 |
|
3DsoundSurge takes a look at the impact of audio in Unreal Tournament |
|
|
|
|
|
| Sound News | Press
Releases | Archives | Week In
Review | Editorials
| Articles | |
|
|
|
||||
|
|||||
|
|
Audio Performance in the Unreal
Tournament demo Last updated December 15, 1999
For benchmarking we used "wicked" from http://www.voodooextreme.com/reverend/. We used two different computers. Computer A has a Celeron 300a o/c to 450Mhz, 128MB RAM, Windows 98 and TNT1 90/110 Mhz. The video configuration was 512x384, 16 bit colour, normal texture details for both the worlds and skins and no detail textures. Increasing the resolution to 640x480 only decreased the frame rate with 5% so I'm obviously not fillrate limited at 512x384. Mark (Computer B) used a Celeron 333a o/c 416 Mhz, 64MB RAM and a Voodoo2 12MB. He had high textures details for bth the worlds and skins and detail textures checked. The resolution is the same. In the normal audio configuration we both left all settings to default, which means high audio quality. The Vortex 2 boards were used with the 2040 reference drivers and theLive was tested with LW 2.1 You have several settings to play with in the advanced audio configuration including 3D hardware, reverb, filter, stereo, surround, output frequency, effect channels, digital music, CD audio and volume controls for the different settings. Surround is if you have a Dolby Prologic system and will use that instead of your soundcard's 3D positioning. Reverb is software reverb and has no effect when 3D hardware is set to true for the Live or Vortex2 card. I'm not sure exactly what filter does but I suggest you leave it enabled. I found that reverb, filter, stereo and surround has no effect on performance (framerates). I also found that headphones, 2 and 4 speakers always got the same result. Soundblaster Live
The performance impact of DS3D and EAX ranges from 11% to 7.6%. Also interesting to note that output rate seems to have no effect on performance when 3D sound is enabled. The performance impact of going to 32 from 16 effects channels is rather small and choosing just 8 is a rather big trade off so you're probably best off choosing 32 channels and an output rate of 48Khz. Digitial music has a performance hit of 2%. Computer B
Other EAX cards Computer A 4DWave-NX
The performance is very impressive, with 32 streams and sample rate of 22 kHz it's actually 5% ahead of the Live. Increasing the sample output rate to 48 kHz decrease the difference to 2.5%. It's still a rather small difference between 22 and 38 kHz samplerate so no reason to not choose 48 kHz. 368DSP
As you see the 368DSP using 32 3D streams and a sample output rate is 18% behind the Live. Increasing the sample output rate to 48 kHz increase the difference too 48%. It’s clear that 48 kHz isn’t the strong point of the 368DSP, even in stereo mode 48 kHz cause a significant performance pact. I also tried the 44.1 kHz output rate but the performance difference is only slightly better so probably best to stay with 22 kHz. The numbers of streams have little impact though so no reason to choose anything less than 32 3D streams. That performance is so poor in stereo mode (18.5% behind the Live) is a problem we have noticed in several other games even though it usually is just about 7% it's worse than any other card I've tried including the old SB16 cards. Yamaha's YMF744 (2013 drivers)
Not excellent performance but only 3.7% behind the Live and you get a performance impact of enabling 3D sound and EAX of about 11.5%. The difference between 16 and 32 streams are very small and there is really no difference at all between 22 and 48 kHz. ForteMedia's FM801 (2.03 drivers)
Not excellent performance but only 5.2% behind the Live at worst and you get a performance impact of enabling 3D sound and EAX of about 5.3%. That's less than many other cards but that's only because the stereo performance is pretty poor, or 8.8% behind the Live. This is in 4-speaker mode, 2-speaker mode and headphones are slightly faster in stereo mode for some unknown reason. None of the other cards show much of a difference. The difference between 16 and 32 streams are very small and there is really not much difference between 22 and 48 kHz. Vortex 2 old chip (all vortex2 boards except Aureal's upcoming SQ2500 use a chip with this performance) computer A
You can see that the performance hit for 3D sound with wavetracing ranges from 37% down to 23%. You can also see that the performance hit for digital music, 2.5% is not significant. Number of effects channel and output rate has no major impact unless you are below 16 effects channels and/or 22 Khz, but that's a significant tradeoff in terms of sound quality. You don't gain much by going above 16 channels and 22Khz though. The Console Commands It turns out that you can access the A3D console commands in a similar manner as in Heretic 2, Sin and Half Life. To understand what we are testing and to optimize your system you need to know the following: All toggle variables take a 0 for off and a 1 for on.
The following results are on Computer A with 16 effects channels, 22,000 Hz output rate, no digital music. The default s_maxpoly 500 was used in all cases except where noted.
Computer B - no console commands run yet
New Vortex 2 chip used in Aureal's upcoming SQ2500 computer B - no console commands run yet.
What’s interesting to note from these partial results is the wicked benchmark shows no performance difference between Aureal’s SuperQuad board and their new SQ2500 board which runs with an optimized Vortex 2 sound processor. If you read our SQ2500 preview you will know that on Computer B, the SQ2500 showed a 15.1 percent performance increase over the SuperQuad in Half-Life and a 2.4 percent performance increase over the SuperQuad in Heretic II. Aureal is currently looking into these results. The other thing the Vortex 2 results on both Computer A and B brings to light can be viewed as either disturbing or reason to be optimistic. On one side, the performance impact of wavetracing in Heretic II, which sounds great, is now almost non-existent after originally starting out in the order of 25 percent slower than direct sound (with no 3d). On the other side we have the older hybrid wavetracing implementation of Half-Life (hybrid because Half-Life has its own software reverb engine). It started out at around a 31 percent performance impact vs direct sound and on original Vortex 2 boards has been improved by a few percent to around 28 percent. Half-Life as noted is where the SQ2500 showed the biggest performance gains to date with wavetracing having only around a 17 percent hit on performance compared to direct sound. Right now the Unreal Tournament impact is in the order of Half-Life on the SuperQuad for both original and revised Vortex 2 boards. What we must hope for is that Unreal Tournament’s A3D 2.0 performance impact, which like Half-Life has its own software reverb engine, can be optimized to similar impacts that we see in Heretic II. Aureal is again looking into optimizations in Unreal Tournaments A3D 2.0 enigine. In the mean time, if you have a Vortex 2 card and are having unacceptable performance slowdowns after enabling 3Dhardware, you have another option to turning it off that will gain back quite a few frames per second. That is set s_reflect to 0 and s_occlude to 0. You won't have those effects but in the case of Machine A you will gain back almost 5 frames per second and still have the benefits of 3D audio. You can play around with other setting including s_maxpoly but keep in mind that all of these settings will have an impact on what you hear.
Both of us find that 3D sound adds significantly to the Unreal Tournament Experience. That's no surprise, but of course we would like more. That comes by way of three modes. Software reverb, EAX and A3D 2.0’s wavetracing effects. Our impressions so far is that the software reverb adds to the overall quality of sound in the game and will be there for even those people playing on a Sound Blaster 16 but with no 3d audio. The EAX implimentation sounds very much like the software reverb using the Live or 4DWave-NX while the 368DSP is of far less quality with current drivers. For both software reverb and EAX the aural environment changes are not jumping out on us. Wavetracing via A3D 2.0 gets the nod from both of us for the best sounding audio mode because of its dynamic aural changes in accordance with your location in the enviroment, but its not a night and day difference from software reverb and EAX.
Have comments or questions on this feature? Related Information Read our Half-Life benchmark report |
Articles A Gamer's Guide to Sensaura 3D Audio A Gamer's Guide to 3D Sound and Reverb APIs A Gamer's Guide to 3D Sound and Reverb Engines A Gamer's Guide to DS3D and A3D 1.x Virtual Surround Sound Technologies The Advantages of the Center and LFE channels Stereo to 4 or 5.1 Expansion Technologies Visit to the Creative EMU Technology Center Sound Card Reviews Hurricane Extreme Initial Review SoundBlaster Live! reference review SoundBlaster Live! Platinum 5.1 European Edition SoundBlaster Live! Platinum 5.1 America's Edition Sound Blaster Live! Player 5.1 Sound Blaster Live! X-Gamer 5.1 Hoontech SoundTrack Digital-XG Boostaroo Headphone Amplifier / Splitter Sound Blaster Live!Drive I and II Sound Blaster Creative Digital I/O 2 |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|
|
All content, design and work is ©
2001 - 3D Sound Surge Please respect the copyrights of the articles and writers herein. All copyrights are enforced by 3DSS.