|
Introduction
Within
the options.xml file there is an intriguing "XBOX MODE"
statement, that is set to FALSE. Some gamers have claimed
that setting the XBox mode to TRUE improves the frame rate,
one modder even included it in the Readme file for his mod.
On the other hand, Red Storm have
stated that the variable was never implemented, altering XBOX
can have no effect. The exact quote is
"The value read in has absolutely no
effect. It is never assigned to a variable in the program.
The only thing we can surmise is that someone edits the
file, but saves it in a format that the program can't read.
When that happens, everything reverts to the default values,
and a clean version of the file is written out, thereby
removing all previous options settings." Jeff
RSE
Testing the Theory
Undeterred some gamers ever hungry
for more FPS, some went about testing their theory. Great
Gold was one of the first to publish his results, which are
summarised below.
| |
Average FPS
|
Highest FPS
|
|
|
Xbox False
|
Xbox True
|
Xbox False
|
Xbox True
|
| Graphics Low |
14
|
18
|
22
|
32
|
| Graphics Med |
10
|
15
|
22
|
32
|
| Graphics High |
8
|
10
|
22
|
32
|
The full test results including the test
specification and method are listed at the foot of the article.
Conclusion
What the test set out to do was achieve
a set of results that could be replicated, under a controlled
set of variables. As far as is possible within a game, I think
this was achieved, and the results show improvements in Frame
Rates, across the board, significantly in most cases.
What this test does not take into account
is the Placebo
Effect, correctly pointed out in the GR.com
discussion. In order to carry out a test where the results
are not subconsciously weighed in favour of the testers preferred
results, the test would have to be carried out "blind".
That would require having two options.xml files, identical
apart from the XBOX value, one would be set at TRUE, the other
at FALSE. Someone other then the tester would then select
one of the options.xml files and place it in the GR folder.
Then the tester could launch GR and start recording results.
That's a fair bit of organising for something
that Red Storm have categorically stated has no effect on
the game at all. Nevertheless, I am sure some determined Ghost
Recon fan will get down to business, if you do, mail
me the full results - preferably using the same system
shown here, and I'll append it to this report.
Full Testing and Results Information
Test Machine Specs
500Mhz Celeron, 128 RAM, no video card other than the generic
Compaq Presario chipset, 12X DVD, basically your average calculator.
The Test
For the all tests, the computer was
rebooted each time. Ghost Recon was started directly, with
no other applications opened first. A clean installation with
no mods activated was used. All tests were run in the Training
Facility and lasted for approximately 5 minutes.
Full Results
TEST 1 WITH XBOX=FALSE:
ALL GRAPHICS SETTINGS ON LOW at 1024x768x16
with compressed textures. No shadows.
Approximately 14 FPS when averaged. The high was 22 FPS when
looking at the sky, to a low of 8 FPS when running around
like a crazed maniac and firing.
ALL GRAPHICS SETTINGS ON MEDIUM at 1024x768x16
with compressed textures. No shadows.
Approximately 10 FPS when averaged. The high was 22 FPS when
looking at the sky, to a low of 6 FPS when running around
like a crazed maniac and firing.
ALL GRAPHICS SETTINGS ON HIGH at 1024x768x16
with compressed textures. No shadows.
Approximately 8 FPS when averaged. The high was 22 FPS when
looking at the sky, to a low of 2 FPS when running around
like a crazed maniac and firing.
TEST TWO WITH XBOX=TRUE:
ALL GRAPHICS SETTINGS ON LOW at 1024x768x16
with compressed textures. No shadows.
Approximately 18 FPS when averaged. The high was 32 FPS when
looking at the sky, to a low of 8 FPS when running around
like a crazed maniac and firing.
ALL GRAPHICS SETTINGS ON MEDIUM at 1024x768x16
with compressed textures. No shadows.
Approximately 15 FPS when averaged. The high was 32 FPS when
looking at the sky, to a low of 6 FPS when running around
like a crazed maniac and firing.
ALL GRAPHICS SETTINGS ON HIGH at 1024x768x16
with compressed textures. No shadows.
Approximately 10 FPS when averaged. The high was 32 FPS when
looking at the sky, to a low of 2 FPS when running around
like a crazed maniac and firing.
|