Welcome back! I have been playing Battlefield 3 for the last few weeks and i have to say my expectations have been shattered. Battlefield 3 lived up to everything i thought it would be and more. Since i started this blog, my plans for it have been constantly changing based on my experiences, talking with Dr. Palmer, and a request from a commenter on my last post. Originally this post was going to be on other aspects of the game but i feel benchmarking Battlefield 3 in terms of its graphics is better suited.
Since
my last post i have since upgraded my video card in my computer to a Nvidia
GTX570, a much better card to handle BF3. It worked in my favor because i now
get to compare my old card to my new card, on low and high settings for my
benchmark tests. This will hopefully show what having better, more up-to-date
hardware, can do for your PC gaming. So for this post i will be...
- Conducting 4 benchmark tests. Benchmarking Battlefield
3 with...
- GTX275 on Lowest Settings
- GTX275 on Highest Settings
- GTX570 on Lowest Settings
- GTX570 on Highest Settings
- Giving a quick explanation of Anti-Aliasing, what it
is, and what it does.
- Conducting a few benchmarks that demonstrate the effects on frame rates when changing the Anti-Aliasing setting in Battlefield 3. (As inspired by nathangsxd in my last post. Thanks!
***Below are my results continued from my last post.***
NOTE: Before i start, I wanted to quickly go over something that
was brought up to me in a comment from nathangsxd on my last post. He said,
"are you sure that fraps maxes at at 60fps? I'm pretty sure that it can
record higher frame rates than that...". This is a great comment that i
should address. On the Fraps.com FAQ, it states, "You can choose the frame
rate of the output movie from 25fps, 30fps, 50fps, 60fps, or a custom value up
to 100fps." I didn't realize you could input a custom frame rate higher
than 60fps when i started doing my tests.
So during all my tests, this is the
reason the benchmark results are limited to 60fps. When i do the benchmarking
without recording, my numbers are much higher, hitting well over 100fps in BF3
on the highest settings with my 570.
While this isn't representative of
the computer’s ability to play the game, all my tests have this limitation,
making them accurate in showing frame rate changes based on hardware and
settings. Also remember that these tests are conducted while recording full
1080P video so the frame rates are lower than i get during normal gameplay.
Each
section below shows the results of my 4 benchmark tests and an explanation of
my results are below the images
Prediction
Finally getting to play the full
retail game on my own computer, i am able to see how my PC handles it. When i
had my GTX275, i couldn't play the game on the highest settings but could
easily play it on the lowest. With my GTX570, i can play the game on the
highest settings without any graphics lag or visual tearing.
I
predict that my frame rates will be representative of my experiences, with my
GTX275 having lower frame rates than my GTX570. The question is how will the
275 perform on lowest settings compared to the 570 on the highest settings.
Look below to find out...
Battlefield
3: GTX275 on Lowest Settings
Results
With
my GTX275 on the lowest settings i was able to play the game without any
issues. The average frame rate was about 42.4 frames per second, about 18
frames less than my results with Battlefield 2 on he highest settings with the
same graphics card. The CPU temperature topped out around 42C and the graphics
card topped out around 67C.
Battlefield
3: GTX275 on Highest Settings
Results
With my GTX275 on the highest settings the game was
barely playable at all. I had a very difficult time flying the jet, as you can
see in the video. The average frame rate was about 26.8 frames per second, much
less than my results on the lowest settings with the same graphics card. The
CPU temperature topped out around the same, 42C and the graphics card topped
out a little higher around 74C.
Results
With
my GTX570 on the lowest settings the game performed perfectly, no issues at
all. The average frame rate was about 42.7 frames per second, interestingly
about the same as the GTX275 on the same settings. The CPU temperature topped
out around 45C, 3 degrees higher than the 275 on both settings, and the
graphics card topped out at 84C, much higher than the 275 with both settings.
Battlefield
3: GTX570 on Highest Settings
Results
With
my GTX570 on the highest settings the game again performed perfectly, no issues
at all and i have to say that the game looks absolutely incredible. The video
doesn't do it justice. The average frame rate was about 41.7 frames per second,
1 frame less than the same card on the lowest settings. The CPU temperature
topped out around 47C, 2 degrees higher than the same card on the lowest settings,
and the graphics card topped out at 84C, the same as the low settings with this
card.
Summary
of Results
Overall, my results matched my
predictions, except for both the average frame rates of my GTX570. I expected
them to be higher than that of my GTX275 on the lowest settings. What i think
is happening with my results is fraps is limiting the performance of my
computer due to the recording of the video while benchmarking. In hind-sight,
this may have been a mistake on my part and i would have only done the
benchmarking without the video recording. That would have yielded more accurate
results and a better representation of my computers ability to play the
game. The tempetures for both my GPU and CPU during the 570 tests are higher
than that of the 275 for two reasons:
1). With my 275, the cooler design
is different. It vents out the back of the computer and is a closed system. My
570 on the other hand isan open system and vents into the computer itself,
resulting in slightly higher temps for both my GPU and my CPU.
2). I had my 275 fan set to 90%
speed at all times, helping cool the card better than my 570, which had a fan
speed that was automatically controlled by the computer based on
temperature.
Anti-Aliasing: Effects on Performance
Like i said, i will be
conducting a small, final test which was inspired from a comment on my last
post by nathangsxd in which he said "Could you explain what settings were
used in the actual game and maybe show comparisons of their results on
performance? I know settings like anti-aliasing can be a pretty big performance
hit when set to a high value". This inspired me to see the effects on the
game when changing that setting. For those of you that don't know what
Anti-Aliasing or AA is, here is a video that details it very nicely and it
explains it better than i could:
Below
are 4 pictures. 2 show the settings being used, 1 is a screenshot with no AA,
and 1 is a screenshot with AA. They have a yellow number in the top left which
is showing the current FPS.
Like
the video said, at my resolution it is almost impossible to see a
difference with AA on or off. But, with AA on i was getting 60 FPS and
with it off i was getting 80 FPS. I increased my frame rate by 20FPS with a
minimal visual impact. Thanks for the comment nathangsxd!
Very interesting, I never understood AA myself, as i noticed no difference either XD, but thanks to this I now know it does make minimal difference. I like that You compared two different pieces of hardware, I remember when I bumped up my graphics card and how the jump was for my games at the time.
ReplyDeleteI concur that having the video recording also playing in the background can skew your result. You should try what I did next time on my blog, I tok out a video recorder from the library and setup the tripod in front of my display and recorded that way, might help :).
Also perhaps more experimenting with the different settings might be a way to go. See if some of the other settings make little to no difference to the quality while increasing your FPS.
Ya, i never really understood AA either until now.
ReplyDeleteI hear you and i think that's what i may do. For my next blog post, since I'm unsure what to do exactly, i may do a very detailed analysis of each setting in the game and its effects on performance and the FPS. thanks for the tips!
liked this post a lot. when you ran the benchmarking for each setting was it only running during actual gameplay or did you start it when you loaded the game? i know static menus can skew benchmarking results. it is really weird that you got the same average frame rates although i too have had issues with fraps slowing everything down. also its cool you experimented with AA as well, usually ill turn it off for the speed boost although it does make things look a little smoother.
ReplyDeleteWhen i did the benchmarking, it was in an online game sitting in he same position not moving. No game menus.
ReplyDeleteYa, AA is optional for me now. I used to have somewhat of an addiction to have every setting on high. lol. but i'm starting to see that some settings are worth not having on, like motion blur, i find that i play better when i can turn fast and see a guy right away. and AA makes my frames skyrocket when its not on or on a lower setting so i try that now. obviously I play better when i'm getting higher frames.