Since Poser didn't really rely on the video card but loved the ram, I never paid much attention it. If I get this right D/S wants a GOOD video card, right? If this is true I would like some suggestions as to video cards. I'm running AMD Athlon XP 2200+ Processor (1.80 GHz),XP Home, 1Gig DDR Ram and have an open AGP Slot.

Attached Link: http://www.tomshardware.com/

The two most popular/powerful video cards (3D chipsets) these days are from nVidia and ATI. ATI makes and sells their own cards and also sells their chipset (and/or card itself) to a few other card sellers. nVidia only sells through other card manufacturers and there are lots of them. You pretty much want the same features on a card that make it a rocking game machine. Most 3D software (including D|S) won't use most of the features that a modern 3D game uses, but those cards are designed to do the things that 3D apps DO use - very fast. You also pretty much want to get the best one you can afford, except that: personally, I'd steer clear of the top-of-the-line model from any vender (you pay a premium for a few minor percent increase in gaming performance and those cards tend to be tweeked to near breaking point) and, pretty much ANY of the modern lineup will do a great job in your 3D apps... I'd be more concerned about getting a card with plenty of on-board memory (64mb minimum, the more the merrier), Some people think he's biased, but you can always get some good background facts and specs at Tom's Hardware (see link). He currently has a review of 46 cards there. My personal bias/preference is for nVidia-based cards. They have excellent OpenGL support and a unified driver archetecture that makes it easy to keep up to date with their latest drivers from their site.

Damn!! Thanks Spanki,for the info to consider & the link which I will be checking out. :-)

Spanki, Re latest drivers. Maybe you (or someone) can advise me. I have the Nvidia TNT2 32 meg card (pretty old, I know). Will changing driver do anything worthwhile? The drivers I have are from about 2 years ago. It's kind of academic because I'll be changing the whole machine in a few months, but I just wondered. I've never really taken any interest in hardware and it's all greek to me. mac

Mac, I'd say it's worth a try if you're having problems. You're right, that card is firly old, but new drivers might breathe new life into it. Of course, when you get ready to upgrade, you can buy the low end of the latest rev cards (a Nvidia GeForce FX 5200-based card, for example, which is about 4-5 times better than the card you have) for about $100 at Wally World or Best Buy.

Thanks spanki. I actually went to the Nvidia site, but couldn't figure out what the hell to download. I'll try again when I'm not so tired. mac

Attached Link: http://www.nvidia.com/content/drivers/drivers.asp

Try this link (attached), then click on "Graphics Driver" in column 1, then "GeForce and TNT2" in the second column, then choose your OS in the 3rd column and go from there. Cheers, - Keith

I happen to have a GeForce FX 5200 (128MB), and DAZ|Studio runs very smoothly, even with multiple figures loaded. I have some visual glitches with the user interface (menus, tabs, buttons), but I haven't determined yet whether that's due to the graphics hardware, drivers, OS, or some combination of the three. Based on other reports, however, I'm starting to suspect that it's Win98.

I'm running the same card (an Asylum GeForce FX 5200 made by BFG Tech) and not having any of the glitches you mention under WinXP, so you may be right.

I almost bought a BFG Asylum. Went with a PNY Verto, instead. Would have ordered a Diamond Stealth, if I'd known at the time that they were back in the graphics card business.

Thanks for the link. spanki. I d/ld the latest one and I'll give it a try. cheers mac

how's this for someone who's clueless: how do I find out what kind of video card I have, do I even have one? running win98 btw.

ok, found my computer specs how is this for a card? I know it's old. Intel Direct 3D 2X AGP (shared memory)

That isn't a video card, I think, but an embedded controller - it uses your computer's main RAM for its data storage, and is less self-contained than a stand-alone card.

See, totally clueless here

Intel's graphics solutions are worthless. They're only really useful for the $500 budget box market; if you need to do more than read e-mail, run TurboTax, and surf the web, it won't hold up. Avoid Toms Hardware. It's not at all worth your time to read any more. Try instead www.xbitlabs.com, www.techreport.com. At least you can believe what they post.

I just spent half a day browsing the sites you listed and I found xbit to be quite (ATI) biased and somewhat loose with proclaimations. The techreport site seemed a lot more level-handed in it's reviews and news... though that site didn't have quite the variety of hardware tests that you can find at Tom's (no reviews of Gigabyte motherboards, for example). Thanks for the links... it's always nice to get multiple takes on something.

For simplicity sake, integrated graphics controllers are still usually called "video cards". ie: "My video card is an onboard Intel blah blah". Most integrated video cards are crap anyway, but, as Thalaxis pointed out, Intel's are the worst by far. Upgrade if you can.

Ah, that I get! My video card is the worst of the worst and I need a new one. Having the worst makes things simple really, it's likely that anything that I buy will be a step up.... Still going to wait for recommendations for what works best with DS.

According to most reviews, ATI-based cards currently have a performance edge in single-textured rendering (as in using one texture on any particular polygon) and DirectX 9 shader rendering, while NVidia-based cards have the edge on multi-texturing (more important to games than 3D apps) and NVidia has better OpenGL drivers/support than ATI does (more important to D|S). For all intents and purposes, even the current cheapest/low-end add-on cards from either company will be many times better than an integrated Intel solution, so take your pick ;).

Right now, ATI has faster hardware, especially with DirectX9 shaders and such. XBit isn't the only one reporting that. With Tom's you trade of variety for integrity and quality. Don't use them as a guide for how you spend your money.

Let me back up what Thalaxis says re: Tom's Hardware. It is widely reported and PROVED that Tom's reviews can be bought and that he brings a specific bias to many of his reviews in favor of certain products/companies. Avoid his site at ALL costs, as the reviews are extremely questionable. Go to any tech site (Ars, HardOCP, nVnews, slashdot) and they all have stories/proof that you can find (with "search" of course).

Right now, ATI has faster hardware, especially with DirectX9 shaders and such... I really don't mean to belabor the issue, but just saying it has 'faster hardware' doesn't really tell the whole story. As I mentioned above, it depends on what you want to do with it. ATI does currently render single-textured DirectX 9 data faster than NVidia hardware does and also apparently has an edge in it's ani-aliasing quality capabilities. But NVidia hardware has better OpenGL drivers/support and renders multi-textured (games mostly) data faster. For something like D|S, which (currently at least) only does single-texturing but runs under OpenGL, the two cards are probably roughly equivilant (though if given the choice, I'd prefer more robust/reliable OpenGL and perfectly adequite speed over potential OpenGL issues but faster editor-window rendering). They both make great hardware... I don't think buying either is a bad choice - particularly if your current video solution is more than 2 years old or is integrated on the motherboard. As for Tom's site... I don't really have any bias one way or another about it. I am mostly looking for facts/specs and that info is there - it's pretty easy to skip over any editorial bias he might have. Personally, I found the xbit info/site to be at least if not more biased than anything I read at Tom's... that guy's an ATI evangelist ;). Anyway, there are plenty of similar sites out there and it doesn't hurt to visit several of them and average the results of what you read to level out any bias. Thanks for the extra links.

Actually, most tech sites are pretty miffed at nVidia, since they tried to hide the fact that ATI has better hardware behind driver cheats. nVidia's OpenGL drivers are still better than ATI's, but my Radeon 9800Pro (in a P4 3.2 GHz) runs LightWave, Motion Builder, Cinema4D, and Messiah:Studio extremely well. I was very pleasantly surprised by that, since I had all sorts of trouble with the M:S pre-release version on my Radeon 8500 in my Athlon for about 3 driver revisions. Using more than one site before making buying decisions is also a good idea. BTW, add www.anandtech.com to the list -- he's also pretty thorough with his reviews.

Honestly, if you are using the card primarily for 3D graphics, you really won't lose out buying a product from either company (or their partners). The features that really set them apart are used 99.9% only by games, and mostly as marketing talking points. For 3D previews, cards from both companies will suit you well. Both have stable drivers (with ATI finally catching up to nVidia, more or less). The best decision is to get the best card that you can afford - each brand has many cards at various price points. Spend what you can and enjoy the card. If you plan on using it for games as well, I would lean a bit more to the ATI over nVidia. Otherwise, check them both out.

Spanki, I am not sure where you get your data but ATI's current hardware is faster than Nvidia's all-round. Especially in games.

Oh wait, take that back.. Nvidia's 'GeForceFX 5950 Ultra' slightly edges out ATI's 9800XT in some games. What I am curious to know though is, does D|S use HARDWARE for rendering? I know it supports OpenGL, but so do many other apps and those apps dont actually use hardware for the actual render. They use hardware for scene setup/viewports, but software for rendering.

Freon, I get (some of) my data from many of the sites listed above. Just as examples, take a look at this link: http://www.techreport.com/reviews/2003q4/geforcefx-5950ultra/index.x?pg=2 ...which is a review of NVidia's latest top-end card. At the top of that page, you see a chart listing the tech-specs of the chipsets (I'll use cards/chipsets interchangeably here, even though they are 2 different animals). Basically, the ATI chipset has 8 pixel pipelines and NVidia only has 4 so the 'peak fill rate' for PIXELS per second is obviously higher on the ATI cards. Now look at the next column... note that the ATI chipset only has one texture unit per pixel-pipeline (x 8 pipelines = 8 texture units) while the NVidia chipset hs 2 texture units per pixel-pipelines (x 4 pipelines = 8 texture units). The next column shows the maximum fill rate for 'textured pixels'... and as you can see, the fastest ATI card doesn't even beat out the slowest NVidia card. How does the NVidia keep up with only 4 pixel-pipelines? read on... Now check the memory bandwidth columns... again, only the 2 top-of-the-line ATI cards beat the bottom-of-the-line NVidia card. The top end NVidia card has a nearly 30% advantage in memory bandwidth. Have a look at the charts below that for the results... as I stated a few times above already, for single-texturing, ATI wins. For multi-texturing (blending multiple textures on any particular polygon as is done in many modern games), NVidia wins. The next page is a good read too, which talks about NVidia's advantage in OpenGL-based games/apps and it's shortcommings in DirectX9-based apps/games. The rest of that review covers benchmark scores and you'll find NVidia hardware on top in many of those too. Overall, it's a pretty good, un-biased review and on the conclusion page: http://www.techreport.com/reviews/2003q4/geforcefx-5950ultra/index.x?pg=12 ...you'll see that the author picks the ATI card based on the things HE wants from a $500 card. So you can see that I'm not trying to gloss over the fact that the ATI card has some advantages, I'm just pointing out that the NVidia cards have advantages as well. Generalized statements like "but ATI's current hardware is faster than Nvidia's all-round. Especially in games" is just not true, or at the least, misleading. Again, my "current" card of choice is NVidia. I've owned several in the past as they've dominated the field for performance until only recently. I've also owned ATI cards and (aside from some OpenGL driver concerns), don't have anything bad to say about them either. When it comes time for me to upgrade, I'll evaluate the latest offerings again and choose the one that suits my needs. If that's an ATI or S3 or Matrox card, so be it - my product loyalty only lasts as long as you can keep my confidence in your product ;). At the risk of repeating myself, I'm NOT here to defend NVidia or sell their hardware for them. I'm just trying to fill in BOTH sides of the story. I don't care to particpate in "my hardware is better than your hardware" type discussions (and don't want this to devolve into that), so I think I've said my peace now. Cheers.

looks like we x-posted... you are correct, D|S does NOT use the hardware for final rendering, only for editor-window rendering. Which is why I've been saying all along that it really doesn't make THAT much difference which one you choose... except that (contrary to statements others have made in this forum) some ATI-using folks seem to be having more driver-related issues than NVidia-using folks. Having said that, I'm sure that Daz will work out (or work around) whatever issues they can to make it run equally well on everyone's system.

One slight (symantics, mostly) correction... there IS an optoin to 'Enable OpenGL' for the final render. And if you enable that, it DOES use the hardware for rendering, but you get the exact same render as you do in the editor window, so it's really only useful for things like seeing what your final cropping is going to be (if say you render a 400x800 image when you're working in a 800x800 workspace, the sides will be cropped and the image centered, so you could get a quick idea of how it's going to be cropped).

You didnt need to type all that out Spanki.. I happen to be quite knowledgeable in the area of computers and 3D hardware. I build my own systems, been following the 3D hardware scene since the S3 Virge3D days, and I am also an avid gamer. As I said in my correction, the 5950 does slightly edge out the 9800XT in some games.. However, one thing you should note from that review is dont get caught up in synthetic benchmarks (3DMark), and make comparisons based on them. For example, look at the 3DMark result for multitexturing. The Nvidia card has what looks to be a big lead: 493Mtexel/sec. But when you look at the score in a multitextured game like Quake3, it is hardly faster. That lead drops a lot when you turn on AA and AF. *A sign Nvidia hardware is not working very efficiently with that huge memory bandwidth lead (you can see the same thing in every test with AA and AF on). Move to a game like Unreal Tournament, which uses Direct3D (DX 8), and far more texture passes than Quake3 btw, now suddenly the 9800XT is ahead. How is that possible if the Nvidia card is faster at multitexturing situations? Hardware efficiency, drivers, etc, aka - real world performance, come into play. 3DMark is not a measure of real world performance and its figures for texturing do not translate to real world situations. It has also been proven by many sites that Nvidia's drivers were optimized explicitly for best performance in 3DMark. So much in fact, even Futuremark themselves said Nvidia was cheating, and game developers disclosed they had to make special optimizations to their games just to make it work right with Nvidia's hardware! So dont go by the results that meaningless benchmark gives. Another problem that Nvidia has faced lately is cheating in the image quality area. They took shortcuts in IQ to increase performance, but we dont need to get into that here as this has already gone well beyond anything to do with D|S. LOL BTW, dont get me wrong.. Im not some ATI freak or anything. Nvidia still has a better OpenGL implementation, ive been using Nvidia hardware since the TNT2, and I still use my trusty GeForce4 ti4600 (though it is showing its age these days lol).

Your all great. Not sure if I got my question answered or if I got more confused, but I definately learned allot from all of you and am willing to learn more if anyone wants to share. :-)

Strix, thanks for bearing with us as we hi-jack your thread... so much the better if some of it is actually useful info ;). Freon, >> You didnt need to type all that out Spanki.. Ok.. it's not always easy to tell in this online means of communication. I thought I'd fairly covered the issue, giving pros and cons on both sides and your response was the online equivelent of "Is not!" ;). I could have just come back with "Is so!" and we could take up from there, but I didn't think that would be overly productive ;). >> I happen to be quite knowledgeable in the area of computers and 3D hardware. I build my own systems, been following the 3D hardware scene since the S3 Virge3D days, and I am also an avid gamer. Cool, thanks for the info.. now I have better perspective to talk from. As for me, I started playing with and programming computers back in the early 1980s. I was a (mostly 3D related) Programmer and Producer in the Online Gaming Industry for 7 years and have been building computers, gaming, messing around with, programming and following 3D-related topics of all types for about the past 10 years of that time. I've written software 3D pipelines from scratch, as well as programmed in both DirectX (Direct3D) and OpenGL. >> However, one thing you should note from that review is dont get caught up in synthetic benchmarks (3DMark), and make comparisons based on them. Yep.. agreed, 100%. And I started to add similar statements to my previous reply, but - as you noted - it was already getting wordy. But let's step back and think about this for a minute... just 'how' do we determine that card X is 'faster' than card Y? We have a few choices... - we can look at the specs and devine the theoretical limits of each card, but that rarely gives the whole picture. It's too dependant on how efficient the drivers are for one thing and there's no 1-to-1 correspondence with actual real-world game/3D app performance. - we can run some synthetic benchmarks. these are usually set up to try to test the afore-mentioned theoretical limits of the hardware, but (again, as you mentioned), aren't always as meaningful as real-world apps/games. - we can run some real world app/game-based benchmarks. But keep in mind that there may be issues clouding these results as well (driver cheats being one, but the particular programing style/methods used may also favor one set of hardware over another). - lastly, we can just go by 'gut instinct' - but I doubt you'd get much respect as a hardware reviewer if you just went by that ;). ...When you factor in personal bias based on personal work/gameplay requirements/preferences, good/bad experience with previous products, reports of foul-play that taint your views of company X forever forward, fan-worship of a company/product/or reviewer X over reviewer Y, or otherwise, it's not always easy to come up with a 'definative' answer. I can definately state with 100% confidence that the current top of the line ATI card is faster in every repsect than the original Virge3D, 3Dfx Voodoo, etc., but when someone says the same thing about the latest rev NVidia hardware, I don't find the answer to that question as such a black and white issue. Each of the above methods have inherent problems/issues and none can tell the full story alone, so you try to use every method available to get an overall feel (gut instinct? ;) for which is faster or better suits your needs. Anyway, just one final note here before I wrap this up... you mentioned a few times issues about NVidia 'cheating' in their drivers... First off, not ALL of what has been reported as cheating is actually cheating, but SOME of it was and they got caught (and shame on them for doing it). But the fact that they did causes some exageration, speculation and lack of sympahy (or trust) on the part of some reviewers about the issue. For example, most fail to mention that this caused NVidia to implement stringent new company policies in regards to any future driver optimizations. Most fail to mention (or gloss over the fact) that they removed those cheats (that were cheats) from the newer drivers. Most fail to mention (or gloss over the fact) that ATI is guilty of similar 'cheats', etc. But as you say, we don't really need to get into all of that here ;). Sorry for the new novel... ;). - Keith

Thank you Spanki & 1Freon1. I really do appreciate your time and explanations. While it's a bit confusing and I still don't know which card to go for, I feel that the xchange between you two has expanded my(our) understanding on this issue or issues, more than just the comments of one or the other. I feel like this question will arise more and more as more folks try D/S and more folks come into 3D graphics overall. Many (like me)have spent years fiddling with software and little time trying to understand hardware (to scary :-)

Keith, The cheating is not the biggest issue here. There are inherent and definite problems with the current generation of nvidia cards that affect anti-aliased images. This is a stated fact. Even they have admitted to it. This does not have much, if any, impact on 3D previews, but definately comes into play with DirectX 9 games. Additionally, there is NO WAY that these cards can be used for hardware rendering. That is 100% absolutely performed by the computer CPU. There is NO WAY for any offloading to the graphics card. The only cards that do this cost in the multiple $1000.

"There is NO WAY for any offloading to the graphics card. The only cards that do this cost in the multiple $1000." That's incorrect. Go look at www.mad-fx.com.

Attached Link: http://www.renderosity.com/messages.ez?ForumID=12452&Form.ShowMessage=1598358

ryamka, from my post #23 above: >> ATI does currently render single-textured DirectX 9 data faster than NVidia hardware does and also apparently has an edge in it's ani-aliasing quality capabilities. from my post #29 above: >> you are correct, D|S does NOT use the hardware for final rendering, only for editor-window rendering. Which is why I've been saying all along that it really doesn't make THAT much difference which one you choose... except that (contrary to statements others have made in this forum) some ATI-using folks seem to be having more driver-related issues than NVidia-using folks. ...I know it's a lot to read, so maybe you missed that, or maybe your comments were directed to someone else? BTW, anyone interested in this thread might also find the provided link interesting. Cheers, - Keith

Thanks for that link Thalaxis... I've been wondering when someone would implement all the DirectX 9 stuff in a 3D rendering app (as opposed to game).

My guess is that it will become more popular with PCI- Express, DirectX-Next, and LongHorn, especially as cG evolves to provide access to the increasing power and flexibility of the hardware. Right now it's great for preview rendering, and not up to snuff for production rendering. I think that will change, though I can't guess at how soon. :)

right... lets see now.. arguements aside in statements by the companies..how many poly's shifted and how fast..how many megawosits we run at etc... it comes down the to following. Does it run OpenGL to an acceptable Standard? Is it within your budget? when boiled down to those important points, then either ATI or Nvidia will do, so look for what extras are being offered on the card in your budget range and what Ram size it's got ... trust me.. you won't, with the naked human eye, notice much of a dfference.

...recommend an ATI Radeon 9700/9800 Pro card. ^_^ Runs great for me, on WinXP!

Thank you also -Yggdrasil- :-)