Date: April 16, 2008
Author(s): Rory Buszka
Not long ago, we received word from ASUS that they had managed to incorporate EAX effects into the drivers of their Xonar cards. Soon afterward, Creative told everyone that was false, and that ASUS “EAX” was not true EAX. Who’s telling the straight story, and will you even notice a difference?
When we first reviewed ASUS’ Xonar D2 sound card in September of last year, ‘impressed’ didn’t begin to describe our initial reaction – quite simply put, it rocked our faces off. Several other Techgage authors quickly picked up cards of their own, and were similarly stunned by the competency, clarity, and fidelity. We were sure that Creative’s X-Fi cards had finally met their match, but one gripe that we repeatedly heard from our readers was lack of support for Creative’s EAX versions 3.0 and above.
EAX, now in version 5.0, is a proprietary Creative Labs technology, but versions 1.0 and 2.0 of the specification were made freely available for use by other manufacturers, a practice that ceased with version 3.0. It was a clever scheme – versions 1 and 2 of EAX were hugely popular, and gamers hoping to take advantage of the technology’s newest iteration would have to choose Creative hardware exclusively.
So what exactly is EAX? The Environmental Audio eXtensions (EAX) is a proprietary sound API that allows developers to use the onboard digital signal processing capabilities of Creative’s audio chips to produce more realistic virtual environments through simulated environmental reverb and occlusion effects.
The centerpiece of EAX is hardware-based signal processing, controlled directly through the software itself, which is why we were surprised to learn that ASUS had released a new driver revision for their Xonar cards, enabling “EAX 5.0 Support”. We initially assumed that ASUS’ driver wizards had come up with a way to mimic the Creative X-Fi’s onboard DSP through the driver software itself, in the same way that Dolby and DTS processing are handled.
In a sense, we were correct. To finagle the functionality of Creative’s EAX 5.0 from the ASUS AV200 (D2, D2X) and AV100 (DX) audio processors (better known as the C-Media CMI8788 OxygenHD chipset), which rely upon software-based DSP, ASUS reworked the DS3D GX extension set to provide all the functionality of Creative’s EAX Advanced HD 5.0 through the DirectSound engine itself – in software. That means 128-voice 3D positional audio and environmental reverberation. ASUS admits that this isn’t true ‘EAX’ – rather an emulation of the functionality of EAX 5.0 through the software driver.
According to ASUS, the benefit of this approach is that it removes the need for an abstraction layer to provide a direct connection between DirectSound and the audio driver. Windows Vista eliminated the Hardware Abstraction Layer (HAL) of Windows XP and 2000, severing this direct software link between drivers and applications, and leaving owners of Creative X-Fi and Audigy cards without surround sound or EAX effects under Vista.
Creative responded to this problem with the ALchemy software upgrade, which mapped DirectSound3D calls to equivalent OpenAL calls. Because OpenAL is third-party software, it was not crippled by the loss of the HAL, and this approach re-enabled the lost functionality. But while this software fix is available to X-Fi owners at no extra charge, Audigy card owners will have to cough up $9.99 simply to download the compatible fix. In addition, not all game titles are supported. ASUS says their Xonar sound cards function without this limitation, even while using EAX functionality.
It wasn’t long at all before Creative had something to say about ASUS’ claims regarding EAX 5.0 functionality, however, and we soon received an email from Creative Technology’s VP of Marketing Communications, Phil O’Shaughnessy, claiming that by mapping EAX reverb functions to their DirectSound3D GX (“Game eXtensions”) effects engine, ASUS was not providing a “genuine” EAX experience, and was “misleading” customers by claiming to support EAX 5.0 functionality.
What we’ve gathered from Creative’s statement is that the ASUS approach simply allows the Xonar sound card to render 128 simultaneous voices, but it’s still limited to the EAX 2.0 effects palette (Environmental reverb and occlusion). Here’s a quick breakdown of how the EAX versions compare to one another:
ASUS countered that while their DS3D GX 2.0 engine simply emulates the environmental reverb effects of EAX 5.0 in software, instead of relying upon a dedicated DSP chip, they believe the software-based DSP approach is superior because it can work directly through DirectSound3D, instead of requiring an ALchemy-style solution based on a third-party software API, so it’s automatically compatible with all DirectSound3D games under Windows Vista, while ALchemy is only verified to work with a limited number.
In addition, though ASUS states that they “respect the capability of a dedicated DSP processor to offload the CPU work”, they found that performing the signal processing associated with positional 3D audio in software instead of on dedicated hardware had a minimal impact on overall system performance during gaming, and question the necessity of hardware-accelerated positional audio and environmental effects on modern gaming PCs.
In their response to Creative’s criticism, ASUS also jabbed at the exclusive nature of Creative EAX, writing that their implementation was not “locked behind specific X-Fi cards”, and that software DSP effects such as those provided by DS3D GX 2.0 offered advantages in “flexibility and universally guaranteed user experiences.”
Taking another look at the above list of EAX versions and their features, however, there are other EAX functions that ASUS does not claim to have matched with the new DS3D GX 2.0 engine. Among them are EAX Environmental Panning, simultaneous environments, flanging, distortion, ring modulation, EAX Voice, and Environment Occlusion. They might be in there (or their equivalents), but we haven’t heard.
So, to recap: While ASUS support of EAX 5.0 isn’t true “EAX” – they haven’t licensed anything from Creative – the DirectSound3D GX 2.0 extension set aims to free the positional 3D audio market from Creative’s grip by performing the same functions in EAX 5.0-enabled games, without the Vista compatibility issues of hardware-acceleration, and the half-baked ALchemy solution.
This article seeks to answer one question – Are you likely to notice a difference between ASUS’ DS3D GX 2.0 and Creative’s EAX Advanced HD? We’ve got cards, and we’re not afraid to use them…
Before we begin with my own listening tests, I’d like to toss in a word about exactly how I’ll be going about testing this. Some of the most sophisticated instruments for audio testing are right on the sides of your head – your ears have a frequency response that ranges from 20 Hz to 20 kHz, and a dynamic range that spans over 120dB, with a threshold of hearing that begins at about 20dB, depending on how you’ve abused your ears over the years.
Yet, many people don’t realize that they can trust their ears just as well as they can trust their eyes or nose. So to test ASUS’ claims against Creative’s counterclaims, I’ll be relying principally on my ears to detect differences in the quality of each card’s positional and environmental effects.
When we approached Creative with the idea for this article, they happily obliged by sending along a Creative Sound Blaster X-Fi XtremeGamer Fatal1ty Professional card to use in testing. Besides being a real mouthful to say, this card is targeted toward gamers who seek robust audio hardware, but can do without a plethora of accessories.
Representing the ASUS side in this battle is the high-end ASUS Xonar D2 sound card, which I reviewed in September of last year, and currently own. Another advantage of ASUS’ software-based DSP approach is apparent when considering that all it takes to add the improved DS3D GX 2.0 functionality is a driver upgrade, which I installed on my machine before beginning the testing. By using the higher-quality Xonar D2 instead of the newer but more value-oriented DX model, I’m giving the Xonar side in this fight the best representation it can possibly have against the X-Fi.
All testing for this article was performed using my reference machine, a modest gaming rig that also features a silent cooling system controlled by a programmable microprocessor controller, reducing background noise. In addition, all tests were performed in my quiet room, with no external noise sources active during testing.
To monitor the audio system’s output, I’m using my reference headphones, the Beyerdynamic DT 770. These stereo headphones fall within the $250 price category, and are highly revealing of any flaws in the source. They aren’t surround, but they produce a coherent sonic image, and actually work extremely well for positional audio.
Game testing was performed under Windows XP, to avoid any potential compatibility issues with Windows Vista that could get in the way of critical listening. All games were tested with EAX functions active, and the appropriate device selected in the Audio control panel.
To test each game, I first played a level with the Xonar card active, then the X-Fi, repeating the same part of the level for each card under test. For this article, I’m testing three games: 2K’s BioShock (EAX 5.0), Eidos’ Kane & Lynch: Dead Men (EAX 3.0), and Human Head’s Prey (EAX 5.0). In each game, I’m listening for three things: 1) natural-sounding reverb effects, 2) accurate occlusion, and 3) smooth transitions between sonic environments.
With those details out of the way, let’s move on to the first game test: BioShock. And let no one accuse me of starting off easy…
2K’s BioShock plays like a first-person shooter, reads like a horror thriller, and paints rich visual and sonic landscapes with a variety of materials and textures. It’s one of only a handful of EAX 5.0-enabled titles, designed to showcase the full capabilities of an EAX 5.0-capable sound card. The underwater city of Rapture comes to life with sound, as the sharp reports of gunshots echo off glass walls, and dialogue of the “Splicers” – genetically-altered once-human monsters reverberates throughout cavernous rooms, just to name a couple of the sonic treats that await gamer-audiophiles in this popular title.
Beginning with the Xonar card, plus an updated driver file that fixes a bug that caused BioShock not to recognize that an EAX-capable sound card was available, I played the first level from the ill-fated plane crash to the Medical Pavilion, absorbing the sonic and visual scenery along the way. When entering the lighthouse after swimming from the flaming wreckage of the plane, the reverberation effects were plainly discernible, and seemed to pan around me as I descended the staircase to the bathysphere, which had its own reverberant sonic signature.
The reverberation effects in BioShock on the Xonar card added a sense of spaciousness and ambiance that was different with each environment I entered, just as one would expect from an EAX title. Footsteps in carpeted rooms didn’t echo with the same delay as they did in long, hard-floored hallways, and it was easy to place sound sources in space. Near the end of my play in the level, as shootouts became more complex, the Xonar still rendered reverberant effects that retained good separation of the sound sources.
With the Xonar card in BioShock under Windows XP, I did notice something troubling, despite the fact that I was using an updated driver file that had been intended to fix issues with BioShock – intermittent crackles and pops littered the audio at random intervals. In addition, when the overall sound levels were low, soft background noise could be heard at times, and some especially quiet sound effects could sound slightly garbled. It sounds like ASUS’ driver wizards have a bit more work to do.
Switching to the X-Fi card and repeating the same part of the first level, I noticed that the quality of the reverberant effects was improved somewhat. I particularly noticed that the sense of distance had increased – sound sources located further away sounded significantly further away than they had with the Xonar card. Also, transitions between acoustic spaces sounded a bit more natural. Instead of sudden increases in reverberation as I passed through a doorway, the space seemed to open up in front of me as I passed into it. Each “Big Daddy” robot’s bestial grunting seemed to excite a larger volume of air in the room, and explosions seemed to burn away into thin air. Overall, I noticed better sonic refinement.
So, who wins this round? It’s clear that the Xonar card is capable of performing under an EAX 5.0 game, but while the Xonar card sounded clearer and crisper to my ears, with greater dynamic range, the X-Fi XtremeGamer sounded slightly more realistic. The main difference I noticed was with the quality of the reverberant effects – this is something that ASUS can definitely improve upon with future driver releases, in addition to eliminating the occasional pop or crackle that I noticed with the Xonar card, and I hope that ASUS will continue to pursue improvement and refinement of their reverb algorithm.
Next, we’ll take a look at each card’s performance with an altogether different sort of game, Kane & Lynch: Dead Men.
Eidos’ Kane & Lynch: Dead Men is an EAX 3.0 game, but according to ASUS, EAX 3.0 functionality should still be supported under DS3D GX 2.0. This title is another one that surrounds the user with action, however its positional audio effects are more subtle than BioShock’s. The first chapter, “Impact”, begins with Kane and another prisoner (who will later be identified as Lynch) escaping from a prison van and being rescued by masked gunmen through a hail of gunfire.
With the Xonar card, during the first chapter of Kane & Lynch, environmental sound changes were so subtle that they might be easily missed. However, the audio of the Xonar sound card was clean and free of the random pops and crackles that were noticed while testing in BioShock. As Kane runs into the alley after being rescued from the back of a truck, the sound of the gunfight in the street is occluded by the wall, and the amount of reverberation increases slightly.
The most noticeable change in environmental acoustics during the first chapter occurs when Kane, the other prisoner, and the gunmen make a stand in an abandoned donut shop. As Kane enters the smaller space covered in hard surfaces, there’s an audible change in the sounds of gunshots originating from outside the structure, as well as an additional reverberant character to voices.
In the second chapter, “Trial”, the abandoned shopping mall atrium sounds positively cavernous, with a long delay on early reflections. As police officers stream into the space and begin firing and shouting, the number of simultaneous voices climbs, yet echoes sound clear and well-defined even within the mishmash of direct and reflected sound.
In the case of Kane & Lynch, because the in-game EAX effects themselves are so subtle, there was virtually no noticeable difference when the game was run with the X-Fi card instead of the Xonar. I was a bit disappointed; I tend to prefer it when these things are clean-cut, but in this case there was no clear winner between the two sound cards.
If I had to pick which card I preferred, I’d have to go with the Xonar this time, but in reality it’s too close to call. So why test Kane & Lynch, if it’s only an EAX 3.0 game? Simply to prove that DS3D GX 2.0 also works within EAX 3.0 games, and can deliver an indistinguishably similar result.
Next, we’ll take a look at a third and final game, one which Creative suggested that we use in our testing: Human Head’s Prey.
I’m convinced that some games are developed simply to give somebody something to do for a while. Human Head’s Prey is one of those games – not much more than an interesting combination of tired first-person gaming clichés. However, this game makes use of EAX 5.0, and was recommended highly by Creative for testing, so I decided I’d give it a whirl.
While the EAX 5.0 performance of the Xonar card left a bit to be desired in BioShock, the Xonar performed admirably in Prey, dramatically heightening the ambiance of the half-living alien spaceship and the perception of location and distance of sounds. To test Prey, I played through the “Downward Spiral” level, which takes the game’s main character on a meandering path downward through the alien spacecraft.
EAX makes a huge difference – just out of curiosity, I tried playing the level through without EAX enabled, and the game audio was drab and lifeless. Yet with it enabled, environments took on spaciousness, depth, volume, and air. The Xonar’s EAX 5.0 performance in Prey was problem-free, with none of the random noise that plagued BioShock.
Switching to the X-Fi, I noticed that the effects, again, possessed better detail, and sounded a little bit more realistic. But only a little – in fact, the EAX effects in Prey with the Xonar card sounded almost exactly as they did with the X-Fi. As with Kane & Lynch, the difference between the two cards is so slight that you’re likely not to even notice a difference. That alone should be enough to scare Creative.
See the next page for my final thoughts on the matter.
Perhaps it’s the timing, or the fact that established industry giants are going toe-to-toe, but the recent face off between Creative and ASUS over EAX 5.0 functionality in the new Xonar sound card drivers from ASUS has attracted quite a bit more media attention than I’m used to seeing devoted to PC audio hardware.
3D positional audio with the environmental effects that EAX provides can make a huge difference in your enjoyment of games – they not only help you feel like you’re inside an acoustic space, but the cues from those environmental reverb effects can help you determine where other sound sources are located relative to yourself. With this in mind, it’s easy to see why 3D environmental audio is a subject that matters to gamers.
Creative’s EAX technology has dominated the 3D audio landscape for quite some time now, despite the fact that every version of the specification after 2.0 was proprietary. But ASUS has managed to incorporate the same basic functionality of EAX 5.0 (128-voice environmental reverb and occlusion) into their DS3D GX 2.0 driver extension, challenging Creative’s dominance and forcing us all to ask if the days of proprietary, hardware-bound environmental DSP could be behind us.
In the three games I tested, I found that while EAX effects on the Creative X-Fi had a slight edge in sound quality when compared to the Xonar, it was hardly a night-and-day difference. The strange artifacts I heard under BioShock were likely due to the fact that I was using a patched driver to conduct the tests – ASUS sent me an updated file to be included in the driver installation that would allow EAX to be enabled within BioShock.
It’s clear that the Xonar drivers and ASUS’ DS3D GX extension set both need a bit more work to bring them up to the level of Creative’s hardware-based DSP, but the success of the ASUS card under both BioShock and Prey definitely proves that software-based DSP effects are coming of age.
The significance of the testing and conclusions detailed in the preceding three pages is thus: While ASUS hasn’t managed to de-throne the X-Fi’s hardware-accelerated EAX environmental and positional audio effects where sound quality is concerned, they’ve come extremely close, and have managed to duplicate much of the functionality of EAX 5 to boot.
In fact, as far as practical conclusions are concerned, in most games you’re not even likely to notice a difference, and if you do, it’s a slight one at that. There was also no perceptible effect on smoothness of gameplay when using the ASUS Xonar sound card, despite the fact that my ‘reference’ rig might have been cutting edge, oh, say, a couple years ago (with the exception of the video and audio subsystems).
If there’s one thing we see over and over again in this industry, it’s that competition produces innovation, or at the very least, consistent improvement. Creative appears to be very interested in defending their position as the heavyweight of the PC audio market, but at this point that effort seems to be based largely in attacking and undermining the performance claims of competitors, instead of developing and delivering improved solutions.
Here we begin to see the true gravity of ASUS’ software-based DSP approach: When the technology evolves, companies like ASUS need only to update the driver software, instead of embarking on the costly pursuit of developing more powerful hardware DSP components.
Creative might yet succeed in holding off its tenacious competition for the time being, but a look at the short list of available EAX 5.0 game titles when compared to the sheer number of EAX 4.0 and 3.0 titles suggests that interest in EAX technology might be waning, in favor of non-hardware-specific 3D positional and environmental audio solutions.
ASUS appears to be ahead of the curve there, as their solution provides them with the agility to adapt to whatever standard the game industry throws at them next. And as we’ve all seen, trying to stifle the competition is ultimately a losing battle – Creative will eventually need to respond either with improved hardware, or with a software DSP solution of their own. With these latest developments by ASUS, one thing’s for darned certain: the world of PC gaming audio just got a lot more interesting.
If you have a comment you wish to make on this article, feel free to head on into our forums! There is no need to register in order to reply to such threads.
Copyright © 2005-2019 Techgage Networks Inc. - All Rights Reserved.