ATI & NVIDIA Video Enhancement Quality Tested

[W1zzard]

High-Definition content is available everywhere nowadays. We take a close look at what the drivers from AMD and NVIDIA offer in order to improve the quality of video playback. HQV Benchmark 2.0 uses 39 different tests that put the available image enhancement features in the spot light.

Show full review

 

[EastCoasthandle]

No surprise there. Is the HQV Benchmark 2.0 something users can try for themselves (free/demo)?

 

[W1zzard]

unfortunately it's not free, if you are a member of the press, they have a form on their site

 

[WarEagleAU]

well you could give us a link to said form, ha ha. Glad to see both ATI and NVidia both are taking control of video processing and realizing it is more movies and content than about games these days.

 

[DannibusX]

Nice review! I never even thought of benchmarking for video playback.

 

[mstenholm]

Did I miss that part where you did a monitor calibration with a good calibrator? Skin tones will be off with most consumer monitors such as the LG out of the box.

 

[W1zzard]

Did I miss that part where you did a monitor calibration with a good calibrator? Skin tones will be off with most consumer monitors such as the LG out of the box.

hqv accounts for that by switching a "normal" skin image with a yellow and red image

 

[W1zzard]

those are the 3 test images, no matter your calibration you can clearly see the differences.

with proper correction all 3 skin colors will look right and the shirt and background color won't change

 

[Delta6326]

sweet review!

 

[DaC]

Does someone knows where to finda a tutorial about image calibration ? I really can't make my mind on how to configure playback settings.
BTW, really nice review Wizz.

 

[mstenholm]

Does someone knows where to finda a tutorial about image calibration ? I really can't make my mind on how to configure playback settings.
BTW, really nice review Wizz.

First step would be to buy/borrow a calibrator. An cheaper and way less accurate methode is to obtain a correct "exposed" photo and adjust your monitor until you are happy. Lets assume that you took any of these three photos. If your used the first one and got your monitor to show natural skin tones then all other photos/video/what ever would have a wrong color/temperature/gamma.

 

[Izliecies]

those are the 3 test images, no matter your calibration you can clearly see the differences.

with proper correction all 3 skin colors will look right and the shirt and background color won't change

Which one is the "right one"?

 

[Mistral]

Which one is the "right one"?

If I had to guess, the first. Or maybe the third. The second one is way off, too yellowish. Then again, that's on my screen.

Great read, thanks W1zz.

 

[newtekie1]

Very interesting read. However, I wonder how much of a difference this actually makes. I mean, it seems to me like this is one of those "you'll notice it when on freeze the image and look at it really really hard, but not when you are actually watching the movie" type of things. I'm being serious here, how much of a real world difference is there between the worst and the best in this test? Is there really going to be a noticeable difference between an HD4290 w/ Defaults or Optimized and an HD5870 w/ Default or Optimized?

 

[erocker]

Very interesting read. However, I wonder how much of a difference this actually makes. I mean, it seems to me like this is one of those "you'll notice it when on freeze the image and look at it really really hard, but not when you are actually watching the movie" type of things. I'm being serious here, how much of a real world difference is there between the worst and the best in this test? Is there really going to be a noticeable difference between an HD4290 w/ Defaults or Optimized and an HD5870 w/ Default or Optimized?

I know I get a pretty drastic difference between my "out of the box" settings for my 5850 and going in CCC and setting my video settings. I remember helping DonInKansas adjust his video settings and he was pretty amazed at the difference.

I do wish both ATi and Nvidia would have a nice mix of different presets to choose from.

 

[Sasqui]

Way interesting. Personally, I've been watching more and more DVD content from my PC and noticed a big improvement with my 5870 in 1080 upscale quality from my old 4870 card.(perhaps it's the software...?)

Would like to hear details on what is "optimized"? Was that simply trial and error to find the best driver settings?

 

[air_ii]

I read on another forum that the "optimised" settings will be integrated as default in the next Catalyst release (no clue why they didn't do it in the 10.6).

 

[W1zzard]

I read on another forum that the "optimised" settings will be integrated as default in the next Catalyst release (no clue why they didn't do it in the 10.6).

yes i mentioned that too. what they will enable by default is our "default on" settings.
since our "optimized" ones were test by test, there is no setting that fits all tests

 

[W1zzard]

Would like to hear details on what is "optimized"? Was that simply trial and error to find the best driver settings?

yes that's basically it. you can change the sliders with ccc in the foreground and the video in the background will change instantly

 

[wahdangun]

wizz, can u give a photo along side the score ? its hard to imagine with just looking in the score.

maybe each photo in out-of-the box, default on, and optimized.

 

[trt740]

Very nice review well done

 

[erocker]

Didn't even notice you are using Catalyst 10.6 in this review. Looks like AMD is updating their site right now. Can't wait to try them out.

 

[trickson]

THANK YOU FOR THE GREAT REVIEW !!!!!! I want to see them 10.6 drivers and would really love to use that benchmark for mine . But it is not free ( yet ) . Any way keep up the great work .

 

[Sasqui]

yes that's basically it. you can change the sliders with ccc in the foreground and the video in the background will change instantly

I can't recall if all of the sliders have the performance vs. quality level listed with them. If so, I think you're saying you simply bumped all of them to highest quality for the best test.

I can't imagine a 5870 even blinking with everything set to highest.

 

[DannibusX]

W1z, just wanted to give you a heads up if you didn't know, but Terry Makedon linked this review on his Twitter account.

http://twitter.com/CatalystMaker