
Perfect World Beta 

Life After DX 11 

Gigabyte Cluster 

FiringSquad User Articles 

Shuttle XPC Design Contest 

NVIDIA SLI GIVEAWAY CLUSTER 

AMD Contest Group 

Foxconn Graphic Design Contest 

Editors Challenge Sponsored by Inte... 

FiringSquad Editors Challenge Round... 

sweno 

|

|
|
|
12 entry(ies) in this category
|
 Does this resolution make me look fat? (10 comments ) by: Swatt (25) | Posted in cluster Editors Challenge Sponsored by Intel Round 2 Posted 75 months ago ( edited 75 months ago ) in category DEFAULT
» MEDIA (9)
 Figure 1

 Figure 2

 Figure 3

 Figure 4

 3dmark 05 Results

 3dmark 05 Framerates

 3dmark 06 Results

 3dmark 06 Framerates

 Oblivion Benchmarks

| Its time to face the facts people, widescreen gaming is where it’s at. With the transition from the standard definition TV in its 4:3 aspect ratio to new HDTV’s and their fancy 16:9 aspect ratio the industry is on the move. Sony and Microsoft saw it coming and even Nintendo (who didn’t hop on the HD bandwagon with the Wii) had the foresight to include a 16:9 option. Caught up in the middle of all this are PC gamers.
While getting your hands on a widescreen monitor is easy enough, finding games that natively support it is another matter entirely. While Microsoft is beginning to enforce that games for PC support widescreen with their “games for windows” program, we are still left with a large amount of the games we play unsupportive of widescreen resolutions. What does that mean? It means that if you have a widescreen monitor, chances are your favorite characters in your favorite games look like they could loose a little weight. In fact, only about 120 PC games support widescreen displays natively to date and trust me when I say that not all of those games are widely played (Left Behind: Eternal Forces anyone?)
The good news is that widescreen displays are no longer niche and are becoming more ‘mainstream’, along with workarounds for many unsupported games which are available for the widescreen gamer. Many games currently in development are stating that they natively support widescreen resolutions as well. While this is all fine and good, what we are here today to do is examine the effects of widescreen resolutions on gameplay and pc performance.
Many newer games such as Half Life 2 support widescreen resolutions right out of the box. As displayed in fig 1 the resolution and aspect ratio options are robust and easily modified to fit your needs. In fig 2 – 4 we can see the difference the resolution can make within the game itself. You can see in fig 4 that there is quite a difference in overall screen size.
This increased size does not come without a cost. By increasing the size of the screen we have increased the resolution as well. The resolution of the widescreen panel used in testing is 1680 x 1050 which means the 1,764,000 pixels total. This is a common resolution for a mid range widescreen display. A standard 4:3 LCD monitor will display a native resolution of 1280 x 1024 which is 1,310,720. That’s a difference of 453,280 pixels. After crunching the numbers we find that the widescreen panel is displaying 26% more pixels than the 4:3 panel. The implications of this means that to take advantage of the native resolutions of the panel you’re asking your video card to process 26% more image per frame.
So another question is how do video cards handle the extra work? That’s what were about to find out. We’re took two different displays, one 16:9 and one 4:3, and run them in their native resolutions. Well measured fps (frames per Second) on both monitors from different games and comparing the results.
Test Setup:
For our test we used a mainstream system outfitted as follows:
CPU: Athlon 64 3700+ (San Diego)
Video Card: EVGA 7600 GT
Motherboard: Asus A8N5X (nForce 4)
Ram: 2 x 1GB DDR 3200
NVIDIA Driver: 6.14.10.9371
Widescreen Monitor: BenQ FP202W (native resolution 1680 x 1050)
4:3 Monitor: BenQ FP19GX (native resolution 1280 x 1024)
Please note that we are not measuring the performance of the monitors but rather the performance of the PC in widescreen and fullscreen resolutions. We should fully expect that there will be a performance decrease on the larger resolution and the objective is to illustrate what the performance variation is in games.
To illustrate the effects of the different resolutions we have opted to run all tests in the native resolutions of the monitors. You will notice that the vertical resolution is very close to the same where the horizontal resolution differs greatly due to the widescreen nature. Also, running in the native resolution of a monitor ensures the clearest picture possible due to the lack of scaling. While scaling does not impact PC performance, it is always a good idea to run your display in its native resolution. Also, 1680 x 1050 is the only supported resolution that has a similar vertical resolution (1050 on the widescreen vs. 1024 on the fullscreen) to ensure a proper test. As indicated before, there are approximately 26% more pixels on the widescreen monitor.
3dmark 05:
We used the following settings for our 3dmark 05 test:
Anti-Aliasing None
Anti-Aliasing Quality 0
Texture Filtering Optimal
Max Anisotropy 4
VS Profile 3_0
PS Profile 3_0
Force Full Precision No
Disable DST No
Disable Post-Processing No
Force Software Vertex Shader No
Color Mipmaps No
Repeat Tests Off
Fixed Framerate Off
Scores:
3dmark Score:
1280 x 1024 3dmark score: 5209
1680 x 1050 3dmark score: 4601
Charts (images):
3dmark 05 Results
3dmark 05 Framerates
With a difference of 608 3d marks we end up with about a 12% decrease in performance in the widescreen resolution. An interesting note on the widescreen test was that even though we had selected the widescreen resolution, the CPU tests (in 1680 x 1050) refused to run in widescreen and instead ran with the black bars on the sides of the screen. There still seems to have been a difference in framerate in CPU test 1 which caused a considerable difference between the CPU scores.
3dmark 06:
We used the following settings for our 3dmark 05 test:
Anti-Aliasing None
Anti-Aliasing Quality 0
Texture Filtering Optimal
Anisotropic Level 1
VS Profile 3_0
PS Profile 3_0
Force full precision No
Disable HW shadow mapping No
Disable post-processing No
Force software vertex shader No
Color mipmaps No
Force software FP filtering No
Repeat tests Off
Fixed framerate Off
Scores:
3dmark Score:
1280 x 1024 3dmark score: 3056
1680 x 1050 3dmark score: 2681
Charts (images):
3dmark 06 Results
3dmark 06 Framerates
We see a similar trend with a scant 1% increase in the difference between the resolutions. Overall the tests performed similarly and again the CPU test refused to run properly in the widescreen mode instead running with black vertical bars on both sides of the screen. This time however the CPU scores were (for all practical purposes) identical.
The Elder Scrolls IV: Oblivion
Using fraps to capture the benchmark in Oblivion which was run on max settings in both resolutions. No antialiasing was used in these tests.
Benchmarks:
1680 x 1050
Average FPS
Town: 22.417
Outside: 15.533
1280 x 1024
Average FPS
Town: 25.917
Outside: 18.383
Chart (image):
Oblivion Benchmarks
At a glance the town benchmarks appear very close but it averages out to approximately a 12% difference between the resolutions once more. The gap for outdoors is even larger at approximately 16% faster on the 4:3 monitor. One point worth mentioning is that the outdoor max framerates were very dramatically different. While it is only a difference of 8 FPS, if we talk about percentages it is a difference of a healthy 28% between the two resolutions.
Conclusion:
In these tests the average performance difference was about 12% slower on the widescreen resolutions. Although this percentage can sometimes widen by a considerable margin (up to a whopping 28%,) by and large this was not the case. Considering you are displaying 26% more pixels for your PC to process, a performance hit of only 12% seems like a bargain.
The appeal of native widescreen resolutions in games is obvious. With a larger field of view you are able to more effectively see your surroundings and make better decisions. This isn’t exclusive to a single genre but is beneficial in all games. Unfortunately we have found that this comes at the cost of system performance. Depending on what you’re playing the performance can vary but ultimately if you want to run widescreen, you’ll want to make sure your computer will adequately handle the increased resolution and plan your build accordingly. Also, the matter remains that not all games currently support widescreen and often times will “stretch” the image to take up the entire screen. This may or may not bother you, but as for me I find it very distracting and difficult to look at.
Widescreen certainly is the wave of the future and we can expect it to become even more prevalent over the next couple of years. Personally I feel that the widescreen experience enhances the gaming experience greatly even if it means enduring non-widescreen games stretched across the screen. Increasing immersion in games is something that all developers strive for and helps deliver a larger impact in a game. While running modern games like Oblivion in HD widescreen resolutions like 1920 x 1080 with all the eye candy turned on might be a little ways off for many of us, with a little tweaking we should all be able to enjoy 1680 x 1050 resolutions on many newer games with even more mainstream video cards. Even with some games refusing to hop on the 16:9 train, it makes it difficult not to want to go widescreen. |
| 
| 10 User Comment(s) • 9 root comment(s) |


Knuckles (1565) Mar 16, 2007 - 07:21 am
| Interesting article.
In the 3dmark05 fps test the frame rate drops by average 12% with the increase of resolution, but the system is pushing 19% more pixels at the higher resolution.
Also, in 3dmark06 fps test the frame rate drops by average 15%, but it's pushing 14% more pixels.
Finally, based on the average Oblivion scores, the frame rate drops 15.5% in the outside test and pixel pushing increases by 13.7%. But in the Oblivion town fps test the frame rate drops 13.5% and the pixel pushing increases by 16.4%.
So what is your take on that? Could it be that in tests (ie. Oblivion-town) where the frame rate drops less than the pixel pushing increases, that those tests are CPU bound. And for tests (ie. Oblivion-outside) that where the frame rate drops more than the pixel pushing increases are GPU bound.
There are small problems regarding the numbers.
First of all 1680x1050 resolution is 16:10 not 16:9, and 1280x1024 resolution is 5:4 not 4:3. However they are often used on that type of screens.
Second, well actually 1680x1050 is pushing 35% more pixels than 1280x1024, because the lower resolution is the 100%. You could also say that 1280x1024 is pushing 26% less pixels than 1680x1050 resolution, to move the reference point to the higher resolution.» Login to reply to this |


 Swatt (25) Mar 14, 2007 - 08:24 pm | Edited on Mar 14, 2007 - 08:33 pm
| Hey Guys, Thanks for the comments.
Ag4life, I actually took price and size into account when selecting resolutions and panels (also that my friend had the widescreen model while I own the standard model.) What I found was that for the price, these two resolutions were the most common native resolutions. With the verticle resolution being so close to the same, it made the most sense to keep the tests simple and only do these common resolutions.
Initially I had intended to run in many resolutions including comparisons in 1280 x 768 to 1024 x 768 (which would have made for a much better test) but I didn't have the time to spend working on getting the programs to run properly in these resolutions.
Trogdor, I agree that there are many other perfectly valid resolutions to choose from, however, these resolutions were the native resolutions of the panels chosen and also are very common resolutions among other panels of this size and price.
Also, and sadly, I do not have the resources to test other hardware as this is my system and the only one I have with a PCI express video card. I could have used a 9700, 5900, and 6600GT all agp cards with a 2500+ CPU, but that would have been far too involved and a relatively invalid test when it comes to new games.
I would have loved to test more games and more resolutions, but the simple fact of the matter is that I didn't have the time or resources. I am completely aware that there is room for improvement with the article, but I'm afraid that this was all I could muster under the circumstances.
I probably could have included more information about the difficulties and choices I made within the article itslef. Maybe that would have helped ;-)
Thanks again everyone for the comments.» Login to reply to this |


» Note: You need to be logged in to write a comment!Login here, or if you don't have an account with FiringSquad, register here, it's FREE! |

My Media-Blog categories
No categories created yet.
|