[SFP1 Series 1] F-4 Cockpit

[Ticket1 0]

Hi folks,

 

I'm sort of doing some kind of benchmark using different Catalyst drivers, Catalyst drivers really drive people nuts!  I've finally decided to use the 5.15.06 version for the time being because it seems to be the best in terms of better frame rates and Bunyap's 2006  weapon pack despite some game-play features might be lost, I can live with that.  In order to do the benchmark, I need to design a mission which is done.  Now I run into a little bit of trouble during run-time.  As you might know, F2 doesn't do the wide-angle cockpit view in 5.15.06 as it does in Oct 2008 version.  So I have to right mouse to zoom out manually every time I run the mission.  As I need to run it over and over to obtain different performance readings, is it possible  in this version to save the cockpit view?   Your help will be greatly appreciated, many thanks!

 

 

 

Ticket1

 

 

[Ticket1 0]

 The following videos show the difference in frame rates:

 

 

 

 

 

 

 

 

 

 

[Romflyer 25]

I'm not sure what your expectations are for frame rates, but I seriously was having troubles seeing the difference between those two film clips, as a matter fact I thought they looked pretty damn good at 500+ knots at tree level

 

My suggestion to curing your frame-rate problem...........would be to turn off the frame counter .....I seriously doubt you will still be able detect the difference between 50fps and 40fps without the counter telling you how well your computer is doing.........

 

Enjoy the game

Edited August 23, 2010 by romflyer

[Fubar512 1,350]

Sometime ago, one of the gamer sites ran an E-article on frame rates, or rather, people's perceptions of frame rates.

 

To do this, they showed three different videos to a test group. The first video showed a character running through an otherwise featureless landscape. The second showed the same character shooting at, and dodging fire from opponents. The final video sequence was that of a scorched earth scenario with huge explosions, flames, falling wreckage, and characters fleeing for their lives.

 

They then asked which had the highest frame rate. Most people selected the first (the character running through the featureless landscape. They choose wrong The vid with the fastest frame rate, was the one most people had voted the slowest, and that was the scorched earth vid (the third one).

 

They came to the conclusion that the overwhelming majority of people cannot perceive the difference between a video at 30 fps, and one at 100+.

 

The quest for insane frame rates was started by gamers who played first person-shooters, back in the DOS and Windows 9X days, when frame rates were directly tied to mouse refresh and reaction rates. That basically became a moot point around the time that games were being written for Windows XP and DX9.

Edited August 24, 2010 by Fubar512

[+streakeagle 871]

Most people can't see a difference between a 720p and a 1080p tv either. Most gamers don't know what FSAA or AF is either.

 

Find a game where you can very the frame rates from 30 to 60 to 90.

Do a barrell roll in a high roll rate aircraft with the horizon rotating at the center of the view.

If you can't tell the difference and like the 30 or 60 better than the 90, you can donate the hardware that does 90 fps to me :)

 

My acquisition of an LCD monitor a few years back capped me to 60 fps, but my hardware usually limits me to 45 to 50 fps (cpu bound).

In SF2, I can easily tell when the frame rate changes by 10 fps and can tell you the fps +/- 5 by how the aircraft handles in rolls and air-to-air gunnery.

 

At 60+ fps, my gunnery is a lot better than when it is 20-30 fps: smooth tracking of LCOS gunsight = very high accuracy.

If my fps are falling to 30 or lower, not only is the screen/motion a lot more jerky, but it means TrackIR and joystick inputs aren't being handled as well either.

 

Another game where the difference directly affects my performance is Operation Flashpoint/ArmA2 (also Star Wars Battlefront). Not even using TrackIR or a joystick, the smoothness of movement and aiming with the mouse are critical. In OFP, I can easily achieve the FPS I want most of the time (locked at 60) and can handle an M-16A2 like it is a sniper rifle using just the iron sights. I had a friend visit and play Arma2 with me. The laptop is Core I7 with 6GB of RAM, but has a relatively low end ATi laptop graphics chipset, which I run at 1080p despite its limited processing power. My Athlon 64 3800+ has only 2GB of much older PC3200 RAM, but has an ATi HD4890 1GB graphics card. My kill ratio dropped by 10% when I switched to the visually slower laptop (20 to 35 fps). My friends kill rate more than tripled when he switched to my desktop. He now understands all to clearly how much the frame rate matters and how much the graphics card affects frame rate and quality. My desktop's CPU is 2.4GHz single core against the much faster core I7 laptop, yet my desktop plays with 40 to 50 fps at much higher image quality settings (higher FSAA, better shadows, etc).

 

Just because the majority can't tell the difference between 30 fps and 60 fps doesn't mean there aren't people who can't, and it can affect gameplay dramatically. If I had the money, I would try to make all my games play at the highest possible resolution at the highest practical frame rate (2500 x 1600 and 60 or 120 Hz). Instead, I have to live with what I have and play older games or have crappy image quality if I want that kind of frame rates.

[Fubar512 1,350]

Actually Streak, I've not experienced any appreciable degradation in accuracy until the frame rates dropped into the teens.

[Fubar512 1,350]

Anyone care to estimate the frame rate on this video?

 

FPS.wmv

[+hgbn 91]

a qualified guess would be around 25-35 with a occasionally stutter in the lower teens

[Fubar512 1,350]

a qualified guess would be around 25-35 with a occasionally stutter in the lower teens

 

Nope.

 

1) No stutters

2) Guess again

[+hgbn 91]

my best bet then would be a average on 20

[Fubar512 1,350]

my best bet then would be a average on 20

 

Getting waaaay colder

[Wilches 1,172]

Bud, in your FLIGHTENGINE.INI, located inside the Flight folder, find the lines below and replace'em with this:

 

[LowDetailOption]

HorizonDistance=45000.0

DetailMeshSize=3

DetailLevel=0

WaterEffect=0

NoiseTexture=0

MaxTextureRes=1024

 

[MedDetailOption]

HorizonDistance=50000.0

DetailMeshSize=4

DetailLevel=0

WaterEffect=1

NoiseTexture=0

MaxTextureRes=2048

 

[HighDetailOption]

HorizonDistance=60000.0

DetailMeshSize=5

DetailLevel=1

WaterEffect=2

NoiseTexture=1

MaxTextureRes=0

 

[unlimitedDetailOption]

HorizonDistance=75000.0

DetailMeshSize=6

DetailLevel=1

WaterEffect=2

NoiseTexture=1

MaxTextureRes=-1

 

It'll increase the frame rate a lot!

[Fubar512 1,350]

Bud, in your FLIGHTENGINE.INI, located inside the Flight folder, find the lines below and replace'em with this:

 

[LowDetailOption]

HorizonDistance=45000.0

DetailMeshSize=3

DetailLevel=0

WaterEffect=0

NoiseTexture=0

MaxTextureRes=1024

 

[MedDetailOption]

HorizonDistance=50000.0

DetailMeshSize=4

DetailLevel=0

WaterEffect=1

NoiseTexture=0

MaxTextureRes=2048

 

[HighDetailOption]

HorizonDistance=60000.0

DetailMeshSize=5

DetailLevel=1

WaterEffect=2

NoiseTexture=1

MaxTextureRes=0

 

[unlimitedDetailOption]

HorizonDistance=75000.0

DetailMeshSize=6

DetailLevel=1

WaterEffect=2

NoiseTexture=1

MaxTextureRes=-1

 

It'll increase the frame rate a lot!

?

[Ticket1 0]

Hey folks,

 

And to Fubar:  I can tell right away it's definitely below 30FPS!  Low 20s I say!  Reasons are as you pan the views, it's like slow motion, it's not smooth as fluid at all.  I'm comparing it to H.A.W.K. and SFP1 when in medium settings.  

 

The video I've uploaded here are for your to compare side-by-side.  I do understand that using FRAP will discount some of the frames during recording so add say 5 frames to what you've seen in each of them.

 

I think I've got the answer to my original question.  2008 Oct really does the job better in terms of AI aircrafts.  I've designed a mission for testing this aspect in particular.  2008 Oct version AI tracks me while I was 80 nm away from behind whereas 2006 May version doesn't.  I take AI very seriously because of realism, therefore I think I'll have to sacrifice some visual qualities for realism.  This is not the first time, in fact, due to the limitations of my rig which is Q6600 based, ATi 4870 1GB DDR5 and Vista 64bits I can't ask for too much.  I've no problems running the simulator at medium setting with 4:3 resolution and with AA=8X, actually I managed to get over 70 FPS which is more than satisfactory.

 

The thing with the threshold frame rates I believe is accuracy in First Person Shooter games.  Experience tells me that the higher the Graphics setting in Battlefield 2, the more kills you'll get simply because at that settings, you can see farther and a slightly larger dot actually is a moving target which is meant to be killed.  Put aside the Internet connection issue, the only lagging factors would be slower CPU speed or inadequate GPU power.  Same in flight simulator.  If I'm playing a WWII flight simulator, I need even better hardware in order to get higher graphics quality which includes visual candies and the optimum frame rates.

 

Anyway...Fubar what is that simulator?  Is it LOCKON or a modded MiG-29 in one fo the Strike Fighter theatres?

 

 

 

 

Tk1

 

 

[Fubar512 1,350]

It never went below 35 FPS, and averaged 40+. As an encoded WMV file, it's a constant 30 FPS, so I've made my point. I purposely panned the view the way I did in order to deceive the viewer's eye.

 

And it's SF2, using Stary's CE Terrain and the Mirage Factory's MiG-29A.

[MigBuster 2,884]

Hey folks,

 

 2008 Oct version AI tracks me while I was 80 nm away from behind whereas 2006 May version doesn't.  I take AI very seriously because of realism, therefore I think I'll have to sacrifice some visual qualities for realism.

 

 

Not sure how its implemented in the game but remember - if a ground radar site has picked you up then GCI can just vector the bandit to your location.

[Romflyer 25]

?

 

 

I have been hoping that Wilches was gonna come back and answer Fubar's question.......can anyone offer any remarks on what changing these values will do? when I compare the changes to my install it looks like the only changes are to the MaxTextureRes= which are considerably higher values than in my install.......wouldnt these changes slow down FPS??