what the heck?!? Fear runs faster on MAX...

Why does FEAR run faster on MAX settings than it does on MEDIUM?

I did the stress test, and my machine went up by 10fps on the MAX settings.
In 1024 by 768.

I've noticed the same thing with HL2 and CS:Source. in 1280x1024.
It doesn't make any sense. The fps should go down. :?
Does ATi use some optimizations with AA and AF?

not that i'm complaining... it's just counter-intuitive.
30 answers Last reply
More about what heck fear runs faster
  1. I don't have figures to compare, but I noticed that with certain ATi cards, that when you crank the detail up in certain games it seems that you actually get better performance.

    So I can't say better FPS - but more stable FPS range. I just always had this feeling that with ATi you get better performance when you stress the card, and that's why they are the best in my eyes :D The X800 series started the evolution of high res gaming, the 1800 came closer and the 1900 is just what they aimed for, high res, high detail and eye candy and decent FPS to go. I'M VERY BIASED.... :D :D
  2. Quote:

    I did the stress test, and my machine went up by 10fps on the MAX settings.
    In 1024 by 768.

    By "MAX" I'm assuming soft shadows disabled, but hey that's great.
  3. thats the same with me in css with my x700pro, on low it gets about 70 fps in stress test and on high (no AA or af) it gets 110 fps... maybe it has to do with the cards optimum settings :P since medium is the same, runs slower than high... maybe it likes the settings it thinks are best.. i cant say the same for fear, maxxed i get abou 20 fps and on medium its closer to 40 or 50 but ive never done a test thats just my guess..
  4. Ok here it is...

    Max everything at 1024x768 (ss on)

    Bilinear filtering, 1024x768 and other settings at medium. (ss on)

  5. If I knew how to post pictures, I'd try and see if the same applies to me... :oops: :oops:

  6. When I select Maximum, I get 4aa and 16af by default - so I left it at that. Seems that for FEAR, my previous theory is invalid... :roll:

    I might be that the 1900 runs shaders on SM3.0 - although I'm not 100% sure if SM3.0 is a rendering path for FEAR...
  7. That's no suprise. ATI performs better at higher resolution. I've owned previous ATI cards like 9500 Pro 9800pro, X800XL. I just recently owned a 7800 GT and with all settings up it performs like a rabbit geting tired as you increase the res. Now I own a X1900XT and ATI's rep for running smoothly at high resolution still lives on.
  8. i wonder if thats ati's good drivers :P... which would be better at say a nice high res of 1024x768 ...
  9. You're not using soft shadows so don't kid yourself. If you click Maximum in F.E.A.R. it will say softshadows are enabled with Anti aliasing, but if you actually take a look at the game you'll see otherwise.
  10. Your scores seem to be consitant, unlike mine.
    I'll do that CS/HL2 benchmark aswell, and post my results soon.

    Wow, your 1900xtx isnt that much faster at max then my x850xt...
    i actually beat you at maximum fps :P

    It's probably because of shader 3.0.

    about the soft shadows...
    do you mean that its a 3.0 feature and my card doesn't support it?
    That could explain the great performance on MAX. It could be shader 2.0b optimizations.
  11. I did the CS source benchmark at 1280x1024.
    My cpu ran at 3.0, haven't oc-ed yet since i got the new ram and i don't have the bios cd in case something goes wrong.
    Here's the menu so you know what i ran everything on.

    AA: OFF AF: Trilinear FPS= 107.25

    AA: 2x AF: 2x FPS = 103.57

    AA: 4x AF: 4x FPS =103.65

    AA: 6x AF: 16x FPS = 101.86

    So the results were not as impressive as in FEAR.
    But, i did get something similar.
    only a 2 fps drop going from 2x to 16x.
    And it actually ran a tiny bit faster at 4x than 2x, which is what i've noticed while playing CS. :o

  12. Well, I get SS and 4xAA - look at the screen shots. Dunno if this applies to the x850, but looking at the FPS of the benchmark I'd say the possibility is good. Funny thing though, as I had to use the global setting and go to Maximum - as soon as I try and do a custom change on the graphics settings, SS would not work with AA. But using the global setting, it reverts to 4AA and 16AF with SS enabled.
  13. Now that i see your screenie, i don't think mine actually uses soft shadows.
    Here, this is pretty close to yours:

    That pixel shader 3.0 looks nice. Nice lighting effects.
  14. If you setup any game at low settings, example 1024x768. Then the game will push more of the graphics to the CPU. If you higher up the graphics to examples 1280x1024 and higher settings, then the computer will push the graphics to the graphicscard more than on 1024x768...
  15. Soft Shadows isn't part of SM3.0, just a nice feature that chows performance. i play without it as it really isn't that cool to me.

    they are basically just softening the edges of shadows with some sort of AA in multiple layers.

    Once again I'm not 100% sure of the SM3.0 spec being present in FEAR.

    and finally, i must admit that looking at graphics without AA and AF is sumthing i can't do without - and i was one of the many that argued against it sum time ago. hypocript i am :twisted:
  16. SS does make the shadows more realistic, mine just look like straight edges.
    I can't play without AA and AF aswell, i'm an eyecandy wh0re. Although, in some games it makes it actually harder to see enemies like cs.

    Hmmm, i think ATI should have increased those ROPs on the 1900,
    I get almost the same fps as you, with an 850. Thats crazy. I thought the 1900xtx can beat two 7800GTXs in FEAR?
  17. :lol: Just remember - I'm running SS + 4AA and 16AF with everything max and only on a 660 P4. I think it's quite good...

    Where I'm now:

    Where you are now (I think):

    You should check the difference between the X1800XT and the X1900XT in terms of SS, ATi included an instruction Fecth4, for the purpose of rendering Soft Shadows almost as fast as normal shadows.
  18. ohh so the second one is your card without SS?
    Very impressive. :wink:
  19. Thankyou.

    Although I realize now that this might seem like I'm bragging about my card, something I explicitly try to avoid. :oops:

    The second one with my name in it (stupid addition :roll: ) is how you can compare yours with mine. 1024x768 4AA and 16AF.

    I know my friend has a X850XT - and that card can give me a good run for money in terms of performance on games where the tech is equal.
  20. Too bad the softshadows still aren't being rendered correctly. Don't believe me? Shine your flashlight at any shadow in the game and watch it sharpen
  21. One day, when I'm really bored - I'll shine a light on a shadow and watch it disappear...

    What does Soft Shadows in FEAR actually do? Ummmm, it takes multiple light source angles into consideration - so it applies shadows according to different angles as one would get in real life from multiple light sources as opposed to a simple shadow form the closet light source or a generalized light source.

    So, when you shine a torch onto a shadow, then guess what? You kill ambient lighting and the differing shadows and yes - the shadow will revert to a simple one. Simple, innit?
  22. I'd say you're correct but this is not what F.E.A.R. does. It is very nice if you can run F.E.A.R. with softshadows and Antialiasing enabled, however the game currently isn't suppose to utilize both. Did you not question the pixilated edges of the shadows when you shine the flashlight on them? Last I checked the real world doesn't do that.
  23. first off, i apologize for my sarcastic reaction. just seems as if you're after my blood or sumthing... :D

    yes the quality does lack pure quality, but the bottom line is that AA and SS is running together. the ss quality looked the same on my 1800xl 2months ago (minus AA), so hopefully it's sumthing that can be resolved any time soon (patch/driver).

    SS is nice, but in my personal opinion a moot and resource hogg feature. give me bump mapping, good filtering and decent shaders, and i'm happy!
  24. I'm sorry if you thought that; my point is F.E.A.R. can not render softshadows correctly with antialiasing. Softshadows in general are VERY nice, however F.E.A.R.'s are lacking.
  25. Point taken - SS in Fear right now does not look all that cool (pixelated on the edges) with AA. But I have yet to see it function properly, with or without AA.

    Is shadows really such a big issue these days? I'm just asking, as it marginally adds to the game (in my view). I mean, Doom3 relied heavily on shadows and it killed the hardware that was out at that time minus top Nvidia cards, and for what? It was such a dark game, that the shadows rarely had any usefullness (once again in my opinion), so why make it so resource hungry?

    But okay, just like I was someone querying the usefullness of AA and AF a few years ago and can't live without it now, I'll probably be blowing my horn forwards on SS. :D

    HDr, now that is sumthing that blows my mind at this stage and worth scrutiny and discussion. :D

    Once again, sorry the sarcasm!
  26. Yeah, I love F.E.A.R., but there's something that bothers me. Have you ever noticed your flashlight does not cast a shadow? Doom3's flashlight does...
  27. Actually I haven't. Probably sum sort of ultra xray light :D I must say, FEAR was one of the few FPS that could come close to HL2 (in my view). Very good. Let's hope STALKER is as good, if not better.

    will check it sum time, after I've taken over San Andreas... which may take me a year or so...
  28. Right, ultra xray light...
  29. just joking! xray, as in shines through objects without casting a shadow. are you using patch 1.2 and 1.3?
  30. 1.3
Ask a new question

Read More

Radeon Stress Test Graphics