[H]uge image diff w/ FX and RADs in FarCry Demo

[TheGreatGrapeApe]

As part of [H]'s review of the FX5900XT, we see once again the difference between the FX series and the Radeons.

Lighting issues, and the significant difference between PS1.1 and PS2.0. Remember this is a TWIMTBP title for nV.

<A HREF="http://www.hardocp.com/article.html?art=NTk5LDM=" target="_new">http://www.hardocp.com/article.html?art=NTk5LDM=</A>

When looking at the performance figures keep in mind that in many games the FX5700U is running at 800x600, and not 1024x768 like the R9600XT, or at lower AA/AF settings.

The R9600XT does pretty well even against the FX5900XT.

Unfoprtunately the latest 56.64 drivers weren't out in time for the review. It will be interesting to see if the anomalies continue with those drivers.


- You need a licence to buy a gun, but they'll sell anyone a stamp <i>(or internet account)</i> ! - <font color=green>RED </font color=green> <font color=red> GREEN</font color=red> GA to SK

 

[HardWareBoss]

Now I know why nVidia keeps releasing slower or maybe better cards. To try an keep up with ATI in sales. nvidia releases more cards but ATI's are better.

 

[BigMac]

This seems as good a place as anywhere to say this. I bought FarCry (full version) yesterday (March 24, I may have been the first in that particular shop) and it is SWEEET. I do need a CPU upgrade though for full enjoyment, sigh.


BigMac

<A HREF="http://www.p3int.com/product_center_NWO_The_Story.asp" target="_new">New World Order</A>

 

[ChipDeath]

Ah yes... So far nothing I've seen anywhere has made me feel like I made the wrong choice swapping my Ti4600 for a 9800Pro.

In fact that review has reminded me to have a go on NFS:underground again.

BTW: the review talks about a 'micro-pausing' problem in NFS:Underground, but <i>only</i> on the FX cards. When I was playing the game on my Ti4600, it would pause very briefly (10th of a second maybe?) about every 10-15 seconds. It wasn't serious enough to have a major effect on gameplay, but it was a little annoying. at the time I assumed it was because I had only 512Mb RAM and it had to load some bits off the disk. Do you think this is the same problem as described in the review? if so then:
a) the game should now run smoothly on my rig
b) It clearly <i>doesn't</i> only affect the FX line, as my GF4 had the same problem, so it's purely something in the drivers...

---
Epox 8RDA+ rev1.1 w/ Custom NB HS
XP1700+ @205x11 (~2.26Ghz), 1.575Vcore
2x256Mb Corsair PC3200LL 2-2-2-4
Sapphire 9800Pro 420/744

 

[HardWareBoss]

Which vid card do you have?

 

[BigMac]

I upgraded to a 9600XT last december. The cpu is an Athlon XP1800+. The card is ok but the game is rather taxing on my cpu (although the box said a minimum spec of 1Ghz). Of course I try to keep the settings at the best eye candy possible. Looks like the AI and the physics are quite cpu intensive.

BigMac

<A HREF="http://www.p3int.com/product_center_NWO_The_Story.asp" target="_new">New World Order</A>

 

[HardWareBoss]

Nice I'm going to buy 9800Pro. When the new cards hit the store.

 

[Willamette_sucks]

Prices aren't going to drop much (the XT prices will though) when the new cards come out. 9800 production will stop soon after and the last of the cards will be phased out, as they are replacing the whole line, not just their top card (although the top card will be the first to come out).

So buy now!

Me: are you saying I can't provide?
Me: cause I know I can provide.
Me: oh and I can provide money too
Rachel: why do we need money when we can just stay in our room and have sex all day?

 

[pauldh]

I agree, the R9700 Pro and R9800 became hard to find and never went much below $200. I think the same will happen to the R9800 Pro. Only thing that would change that is if the new midrange cards were R9800 Pro caliber. And I doubt very much that will be the case.


ABIT IS7, P4 2.6C, 512MB Corsair TwinX PC3200LL, Radeon 9800 Pro, Santa Cruz, TruePower 430watt

 

[kinney]

I dont get what the point is on most every part of their comparison of IQ.

Besides the farcry one that is.

The Farcry comparison is noticable.

The others merely look like they are being rendered by different cards.. whose to say the ATI is correct and the NV is not? Or vice-versa?
They've ALWAYS rendered differently from each other!

The Farcry example is not something I would have noticed if I was actually playing the game running around actually hunting than staring at the trees far off...
but it does hurt immersiveness so I'm not going to make excuses for NV on this particular case.

For whatever sacrifices that were made (besides the Farcry pics) were worth it if thats all that was necessary to put the 5900XT 20FPS faster than the 9600XT as well as generally just, faster.
I realize this is PS1.1 vs. 2.0 but I dont see much of a difference due to that change in IQ (if any). They are supposedly pointing out rendering issues so this is a moot point but worth keeping in mind.
I would like to see 9600XT benches with PS1.1 in his "manual run"...

Which brings to mind- I wish they would have used a recording in their testing instead of a manual run through FC... something I dont support.
That introduces human error or bias for FPS to a great degree.
Furthermore, I would be interested if the condition continued on the 56.64s.

To much up in the air, too many variables/uncontrolled factors left open in this comparison IMO. I question its significance.
2sense on the situation.

____________________________
<b>RESIDENT FORUM WARRIOR
<font color=purple>I just neutered the cat.
Now he's a liberal.</font color=purple></b>

 

[TheGreatGrapeApe]

Well my main point was FarCry, NFS:U and HALO's light diffuse issues are very minor.

But seriously if it's not what you're supposed to see then it's an issue. As minor as it may be to some. Why bother getting a $500+ top of the line CRT if your card is going to contribute to it looking the same as any other bargain CRT because of the image quality produced by your card. Now that's an extreme example but it does put it in more conrete terms than just preference and noticeable. If people don't notice the small details then fine but for those paying 200-500%+ premium on their monitors, why make the card the weak link that spoils that investment.

In alot of ways it's what you value. Does it matter that the FX5900XT may be rendering it a little off compared to the R9600XT or does the FPS matter more to you. In many games the FPS will be close so why not go with the best quality image. In other scenarios the FX5900XT will have a great advantage frame/s wise so you may get a resolution or AA/AF step not available from the RX9600XT, and that may benifit you or be more noticeable to you than the, what may be to you, slight differences in scene rendering. IT's really a matter of taste. PS1.1 vs PS2.0, I don't care what the R9600XT does in PS 1.1, because it can do it in PS2.0 which is nice. But it would be interesting for argument's sake. Just like the NV40 vs R42X, if PS/VS3.0 have something to offer I would like to have it as long as it doesn't cancel out other benifits. I WILL strongly consider the NV4X if it offers PS/VS3.0 support without sacrificing some other items over the R4XX series. Hey why not see/experience it all, that's why we have these features in modern games, they must offer some benifits, even if it is just shinier water.

But when comparing like cards that cost as much, why not get things looking like they are supposed to or the best they can look? I'd prefer a 24bit laptop LCD over an equally cheap/expensive laptop that can only display 16bit colour. But some people wouldn't notice the difference and would prefer a touchpad over a trackpoint, and that's their major feature.

The funny thing about ATI vs nV is how the Xbox (nV) shows alot of the same rendering as ATI in these titles, so it's nV's drivers that appear to be creating some of these anomalies. Also a reference raster image will usually expose which is doing this or that right if one is available. So it's not just ATI vs nV, but many times that alon can be thecase, but I would expect that more from ATI than nV because of it's limitations. The problem is that the nV at full potential is slow and it gets hampered, which is IMO what causes most of this. But that's just my op-onion as usual.

The main thing is to illuminate the differences to make informed decisions. Everyone initially felt the FX5700U was blowing the R9600XT out of the water, yet now it seems very different, especially when seeing what resolution and AA/AF are useable. And yes the FX5900XT is generally faster, but seriously, it's not really fair to compare the two. The FX5900XT is a great marketing move, but it did canabalize the true R9600XT competition to achieve that win in that segment, and even then it's at a more expensive price tag. Compare it to it's intended target and the story, as I said above is very, VERY different.

Furthermore, I would be interested if the condition continued on the 56.64s.

Yeah like I said, I'd like to see what was addressed with that. I don't expect much image improvement for FarCry, as I have seen reviews not that it's still 1.1 for 56.64, but it would be nice to see the speeds, especially with the 4.3 drivers as well, as they offer speed boosts, especially in PS2.0 heavy games/apps.

As for it's significance, true for speed and such you have a point, runs differ, setups differ. As for the title and this test, I think it's very significant, and quite telling of what alot of titles have to offer the NV3X line in the future, especially games that are not TWIMTBP titles.

Of course, that too, is just my 2 frames worth.


- You need a licence to buy a gun, but they'll sell anyone a stamp <i>(or internet account)</i> ! - <font color=green>RED </font color=green> <font color=red> GREEN</font color=red> GA to SK

 

[kinney]

Cool.
I just wanted to get a different viewpoint out there on this forum.

I do though still not agree that pitting the 5900XT isnt fair to pit against the 9600XT..

I see that just as fair as pitting the 9800 Pro against the 5900XT even though the 9800Pro is about $40US more..
just depends what you want.

Frankly, I like the XT most not only because its NV and their drivers are top-notch (ya theres a few anomalies here and there like this FC issue, but I can pull up more for ATI as well..) and their features are the best IMO (referring to overall driver features such as Nview, game profiles, custom resolution creation and refresh settings, a better working refresh rate implementation, ect.)... add that up with the ability most of the time to turn some 5900XTs into 5950Ultras, or with a lemon, "just" a 5900Ultra (yes you read correct!) from a simple flash makes them the choice for me.

Kinda like the new 9800 nonpro IMO.
And we BOTH know what I thought about those!! I bought TWO!

Just my op-onion (witty ape), as well... not a popular one but one I truly see as being valid.
One the other hand, you can get near 5950Ultra or above performance with just spending the extra $40US and getting a 9800 Pro and so I'd see that an equally appealing choice.

On yet ANOTHER hand (how many hands could I go thru here?), you lose the NV drivers detailed above and you must be more questionable about longterm support from ATI.

I could prob go back and forth here for quite some time but you get my drift.
Trying to be fair and balanced.

____________________________
<b>RESIDENT FORUM WARRIOR
<font color=purple>I just neutered the cat.
Now he's a liberal.</font color=purple></b>