In testing the performance of the X-Fi Xtreme audio, we've taken a couple of approaches. First, we ran RightMark's Audio Analyzer 6.0.6 to gauge the card's vital statistics, configuring the card as outlined in the instructions provided on RightMark's website specifically for the X-Fi. We were forced to limit testing to a maximum of 24bit/96kHz instead of 24bit/192kHz. Although the X-Fi can effectively output at these levels, it is unable to record signals beyond 24bit/96kHz. For CPU performance testing, we ran F.E.A.R. with minimum video settings with no audio at all and measured the frame rates. We then compared these to two more runs, one each with the X-Fi Xtreme (onboard) and X-Fi Fatal1ty (add-on card).
|The RMAA suite is designed for testing quality of analog and digital paths of any audio devices, be it a sound card, an MP3 player, a consumer CD/DVD player or an acoustic set. The results are obtained by playing and recording test signals passed through the tested audio path by means of frequency analysis algorithms. A more common mark is also provided for those unfamiliar with measured technical parameters.|
To clear things up in the above picture, the "Sound Blaster X-Fi" results on the left refer to the onboard solution, while the "SB X-Fi Audio" columns on the right are for the Fatal1ty card. We can see that the Frequency response was tightest with the X-Fi Xtreme at 16-bit/48kHz, but that range nearly doubles at 24-bit/96kHz. From that point on, just about every comparison went in the X-Fi Fatal1ty's favor with a lower reported Noise Level, higher Dynamic Range, and lower distortion and crosstalk values. With that being said, though, taking a closer look at each set of comparisons shows that the differences between the onboard audio and the add-on card are very slight. Typically we don't expect the values seen in the X-Fi Xtreme columns for integrated audio solutions. Take a look at Jeff's results with the integrated HD Audio on a Gigabyte GA-M59SLI-S5 to see what we mean.
Looking at the results, we see that there is no real performance advantage for spending the money on the carded version of the X-Fi Fatal1ty (or other variants) when compared to the integrated X-Fi Xtreme Audio solution. A mere two frames separates the two performance runs, impossible to notice during the demo. When comparing the X-Fi test runs to running the same game without audio enabled, the CPU utilization loss only amounts to about a four percent delta. MSI plans to continue using the X-Fi Xtreme chipset in their motherboards to come, which means external solutions could start going the way of the Dodo.