X-bit labs

Started by: Ajax9000 | Date 02/26/07

Comments: 26 | Last Comment:  08/08/07

[1-18]

1. http://anandtech.com/video/showdoc.aspx?i=2923
looked at this recently also, and got wildly different findings for HQV. They were using HQV HD, but you seem to have used standard HQV. Is this correct, and if so, why didn't you try HQV HD.
BTW the Anandtech methods were strongly criticised:
http://anandtech.com/talkarticle.aspx?frmResourceID=2923
http://www.dtvforum.info/index.php?showtopic=46572

1. HQV and HQV HD are totally different benchmarks with the former shipping and the latter still in beta.

2. We do not have BD/HD DVD optical drives currently in the lab, nor we have BD/HD DVD hardware player for reference. We also do not have BD and HD DVD content.

3. We will use HQV HD in future.

I'm a bit puzzled by your 2nd comment as I didn't mention BD/HDDVD. You made the point on p2 of the review that the main HD video formats utilized on HD-DVD and Blu-ray media are H.264 and VC-1. So by testing playback of H.264 and VC-1, are you not (in effect) testing the type of content held on BD/HDDVD? Not criticizing, just asking.

Of course, not using HQV HD because it is beta is an entirely reasonable thing to do!

2. How does all this compare with a yesterday system , let's say XP3200 with a X800 and 1GB RAM . Let us now so then I might feel the urge to upgrade .
Also I will Never use Power/Win DVD player , I want to see the tests done with open source players :MPlayer/Media player classic/BSPlayer/Zoom etc.
Thank you for 11 pages of telling me exactly where I stand.

I agree completly with you, 'Chew BAKA'. Dual-core makes all the difference, and it will be great to see single-core tests, because not everyone have this kind of high end stuff. And Media Player Classic (or BSplayer or MPlayer or even VLC) test will be really, really great. PowerDVD isn't a good player (expensive, bad interface, filters configurations are poor, etc.). I agree with 'meateater' too when he talks about CoreAVC decoder test.

BTW, thanks for this good review.
Keep going!
;)

3. You should throw in COREAVC software decoder next time. You'll find out what a lot of hype the GFX card decoding is.

Also do some tests with Linux as people want to use the graphics card on their MythTV box to decode video.

4. Thank you very much for such an informative article. It is good to know how the HD acceleration is coming along and how the performace of the graphics cards has only a little impact on decoding performance. I use an Athlon 3200+ and a 7300GS for my HTPC and it works very well. So far, besides HD-DVD, I have not found a clip I cannot play. For the most part, the clips you can download are low bitrate and play fine on my hardware.

However, I have a question about the picture quality of the various cards. In the review you did an excelent job at detailing the differences in acceleration, but you did not even state what cards you were testing in the HQV benchmark. Do all cards perform the same? This is likely the case with DVDs, but when you do your HQV HD benchmarks, I think it would be helpful for us budget buyers to know what HQV scores our 7300GS gets compared to a 7900GTX. The supported Purevideo features are different for HD material and I would like to know what I am missing with my 7300GS.

I am also curious about the the MPEG 2 1080i content. The 7300GS cannot handle inverse telecine in HD content. I have tried it and it drops every other frame. You had IVTC enabled for the test. Did you exeperience this with your test? Did the x1300pro have any problems with interlaced content and 3:2 pulldown enabled? Without IVTC on my 7300GS what are my HQV scores? I cannot enable IVTC every time I switch between HD and SD content. What am I missing out on? These are a few of the questions that me and many other HTPC enthusiasts have been struggling with. We want to know if it makes sense to buy a 7900GTX, 7600GT, or a nice, silent, low power 7300GS.

-Thanks

(BTW, there are numerous errors in the "i"s and "p"s and 1080p/720p graphs and charts. Your VC-1 clip is 720p and you show 1080p and in the chart showing the bitrates you say the mpeg2 clip with 1080p when it is interlaced)

5. A very interesting article but I think there is something that could be added.

Once you know that all graphics cards can play theese content in a new dual core system it would be interesting to see what is, both, graphic anc CPU (and system) power consumption during playback.

For passively-cooled HTPCs it would be more important to know what the power consumption of a TV-set is, so, there are no plans to measure power consumption of computer systems during media playback.

WTF? How does a TV have anything to do with passively cooling the CPU and GPU? All that matters is the amount of power they use and hence how much heat needs to be transferred away.

6. Congratulations for the article. Quite informative, and learned a lot from it.

One question, though. You tested the quality of the decoding using a DVD. Should I assume that using other codecs besides the MPEG-2 the quality, improvements in the image and noise reduction are also as great?

I'm asking because from reading the article I got the sense that ppl using old PC's are basically missing a LOT of video quality due to their older hardware without avivo or purevideo
(I'm thinking specifically if/when using divx/xvid based in MPEG-4 and also with H.264)

7. Error when clicking on 'printable version':

Fatal error: Call to undefined function: x_get_banner() in /usr/home/xbit/www/admin/template_lib.php(364) : eval()'d code on line 31

8. I was wondering if there is any sort of test like this with previous gen graphic cards with a lower performing processor..

9. The best review I have yet read on the subject although it still misses some important parts:

-No checks were made whether DXVA decoding was actually being used or not. For example AVIVO doesn't even support VC-1 decoding. And that there are any players out there that use DXVA for ASP content is also new to me.

-It went unnoticed that NVIDIA cheats by not performing the deblocking that is a key part of both the AVC and VC-1 specs. To your defense it has to be said that no internet hardware site has brought this up yet although it is a well known issue on certain forums.

10. i would also enjoy to have a comparison of the nvidia and ati driver: clone mode, dual display, external reader, .....

11. Thanks for the great review.
may i ask if the video cards are plugged into the Samsung 244 using VGA or DVI?
noticed so card tends to do alot better with VGA ut picture quality isnt the same.
Thanks
Bobby

12. I would recommend a test with the matrox P series card since they support more resolutions than ati or nvidia

13. Did you playback the 1080p DivX videos in full screen mode? I observed a lot of jerky playback when playing full screen. 7950 does not seem to have this problem. I'm using 100.65 driver.

14. “The entry-level solutions are different. Such graphics cards either have slow memory with a narrow memory bus like the GeForce 7300 GS or have few pixel processors on board like the Radeon X1300 Pro. Their video processors are thus greatly constrained and do not suit well for playing HD formats with a high bit-rate or resolution, especially if you’ve got a single-core CPU. The rest of today’s graphics cards, from the modest Radeon X1600 Pro and GeForce 7600 GT to the ragingly fast Radeon X1950 XTX and GeForce 8800 GTX, do their job well.”

If you look at the data you will clearly see that the 7300GS is very close to the 7600GT & 7900GS in the tests that are most relevant for home HD video viewing (Blu-ray & HD-DVD) and the X1600 Pro is not even close. Great article but the summary lets it down.
Due to the competitiveness of 7300GS I to would have liked to see how it compared to the other 7 series cards in the quality tests, due to it having less features implemented.

To reiterate my last post, if you look at the difference in CPU load between the 7300GS, 7600GT & 8800GTX for H.264, VC-1 & MPEG2 (1080i only) playback the differences are negligible. I chose these formats because they relate to commercial HD video releases as opposed to DivX & WMV HD which are more geek formats; no offence intended.
I looked at the average and maximum CPU load and ignored the minimum as it’s not particularly significant. The 7300GS used 1.6% and 4.4% more CPU load than 7600GT & 8800GTX respectively. This shows that the Nvidia software is almost entirely dependent on the GPU clock speed for these formats; the three cards are all around 550 MHz. This contradicts your summary massively and WTF were you thinking?
Looking briefly at the other metrics doesn’t seem to change the picture either.
In fact the 7300GS beats the 7950 GX2 & the X1950 XTX!!!
Now if only you’d told us whether the 7300GS suffered from poor video quality compared to the 7600GT etc, we might have really learned something useful here. Opportunity missed I think. :(

15. "we could not use commercial Blu-ray or HD DVD discs just because we didn’t have an appropriate optical drive."

What!? Are you fracking serious? Pioneer has had an an internal BD player out for a long time now and how about the Xbox add-on HD-DVD drive? What a load of crap. You could've picked one up at your local Fry's or hell buy one off of Newegg or some other vendor. Sheesh.

"What!? Are you fracking serious? Pioneer has had an an internal BD player out for a long time now and how about the Xbox add-on HD-DVD drive?"

There's a review of HD-DVD with PureVideo here: http://www.pcper.com/article.php?aid=328&type=expert&pid=4
It shows very good results with H.264 encoded films.

16. If you are playing a HD-DVD disk with PowerDVD and the Xbox-360 drive be warned that there are issues with this combination when using a NVidia card less than a 7600GT. The Cyberlink HD Ready analysis tool states that a 7300GS is not supported for HD-DVD playback and links you to their website which shows a 7600GT as the minimum requirement. PowerDVD Ultra 7.3.2605 refuses to even attempt to play a HD-DVD with a 7300GS, which seems rather excessive. However, if you go to the configuration page and select Hardware Acceleration it does then allow playback. I successfully viewed a HD disk with a Core Duo T2400 1.83 GHz and a 7300GS @ 550 MHz. CPU utilisation was around 75% but there were no glitches.

17. The 8500/8600 series offer dramatically improved HD decoding, especially for H.264 I think. See http://www.legitreviews.com/article/486/2/
The 8600 GTS reduces CPU load by a factor of ~3 compared to a 7600 GT; you don’t even need dual core anymore.

18. Excelent overview of the different formats and their requirements. Thanks for taking the time and effort to do the testing.

[1-18]