Ok Murcers GIVE UP!

[Nuno]

I still couldn´t find a single game/demo that was unplayable with my G400 (oc´ed to 150/200). I have it since I had a celeron 450A and it scaled up well with every cpu upgrade. Even going from a K7 classic 700 to a K7 tb 900 gave me a nice performance boost.

And with a 17´´ monitor I really don´t want more than 800x600 / 1024x768 res. And a custom 960x720 is very decent on a 17´´ and increases performance over 1024.

[Himself]

<font face="Verdana, Arial, Helvetica" size="2">Originally posted by Gurm:
See, there's the kind of statement that bugs me.

I'm into games. I'm REALLY into games. And I have a G400.

I've been recommending G400's to people for a while now, even GAMERS. Because it's the highest quality card available. And it's pretty darn cheap now.
</font>

Don't mean to bug anybody, just stating it like it is as I see it.

Here are a few of my reasons for not recommending the G400 to people who want to upgrade for the sake of games.

1. They have an ancient card and they are very price / future proof conscious. I'd say a somewhat new retail MX 32MB at $143 CDN beats an ancient G400 oem 32MB SH at $168 CDN (down from $238 when it was introduced over a year ago).

2. They already have a TNT 2 or similar, they wouldn't get a speed increase or any substantially different results from using a Matrox card for playing games. Minor speed hit most likely, which is hard to sell.

3. They have a much faster card already and a slower card is definitely not what they are looking for. If 30fps were what they would accept, they would have a slower card already.

4. They find the quality acceptable to them as it is and they can't be convinced quality can be improved further to any substantial degree. Being told you are drunk or seeing things is no fun.

5. Drivers. Driver support is about bug fixes only, with a constant list of impending fixes to address, speed tends to go in the negative direction if it changes at all. Nice if you are a business user, rather disappointing for gamers when they read and hear about all the after purchase speed enhancements and feature additions in drivers being offered by other companies or their current brand even.

6. Nobody would listen to me even if I did, Matrox has negative gamer name recognition, people would think I was mad if I even suggested it.



[This message has been edited by Himself (edited 17 January 2001).]

[Jorden]

Coming from 5, Himself, we go to 6: Did you ever try 5?

[DuRaNgO]

"2 They already have a TNT 2 or similar, they wouldn't get a speed increase or any substantially different results from using a Matrox card for playing games. Minor speed hit most likely, which is hard to sell."

Actually, the G400 BEATS the tnt2 and the G400Max BEATS the tnt2ultra

"5 Drivers. Driver support is about bug fixes only, with a constant list of impending fixes to address, speed tends to go in the negative direction if it changes at all. Nice if you are a business user, rather disappointing for gamers when they read and hear about all the after purchase speed enhancements and feature additions in drivers being offered by other companies or their current brand even."

that sounds more like nvidea than it does MAtrox, don't you think?

"6"
there is no 6

[High_Jumbllama]

<font face="Verdana, Arial, Helvetica" size="2">Originally posted by Himself:


5. Drivers. Driver support is about bug fixes only, with a constant list of impending fixes to address, speed tends to go in the negative direction if it changes at all. Nice if you are a business user, rather disappointing for gamers when they read and hear about all the after purchase speed enhancements and feature additions in drivers being offered by other companies or
</font>

No Speed increases?! Me thinks himself did not feel the upgrade from Powerdesk/Drivers 5.41 to 5.52/5.55. My speed increased at least 30% (Athlon 600 128mB) I uninstalled the new drivers and went back to the old for a few minutes because I was so astounded and to make sure it was not my imagination.

[This message has been edited by High_Jumbllama (edited 17 January 2001).]

[Liquid Snake]

How many times have we discussed this? What's wrong with 30fps??? Is it neccesary to get 500fps? I can run CS, UT, TFC all at >30fps in 1024x768. I'm happy. Sure, it would be nice if it was a little faster, but for now, I'm happy. The only game I've played that could tax my G400 was Deus Ex. I had 20fps in large outdoor areas.

[Himself]

Jord,

Yeah, yeah, blame the one inch high edit window online forums like this give you.

[Himself]

<font face="Verdana, Arial, Helvetica" size="2">Originally posted by DuRaNgO:
"2 They already have a TNT 2 or similar, they wouldn't get a speed increase or any substantially different results from using a Matrox card for playing games. Minor speed hit most likely, which is hard to sell."

Actually, the G400 BEATS the tnt2 and the G400Max BEATS the tnt2ultra
</font>

In what? No, take that back, who cares? I just happened to notice 3dmark2K scores comparing the two when I was looking up scores for my cpu, the TNT 2 scores for faster cpus were much higher than G400 ones with the same cpu. Like it or not, Q3A and 3DMark2K scores count with people.

<font face="Verdana, Arial, Helvetica" size="2">

"5 Drivers. Driver support is about bug fixes only, with a constant list of impending fixes to address, speed tends to go in the negative direction if it changes at all. Nice if you are a business user, rather disappointing for gamers when they read and hear about all the after purchase speed enhancements and feature additions in drivers being offered by other companies or their current brand even."

that sounds more like nvidea than it does MAtrox, don't you think?
</font>

Not in my experience with the card, which is all I have to go on. Matrox had their chance to make a favourable driver impression on me, they failed. Now, they were much better than NVIDIA, but who wouldn't be, they are taking over from Diamond MM.

And yes, there was no 6, just like there is no capital in your sentence, NVIDIA is spelled that way, not nvidea, and it's Matrox not MAtrox. Strangely, I understood you anyway.

[DuRaNgO]

Let's see....

<font face="Verdana, Arial, Helvetica" size="2">With the "Highest Quality" settings, the G400's performance drops off a bit versus the TNT2, especially without overclocking.
Tricrooked?
Since the two cards were so evenly matched under the "Normal" settings, we were kind of perplexed by this one. I kicked it around a bit, did some digging, and found some interesting reasons for the G400's relative performance drop-off here. Turns out the TNT2 doesn't do true trilinear filtering, and trilinear filtering is the principal feature of Q3's "Highest Quality" graphics settings. Rather than do the kind of smoothing trilinear filtering provides to eliminate the vertical banding caused by mip mapping, the TNT2U OpenGL driver simply dithers its output. Even in 32-bit color modes, this dithering is easily discernible. Have a look:</font>

<font face="Verdana, Arial, Helvetica" size="2">Obviously, the G400 is doing more work in order to produce cleaner output, so its benchmark scores drop off a bit more with trilinear filtering enabled. Meanwhile, to my eye, the TNT2's output actually looks better with only bilinear filtering turned on. You can see a larger G400/TNT2 comparative screen shot in a low-compression JPEG image right here (dial-up modem users beware—it's big).
To illustrate the difference between bilinear and trilinear filtering on the G400, I've got a low-compression JPEG image comparing the two, as well.

This is a perfect example of how little a raw frames-per-second benchmark score can really communicate about the performance of a video card. Although the TNT2's frame rate averages were higher here, the G400 clearly walks away looking like the better card.</font>

keep in mind, this is a comparison between the plain G400 and the nviDea tNt2uLtrA

[This message has been edited by DuRaNgO (edited 17 January 2001).]

[Himself]

Can't really make head or tails of those graphs, they seem to contradict each other, they both look like they get the same ballpark performance to me. Not much reason to upgrade on the basis of performance.

I can produce graphs showing a V3 3000 wiping the floor with both of them in Q3A, but who really cares? They are all pretty much the same to me. Like I said, I wouldn't recommend either of the cards as an upgrade today. Frankly, most gamers probably have already owned one or the other already.

[Targon]

And of course, many people, including myself don't give a damn about Q3, or any other first person shooter that's out there. 3D is important, but I'd much rather play a game that has a point and a plotline. The original Wolfenstein 3D was a shooter with a plot. Doesn't matter how much it reflected on the game itself, you had a purpose, to escape.

Since Doom, the ONLY focus has really been on putting together a good game engine, the overall plotlines, for what it's worth, are virtually non-existant. Half-life did a good job for the first part of the game(single player), but still, it ended up a, "If it moves, shoot it". Now, if they would update the engine for "Die by the Sword", that would be a huge reason to go with a new vid card. How about Vampire: The Masquerade Redemption? Both of these are 3D, without being a shooter. That's why I want a faster card, to handle non-shooter 3D games. And if you think I'm crazy, what takes more skill, killing someone with an automatic weapon, or with a sword?

[Himself]

Hey, if they ever come out with a means of duplicating hand to hand or swordfights in computer games I'm there. Kinda sounds like holodeck stuff to me though. Using a mouse and keyboard to swing a sword is rather lame.

[Sasq]

personally I don't play games much (maybe q2 of SOF) so the fps question confuses me, last I remember from school biology class the human eye can only recognize about 25 frames a second, I'm reasonably sure my monitor is not producing and is not capable of producing 200+ frames a second so the benefit of xxx fps over quality is???

[This message has been edited by Sasq (edited 18 January 2001).]

[dZeus]

The human eye can actually see the difference between 60 and 75 frames a second if trained. The reason you can't really see much difference between 25 and 30fps tv (PAL and NTSC) is due to the way televisions work.

But if you're not too used to very high framerates, 20+ fps can actually be experienced as fluid motion.

[This message has been edited by dZeus (edited 18 January 2001).]

[leech]

To a very trained eye? Well, on LSD you can notice all frames *laughs* Of course sometimes everything flashes like my old 50hz Atari ST screen. Well as far as the FPS vs. Quality war, I think it's stupid. most multiplayer games where you want the higher FPS are limited anyhow by the server, so that it's a fair game. So really the extremely high FPS in a game only applies to single-player games. Actually, one game that I'd like to see REALLY good perfomance in and it'd impress me, would be that game with David Bowie....the name if it escapes me at the moment. Omnicron, I think is what it's called. Anyhow, that is one resource hog of a game. Laters.

Leech