Quake 3 Videocard Shootout

[Broken]

Check this article on Thresh's Firing Squad.
http://www.firingsquad.com/hardware/...up/default.asp

What's you people's call on this one?

[YoKoMo [1]]

First of all let me say I think Quake3 sucks compared to UT. As far as the benchmarks go, Who cares. Matrox will soon have new cards out there to compete with a Gfarce and Wiper II. If your games play fine to you, then the card is fine. I will concede it's not the fastest thing out there, but it sure is the prettiest.

[Broken]

Nowhere in the article did they mention the quality of the image. It's kind of sad, when you think about it. Here we're sitting with the most beautiful looking card, and all everything is about is framerates. My Q3 runs at 36 FPS, with full detail, and I'm happy. Happy in the knowledge that it when it comes to looks, it's way ahead of the rest.

One thing that I did find strange was the fact that they didn't just compare card of the same generation. It's not really fair to run benchmarks witht the GeForce, but then again, I suppose everyone has his bais...

Well, anxiously awaiting the G450, and then the G800. Bring it on!

[This message has been edited by Broken (edited 15 March 2000).]

[terence]

Well said Broken. I've read somewhere that beyond 30-35fps the human eye can't really notice much change. To me its all about quality so long as basic fps requirements are met. I get about 40fps in UT on my Athlon/G400 system at 1024xx768 32 bit, high detail and I'm more then happy with it. The picture quality is outstanding. I may be able to get about a 5fps increase in a TNT2 card, but the drop in picture quality isn't worth it.

[dZeus]

about that 35fps: That's bull! Only with non-interactive things you won't see much difference between 30fps and 50fps (and that even depends from person to person). When a scene gets interactive (a pc game for example), you will easily notice the difference between 30fps and 60fps. The magical limit would be 75Hz, the refreshrate of most monitors. After that it won't be really usefull to have higher FPS unless average monitor technology gets to a higher level.

And about Q3: If you want to buy a card only for Q3A, buy a GeForce or Savage2K. If you want to buy a good all-round card, buy a G400. If you have a large, good quality monitor, you definitely don't want a GeFarce! 2D image quality is horrible compared to the G400. Even at 1280x1024 everyone can see the difference.

[Belshy]

It is a shame, this obsession with Frames Per Second, rather than the overall picture of speed combined with quality.

There was a recent article that did a group test on Geforce, Voodoo 3, TNT 2, the Savage card , and Matrox running Q3A, at Tweak3D's site.

They mentioned all of the cards in depth when they did the FPS bench test. However, when they did the image-quality argument they compared TNT2, Geforce and Savage but left out the Matrox G400. It's one chance to shine above the others....typical sloppy reporting.

I looked at the screenies of Q3A on these other cards and shook my head. NONE of them can match G400 for image quality in Q3A, with colour or textures.

People should test these cards properly and fairly. I get 39.9 fps out of my G400max set-up in Q3A 1024, 32 bit, bilinear, vertex, everything else on high.

My brother gets 44 fps on a virtually identical machine but with a standard Creative Geforce instead.

Just a few frames in it....but Geforce image quality doesn't really compare to Matrox, so I feel I made a good choice with this card.

[Belshy]

It is a shame, this obsession with Frames Per Second, rather than the overall picture of speed combined with quality.

There was a recent article that did a group test on Geforce, Voodoo 3, TNT 2, the Savage card , and Matrox running Q3A, at Tweak3D's site.

They mentioned all of the cards in depth when they did the FPS bench test. However, when they did the image-quality argument they compared TNT2, Geforce and Savage but left out the Matrox G400. It's one chance to shine above the others....typical sloppy reporting.

I looked at the screenies of Q3A on these other cards and shook my head. NONE of them can match G400 for image quality in Q3A, with colour or textures.

People should test these cards properly and fairly. I get 39.9 fps out of my G400max set-up in Q3A 1024, 32 bit, bilinear, vertex, everything else on high.

My brother gets 44 fps on a virtually identical machine but with a standard Creative Geforce instead.

Just a few frames in it....but Geforce image quality doesn't really compare to Matrox, so I feel I made a good choice with this card.

------------------
P3 450, 128 mb ram, Creative PCI 16, Intel SE440BX2, G400 Max, Viewsonic E771.

[Belshy]

Btw, their scores are inexplicably poor on the G400 Max IMHO.

Did they just run it with the original Open GL drivers for the G400 off the Q3A disc, with the G400 card's original ICD drivers?

Anyone who knows a little bit about the Matrox G400 series will know that the G400 Open Gl drivers are more or less a "work in progress". The scores should be WILDLY different with the G400 running 5.52 and the latest Turbo Gl. Check out
http://www.twicks.dircon.co.uk/matrox/

Now THAT'S more like it :0)

------------------
P3 450, 128 mb ram, Creative PCI 16, Intel SE440BX2, G400 Max, Viewsonic E771.

[Pace]

Nice to see Thresh's guy used TurboGL - not! I usually like Firing Squad but this review wa slightly disappointing.

Quake 3 doesn't suck compared to UT - UT has more gameplay options out of the box but Quake 3 will get these soon enough. Quake 3 has a better base for building on - it is a better standalone deathmatch game so when addons like Assault come Q3's way I'll probably be playing Q3 a lot more. UT may be better right now - but not by much and not for long - I'll play either though.

Back to topic then, the benchmarks are worth caring about if you are looking for a new card today - but I agree that they could have talked more about everything else that a Matrox card, and in particular the G400 MAX, offers. -- 2D, DH, VCQ, EMBM, etc.

I don't believe in supporting unreleased cards - the G450 will probably be in my system when it comes out as will the G800. However, saying that Matrox's new cards will be out there to compete with the GeForce is unlikely - the GeForce II (NV15/11) may well be out by then - it is a guessing game and the only facts there are is that the GF is the fastest at the moment - and the G400 is the best looking and is marginally slower (when using TGL at least).

My games do play fine at the moment (just) but I would rather be playing at 1600x1200x32 than SVGA and XGA resolutions - not even the GF lets me play UXGA acceptably - I'll wait for whatever offers me that (though preferably from Matrox)

I also don't think that it is biased to compare the G400 with GeForce - that is today's playing field and any comparisons drawn are fair.

And now to agree with someone - UT and Q3 do look much better on a Matrox than on any other card and I don't care about another few frames per second when the game looks this good.

Bring on the G800 indeed!

Paul.

-- Can't believe 4 posts were made while I typed this first time around!

-- So, somebody agrees with me about G400 being an allrounder

-- They don't mention TGL so I presumed it was the Std ICD

[This message has been edited by Pace (edited 15 March 2000).]

[Gurm]

As far as the whole "frames per second" argument - those who say that we can see 75 distinct frames per second need an anatomy lesson. We can distinguish INSTANTANEOUS MOTION at 75fps - meaning if something is in our vision for 1/75 of a second we will register it as motion. HOWEVER... we cannot see 75 distinct frames per second... meaning if something were in one place for 1/75 second and another place for the next 1/75 second we would register it as one blop... we can only pick out distinct images at roughly 30fps, depending on the person's individual biology.

Hope that clears things up. So yes you can "see" a difference from 30fps to 60fps, but you aren't getting MORE information into your head. Sorry.

- Gurm

------------------
Listen up, you primitive screwheads! See this? This is my BOOMSTICK! Etc. etc.

[bobby2]

30 vs 60 fps issues are important for gameplay. When a big firefight takes place with many people the framerate on a '40' fps card can drop to 20. Whereas a 80 fps card might drop to 40 ....

This is my biggest caveat with gameplay online - I use 640x480 low quality cos I cant work at low framerates - I get killed lots.

The place I notice this most in Q3A is when Im railing - I can follow one person in a 1on1, but I lose that person due to frame drops when there are many people bombimg away. This gets worse when Im 'zooming' for my rail kill too. Thus 640x480 LQ is req'd.

The G400 aint the greatest game card - I dont make money playing games - so I got a G400 for doing my paid work (dual monitors and graphics and quality ...) - i use a machine, this one Im typing on, at work which has a TNT2 - the visuals are pants .... Its irritating not having the stuff I want at both places. But then, I earn more from my home PCs.

The G450 will pull Matrox back level, and the G800 may do just the same for its time. But Id rather see Matrox be all purpose rather than gaming only.

bert

[Pace]

Yes, bert! I agree - also, when a demo is run the CPU etc doesn't have to do other tasks which would be running in a normal game - e.g. networking, AI etc.

Paul.

[Rohde]

Saying that 30 FPS is just as good as 60 FPS is BULLS***.

Yes, it's hard (or even impossible) to tell 30 FPS from 60 FPS visually, but if you deathmatch someone who is as good as yourself and he has 60 FPS (and you're stuck with 30) he will WHOOP your ass real bad !!!!

Try taking down a nightmare bot in Q3 with the railgun with 30 FPS.

Image Quality is nice, but only if the FPS allows it. If you want something nice to look at, go ahead an play with 30 FPS, but if you actually want to frag someone, you'll be better of with lower res./detail and a higher FPS.

Ok, I'm done now.

Torben R.

BTW.: Quake3 is way better than UT, it simply has a better "feel".

[This message has been edited by Rohde (edited 15 March 2000).]

[rei]

rohde:

"Yes, it's hard (or even impossible) to tell 30 FPS from 60 FPS visually, but if you deathmatch someone who is as good as yourself and he has 60 FPS (and you're stuck with 30) he will WHOOP your ass real bad !!!"

you forgot those of us blessed with mad skillz

[rjcarr]

You guys are forgetting the who's article you are reading. This is Thresh's site, and he is about the best fps player ever. Do you think he got good by gazing at the walls and eye candy in the game when he plays? No, he probably has an 800+ MHz Athlon with a GeForce and still plays at 640x480 wiht 16 bit everything and low quaility textures because that way he can pull over 80 fps and yes ... 60 is extremely noticeable over 30 (but anything over 60 is unnecessary (take a look at a physiology book)). Thresh cares about frame rates, period, he isn't going to ever be interested in image quality, so read a different review to get that.

-joe