost a game with in it self. If you sell at the right times, you can upgrade for very little and still have the newest toys.

    Happy gaming
  • MrKaz - Friday, February 10, 2006 - link

    I have a ati 9700 and geforce 6600gt and ati rendering look better.

    There are some annoying layers/plates on the nvidia rendering that i dont like.

    And just one note: the display is the same on both cards.
  • DeathByDuke - Friday, February 10, 2006 - link

    I'd certainly buy one if it was around $299-349, considering it performs closer to a much more expensive X1800XT

Log in

Don't have an account? Sign up now