Radeon over G400?

[drzaius]

i'm looking at the overall chewyness of these 2 cards ,Radeon 32meg and G400 DH 32meg (this is what i have now)
now i know that the Radeon could kick the G400 in frames anyday, but i find the speed of my G400 (150/200) fast enough, and anything faster is just a bonus.
but what is the image quality like? what about driver stability? and DVD playback?\

i won't make a decision about this until the G800 comes out, but i just wanna see the feedback of the Radeon, and how well it is to real people (not online review people)

[cputnam]

If you're running Windows 2000 Pro, don't, DO NOT get an ATi Radeon. ATi's Win2000 drivers have more bugs than a year old twinkee. Scratch that, twinkee's have a half life longer than Plutonium...

You get my drift.

2D image quality, the G400 is tops. As to DVD quality, needless to say, this is where ATi excells (except under Win2000 where you get lovely stop errors....

[sasa]

I havent seen Radeon in action, but my friend has Ati All in Wonder Rage 128 bla bla bla. This card is in TNT1 speed range, but full of multimedia (crap) features. If you are gamer, you should forget ATI at once, because ATI has the lousiest driver I have ever seen. There are problems with most games like missing and corrupt textures.
So, since I am a gamer and I dont care about multimedia crap, I will never buy Radeon whatever the price is. If they still cant write decent drivers for museum card, do you think Radeon has good drivers?

[Chris B]

As far as DVD playback is concerned I can tell you this:

I recently read a test/review in a Danish computer magazine where they tested DVD playback on GeForce2/pro/mx (or whatever the newest version is called), TNT2, Radeon All-in-wonder, G400 and a Hollywood something!!! (a real mpeg card).

Result.... As far as Image quality goes, the G400 kicked everybodys ass... eve´n the hollywood mpeg card. And the only card faster than the G400 was the Hollywood card (the only card made specifically for mpeg playback in the test).

I can´t remember what machine they did the test on, but the result was pretty clear..... If image quality and DVD playback is an isue for you.... go with the G400!!

[This message has been edited by Chris B (edited 30 October 2000).]

[dZeus]

the reviewer in that magazine must have been high on crack.

The Radeon does hardware iDTC and motion compensation, which _MUST_ result into lower CPU-utilisation when playing back MPEG-2.

And hardware iDTC on the radeon also is said to be one of the best (if not the best) quality solutions available.

Not that I think the ATI Radeon is such a great card, but at least for MPEG-2 playback it should be the end-all solution.

[Darin]

I recently upgraded from a G400 to a Radeon, and I can't confirm any of the issues posted here. I have no driver issues, but the only game I ever play is Q3A. 2D AND 3D image quality to me is the same (both are noticably better than nvidia products), and 3D speed is of course much better (especially in 32bit). I haven't yet tried DVD playback on it, but I'm quite perplexed by the reference to the Danish magazine. DVD playback is the one area where every reviewer I have seen to date has unanimously stated that ATI is the best, bar none. I just recently got a player for the computer, and I will try it out, but I didn't have it installed when the G400 was in there so I wouldn't be able to compare anyway. The biggest disapointment to me was no support for 1280x960 resolution (only 1280x1024, which is an odd aspect ratio).

I'm moving my G400 into a server I'm building, so it will be easier to make some comparisons when that's done, but it won't really be apples for apples: the server is going to run Win2k, while I'm still on 98SE for my workstation, and I'm going to use the BNC/VGA switching abilities of my monitor to switch signals between the two, so they won't have equal signal paths.

From my experience so far with the Radeon, I'd say it's an excellent choice if you wanted to buy a card today, but why upgrade if you have no real reason to? Both the G800 and the 2nd generation Radeon should be out "soon" (there's plenty of rumors for both, but I think it's fair to expect each within four months). At this point, I consider the Nvidia products a non-contendor due to their image quality issues. There's always something better around the corner, but why upgrade if you don't need to?

If you want some better comparisons (I've only had the Radeon installed for a week), you might want to post a G400 vs Radeon article in the Radeon forum at www.rage3d.com. There are several people there that have upgraded from G400s. I've seen one guy that feels the Radeon has better 2d quality than the G400, a couple that say the G400 has a slight edge, but most people fall into the same category as me... pretty much the same.

[Bursar]

What's all this constant rubbish about Nvidia having bad 2D?

I've just replaced my G400 MAX with a GeForce 2 Ultra, and to my (admittedly untrained) eye, there doesn't appear to be any difference.

My desktop looks just as sharp at 1280x1024x32@75Hz.

[Greebe]

Bursar, what monitor do you have?

[dZeus]

a 10" Hercules monochrome with VGA -> Hercules convertor fitted between the graphics card and the monitor.

Just kidding

[Greebe]

LOL

[Darin]

Bursar, admittably I don't have any first hand experience with current Nvidia hardware. My G400 replaced a TNT1, and I could tell a difference in picture quality with that upgrade. Of course, differences in 2d quality are going to be more evident with higher resolutions, higher refresh rates, and larger, higher quality screens. At the time of my TNT1 -> G400 upgrade I had a good 17" Trinitron monitor, and I could tell a difference. I now have a 19" FD Trinitron, and couldn't really discern anything that made me feel the quality was better or worse during the G400 -> Radeon upgrade. I can't say that I know that the current Nvidia cards are the same as my old TNT1, but general opinion definately puts the Nvidia cards low in that area. FWIW, I just read a new review on sharkyextreme that includes the G450, Radeon, and several geforce cards, and they list (in order of 2d quality, higher first), G450, Radeon, 3dfx & the Elsa implementation of geforce tied for third, and all other geforces last. If you read the section on the Elsa card, they make mention about the Elsa doing better than the other nvidia cards. Since Sharky provides no table of contents, I'll save you the hassle of paging through the entire review to get the table (at the end) that summarizes everything: http://www.sharkyextreme.com/hardwar...10-00/28.shtml

Like so many hardware reviews, I think this has it's flaws, but it's at least nice to see 2d quality not only mentioned, but ranked with other cards.

[Bursar]

My monitor is an Iiyama Vision Master Pro 17. I don't have the exact model number to hand, but it'll be 4 years old this Christmas. It's connected to the graphics card via a BNC cable (always has been) and the image is very steady and sharp.

I'm not hanging around this board waiting for an opportunity to bash Matrox and praise Nvidia. I only bought my new card because I wanted something faster. If Matrox had the G800 available (or at least some dates for release), I probably would have hung on a bit longer. And if my second machine was AGP capable, I would have slammed it in there without thinking.

I don't do a huge amount of 2D stuff. Most of it is Internet related (surfing and so-on), but the quality of the image seems just as good to my eye. At the end of the day, this is what counts.

Maybe some measuring or testing equipment would show up some differences, but if you can't see them, does it matter?

Edit - The Nvidia cards that Sharky reviews are all based on the significantly cheaper MX chipset. In order for this chipset to reach its audience, certain performance limitations were imposed on the cards, and I wouldn't be surprised to learn that other cost cutting factors also hinder the 2D score of these cards.

[This message has been edited by Bursar (edited 30 October 2000).]

[Darin]

Originally posted by Bursar:
The Nvidia cards that Sharky reviews are all based on the significantly cheaper MX chipset. In order for this chipset to reach its audience, certain performance limitations were imposed on the cards, and I wouldn't be surprised to learn that other cost cutting factors also hinder the 2D score of these cards.

I understand the logic there, but I don't think it holds true in practice. From what I've read, I believe the biggest variables affecting 2d image quality are the components on the video card downstream of the graphics processor, particularly in the filters used to prevent high frequency (higher than the required bandwidth for a video signal) noise from being emitted. I don't think there is going to be much, if any, difference in image quality that can be attributed to an MX chip vs. the "regular" GF2 processor. Nvidia makes the processor, and the rest of the board can (and does) differ from their reference design. If the board makers don't put much thought or concern into the design and quality of the boards, the image will suffer. If the conclusion in Sharky's review is correct, then Elsa has apparently done better in this regard. Of course, since the MX is a budget GF2, it is possible that the card makers put higher quality components in the regular cards, but based on the rather unanimous conclusion that ALL Nvidia based cards are a notch below in image quality leads me to believe this is not true. There are people out there actually hacking their GF cards by modifying or removing the output filters to improve their image quality, but of course by doing this they are voiding their warranties, violating FCC regulations, and increasing the chance that their cards are emitting high frequency radio interference.

I don't think that it's complete coincidence that the cards that are reported to have better quality 2d (3dfx, ATI, & Matrox) are those that are manufactured by the chip makers themselves. Nvidia has developed a chip that does its job really well... it pushes pixels. It is the fastest 3d card available, and my guess is that most of the card makers that use the nvidia chip know 3d speed is the main reason most people would buy an nvidia based card. It has been a general consensus for a long time that all nvidia cards are fairly equal performance wise, with some having some extra features. From a cardmakers point of view, it doesn't make much sense for them to put money into improving their signal quality, since their buyers are going to be looking at what chip it has, what features it has, and the price. 2d quality has historically been overlooked, but it's getting mentioned more and more these days, probably due in part to companies like Matrox that produce cards that are noticably better, and partly because of the proliferation of affordable, larger, higher quality monitors that allow us to see a difference. Maybe Elsa has noticed this trend, and is reacting to it.

I'll be the first to admit that the difference in 2d quality is often times exagerated, but I do think there is a difference.

[drzaius]

thanks for the imput on this topic so far guys.

There's always something better around the corner, but why upgrade if you don't need to?

this is very true, i am happy with what i've got, and my current cash flow isn't very giving towards a new $200-350 video card.
i'd rather spend that money on a new mobo and CPU/DDR RAM anyhow.

[DuRaNgO]

good choice, your eyes will thank you
jim

------------------
P3-700e @ agiga! Check it here!
Abit BE6-2
TwinMos 256mb pc-133
G400 32mb DH
Maxtor 15.3gb 7,200rpm
SB Live!
Toshiba 12x DVD
Iomega 650cd burner
Winblows 98se & DX7
and 384k DSL!