World of Warcraft Video Options

There is a surprising number of options for adjusting image quality in World of Warcraft:

There are a handful of "shader effects" that will run on both DX8 and DX9 hardware (and are basically indiscernible between the two classes of hardware), most of which provide minor changes in image quality.  There is one exception, the Full-Screen Glow Effect shader, which we found to be a bit over the top.  This shader, in particular, makes WoW look a lot like a Playstation 2 game as you can see from the screenshots below.

Mouse over the image below to see Full-Screen Glow enabled.

We opted not to test with the Full-Screen Glow Effect enabled. 

Graphics performance in WoW is primarily limited by three settings: Resolution, Terrain Distance and Anisotropic Filtering.  Not too surprising is the fact that these three settings are also responsible for the greatest impacts on overall image quality. 

Any gamer is familiar with how resolution impacts image quality, so we won't go into much discussion there.  Terrain distance can have a pretty big impact. Here, we have three different settings for terrain distance: the slider at the lowest setting, the 50% setting and the highest setting:

Lowest Terrain Distance


50% Terrain Distance


Maximum Terrain Distance

The difference between the 50% and highest settings are much less pronounced; thus, if you're unhappy with performance, this is one sacrifice that you can make that's pretty reasonable. 

By now, you should be familiar with what anisotropic filtering does (if not, take a look at any major GPU launch and our review of it for a quick primer) and its usefulness in WoW is extremely well pronounced in areas where there's water.

Mouse over the image below to see the Anisotropic Filtering slider set to its maximum value.

The newer cards take a much lower performance hit from maxing out the anisotropic filtering slider and they end up looking better too. 

We wanted to be able to directly compare all GPUs, so we maxed out all of the sliders and options (with the exception of the Glow effect and V-Sync).  The screenshot above of the Video Options screen is actually the settings that we used for benchmark (varying only resolution). 

Index ATI vs. NVIDIA Image Quality
Comments Locked

59 Comments

View All Comments

  • Mizuchi - Thursday, March 24, 2005 - link

    I have a generic laptop that I bought from Hugebee Corp (forgot the name of their smaller retail online storefront) that seemed to be an ECS model with a P4 2.8 GHz w/HT, 1GB mem, 5400 RPM drive and 9600 Pro. I'll see how it runs at it's native res, 1280x800 (WSXGA I think), and see if it's smooth enough.


    As for settings with my 2500+ XP Barton, 512MB, 7200 RPM (8MB Cache), 9700 Pro system, I'm running 1280x1024 with LOD, Shaders enabled (sub options off), and everything else min or off. These settings seem to run the fastest. Getting ~60 FPS normally... I guess I am fretting too much over crowded areas and upon landing from flight.

    For reasons unknown, the game runs choppier at lower resolutions. I went as far as restart my system with the lower res saved for my next log on and it still seemed choppier, just to prove that something was wrong. My monitor drivers are installed and working and it runs at 60 Hz at all supported resos...
  • BopTop - Thursday, March 24, 2005 - link

    Finally some benchmarks on MMO's. I'll agree with the previous posters, although I don't play it, I would also like to see EQ2 broken down.

    As an aside, I've read there's also been an issue with Intel's speedstepping cpu's to cause players to alternatively speed up and slow down ingame. This has actually led to accusations of speed hacking and bannings. Would be interested to know if this is a legitamate bug or a cover-up excuse.
  • Kouri - Thursday, March 24, 2005 - link

    Great article, it's very difficult to bench WoW.
    Too bad you don't have any mobile GPU benchmarks. I've been trying to figure out how to get the best quality/performance ratio out of my Dell Inspiron 9100 (/w a Radeon Mobility 9700 128MB) for a while now. Just going outside and doing normal quests you can get away with a lot of the higher quality settings, but going into cities or encountering a large number of players will get you killed (I think having lower texture settings really helps here).
  • ElFenix - Thursday, March 24, 2005 - link

    thanks for throwing in an AXP!
  • darfur - Thursday, March 24, 2005 - link

    This is just to the person who asked why their AA doesn't work in game. You need to disable the full screen glow effect, then AA should work just fine.
  • Aikouka - Thursday, March 24, 2005 - link

    "Oh, one more question. Why don´t you write anythink about the command rate (1T or 2T) you used - it´s quite important! And if I remember correctly a tRas of 10 is not optimal for NF4 boards. I think its 7 or so. Possibly there are similar problems with the P4-setup."

    #25, I noticed on my Soltek SL-K8AN2E-GR motherboard with Mushkin PC3200 that the tRas makes about no difference whether it's set at 8, 9, 10 or 11. The difference in memory benchmarks was quite negligible. However, command rate (going from 2T to 1T made an obvious difference.)

    One thing I'd like to comment on ... for WoW, I upgraded from my old GeForce FX 5200 to a MSI GeForce 6600GT, and let's say that WoW acts WORSE on my 6600GT. I did raise the graphics settings, but even then, the card has no problem handling them. What it tends to do is this:

    www.aikouka.com/wow_crap.jpg

    The screen is flipped horizontally and then rotated 180 degrees, not to mention a ton of textures are missing. This is one of the "better looking" WoW screw-ups that I get on my new card, and what's even better is when I get the beautiful stop error x8E on nv4_disp.dll :P. It's quite annoying.
  • Gholam - Thursday, March 24, 2005 - link

    By the way, there's another repeatable scenario that you can benchmark on: flight. Take a bat/gryphon/whatever, preferrably along one of the longer paths, and measure the FPS as you go from point A to point B. Since the path is the same, results should be more or less comparable - players running by below fall into the margin of error :)
  • Malladine - Thursday, March 24, 2005 - link

    Ok, i take that back. It did slow down noticeably when large numbers of people were around, say 100+

    AH was usually ok, but tping into IF took 30 secs to become smooth, i assume due to loading all the gfx
  • Malladine - Thursday, March 24, 2005 - link

    Before quitting a month or so ago, I ran this game from release @ 1600x1200 with everything maxed out and it was always smooth.

    AMDXP300+
    1gb PC2700
    9800 non pro
  • digit - Thursday, March 24, 2005 - link

    my bad, just checked, pci-x 16 slot, sorry

Log in

Don't have an account? Sign up now