7800 Owners

[trinibwoy]

-------------------------
The Project - from Crytek
Benchmark Result
-------------------------

Average FPS: 63.84

Configuration

Resolution: 1024x768
Bit per pixed: 32
FSAA: 4x samples
3Dc Texture Compression: enabled
Verticle Sync: disabled

7800GTX 480/1300 HQ - Gamma - TSAA - 78.03

Low ~ 34, High ~ 110. Disabling 3DC or TSAA only made about a 2fps difference.

 

[ArchAngel777]

Originally posted by: Frackal
Note, while some parts were impressive (graphics of the creature's vision) and the character models looked good, I bet with more work they could have made a far more graphically impressive vid for the level of stress it requires...

ie, Lost Coast will likely look way, way better and run 1.5x + as fast IMO

What are you smoking? I want some. Actually, I don't otherwise I would make claims like this above.

This tech demo is nothing short of impressive. You have absoutely no proof behind your theory of HL2 rendering better than Far Cry. This thread isn't going to be hi-jacked, but I could not resist replying to this.

Source and CryEngine are incredable in everything they do... They are both on par, easily. Most people just blindly assume Source is better because Valve is more well known. That is the case with many things... But, I tell you... Put asside BIAS and look at the game and what it does and can do. Far Cry has done both extremely detailed outdoor environments with extreme draw distances. It has also done great indoor rendering as well... I have never seen HL2 do outdoor rendering with even close to the detail of Far Cry.

Anyway, all that asside, this Tech Demo is awesome and it looks great. Shows that Cry-Tek is doing all they can to improve graphics.

 

[Tsosczb]

The future of PC gaming
Hollywood style

Oh brother...


Well anyway, this certainly was somewhat of an eye opening experience. Especially considering I just got a decent card recently (had a GeForce2), so I haven't played any of the newer games yet.

I have a 6800LE unlocked, and consistently got an average of 22-22.5 FPS. One thing was that I was getting lots of artifacts, so I first disabled the unlocked pipelines, but all I got was an average of 17 FPS with all the same artifacts. So next I re-enabled the pipelines and disabled the unlocked vertex processors, and the artifacts disappeared, but I still got exactly the same performance as with everything unlocked.

It turns out only one of the vertex processors was causing the artifacts (which I didn't get in 3DMark'03). But for whatever reason, having the extra vertex processors didn't affect performnce in the least. And I never saw any artifacts before running this, so I could still use the extra vertex processor in other games without whatever graphics that were used to get the artifacts.


And in the demo, is it supposed to have huge blurry areas during some of the scenes? The worst part is when those mercenary people are seen talking on the dock, and the guy yelling is just blurry and freaking hell. I even tried disabling all the filters and optimizations in the NVidia control panel, but it looked the same in every single time I ran it (of course losing the filters made the quality go down a bit, but the average FPS went up to just over 29 FPS).

Plus, the lighting in it is amazing. I'm guessing it has mostly to do with the HDR stuff, but I didn't even have to turn the gamma level up to 2-3 times normal to see the dark areas clearly (which is always a bitch).

 

[Polish3d]

Originally posted by: ArchAngel777

Originally posted by: Frackal
Note, while some parts were impressive (graphics of the creature's vision) and the character models looked good, I bet with more work they could have made a far more graphically impressive vid for the level of stress it requires...

ie, Lost Coast will likely look way, way better and run 1.5x + as fast IMO

What are you smoking? I want some. Actually, I don't otherwise I would make claims like this above.

This tech demo is nothing short of impressive. You have absoutely no proof behind your theory of HL2 rendering better than Far Cry. This thread isn't going to be hi-jacked, but I could not resist replying to this.

Source and CryEngine are incredable in everything they do... They are both on par, easily. Most people just blindly assume Source is better because Valve is more well known. That is the case with many things... But, I tell you... Put asside BIAS and look at the game and what it does and can do. Far Cry has done both extremely detailed outdoor environments with extreme draw distances. It has also done great indoor rendering as well... I have never seen HL2 do outdoor rendering with even close to the detail of Far Cry.

Anyway, all that asside, this Tech Demo is awesome and it looks great. Shows that Cry-Tek is doing all they can to improve graphics.

I don't smoke anything but if you do it may be impairing your spelling, and your reading.

I said nothing about which engine renders outdoor environments better or whatever other crap you applied to me that I didn't say.

I said this:


Some parts of the demo look nice, other parts, for the stress it places on the VPU, aren't that good, such as the wall textures and such...too cartoon-like.

I think they could do more with it than they did.

I also said:

HL2 Lost Coast will almost definitely look better than this demo overall IMO, and I suspect I will get higher FPS at same resolution and AA settings than I do on this demo.

 

[Acanthus]

I hate tests like this...

Let me guess, no dynamic sound, no physics, no AI.

It doesnt really represent any real performance in a real game.

 

[Polish3d]

Yeah its just a movie, you cant move around

 

[Powermoloch]

Originally posted by: Acanthus
I hate tests like this...

Let me guess, no dynamic sound, no physics, no AI.

It doesnt really represent any real performance in a real game.

Lol it's the same as 3dmark05 or any marks around. but that doesn't have to be bad

 

[Ackmed]

Originally posted by: Acanthus
I hate tests like this...

Let me guess, no dynamic sound, no physics, no AI.

It doesnt really represent any real performance in a real game.

Timedemos from games dont have all that either. Which is why I like the idea of fraps a lot.

 

[Powermoloch]

Min, Max, Avg
20, 68, 44.723

 

[Acanthus]

Originally posted by: Ackmed

Originally posted by: Acanthus
I hate tests like this...

Let me guess, no dynamic sound, no physics, no AI.

It doesnt really represent any real performance in a real game.

Timedemos from games dont have all that either. Which is why I like the idea of fraps a lot.

I completely agree. Fraps shows real world performance in games.

 

[xtknight]

Originally posted by: Acanthus
I completely agree. Fraps shows real world performance in games.

Unfortunately, it's inefficient and a resource hog itself. I like UT2004's built-in benchmarking utility. That's the best IMO.

 

[trinibwoy]

Originally posted by: Acanthus
I hate tests like this...

Let me guess, no dynamic sound, no physics, no AI.

It doesnt really represent any real performance in a real game.

None of that matters if you are evaluating the performance of different graphics hardware on the same system as is done in nearly every review. It is no different to 3dmark/aquamark or any of the other synthetic benchmarks.

 

[Ackmed]

Originally posted by: xtknight

Originally posted by: Acanthus
I completely agree. Fraps shows real world performance in games.

Unfortunately, it's inefficient and a resource hog itself. I like UT2004's built-in benchmarking utility. That's the best IMO.

Resource hog? Not hardly. Barely over 3megs in processes. Inefficient? Nope.

 

[xtknight]

Originally posted by: Ackmed

Originally posted by: xtknight

Originally posted by: Acanthus
I completely agree. Fraps shows real world performance in games.

Unfortunately, it's inefficient and a resource hog itself. I like UT2004's built-in benchmarking utility. That's the best IMO.

Resource hog? Not hardly. Barely over 3megs in processes. Inefficient? Nope.

Lots of game slowdowns reported? Yep... unless they were talking strictly about the recording feature, but that would be too obvious. I've heard of FRAPS slowing down SC3 in particular. The in-game measurement is always the most accurate. There's no disputing that.

 

[T101]

Average FPS: 55.22

Configuration

Resolution: 1024x768
Bit per pixed: 32
FSAA: 4x samples
3Dc Texture Compression: disabled
Verticle Sync: disabled


This is with my Point of View 7800GT 256MB card (need to update the profile), Forceware 80.94 drivers. However, the demo stutters intermittently, and the water is just black.