Integrated audio turned up to 11
Most motherboards combine the chipset's integrated audio controller with a separate codec chip that handles analog-to-digital and digital-to-analog conversions for the onboard ports. This approach is very economical, but the sound quality with analog speakers and headphones tends to be fairly marginal, which is why we typically recommend discrete sound cards to folks with decent speakers or headphones.
The G1.Sniper 5's audio implementation is sort of like an integrated discrete card:
Instead of ye olde Realtek codec, the Sniper sports a Creative Sound Core3D chip that handles controller and codec duties. The associated circuitry is isolated from other onboard components to reduce interference, and it's laced with audio-specific Nichicon capacitors that purportedly improve sound quality. There are two amplifier chips onboard, too. One is tied to the front-panel headphone out, while the second drives the same sort of jack in the rear cluster.
We've seen headphone amps on motherboards before. However, the Texas Instruments OPA2134PA chip attached to the Sniper's rear headphone out is socketed, allowing users to swap in different OP-amps. Gigabyte includes a second OP-amp—Analog Devices' AD827JN—plus the oversized tweezers required to yank the chips out of the socket. The mobo maker also sells a separate OP-amp kit with three additional chips: Linear Technology's LT1358, National Semiconductor's LM4562NA, and Texas Instruments' OPA2111KP. Each amplifier has a slightly different acoustic profile, and we'll explore their impact on audio quality in a moment. First, we should address Creative's contribution.
The Sound Core3D identifies itself as a Recon3Di audio device. It comes with Creative's Sound Blaster Pro Studio software, which serves up surround-sound virtualization for stereo devices, a "scout" mode that makes footsteps easier to hear in games, and echo cancellation to improve voice input. The software also offers other ways to mess with the audio signal, but surprisingly, there's no real-time encoding for multi-channel digital output. Folks who want surround sound in games will have to use the analog outputs.
We weren't able to conduct blind listening tests with the Sniper, but I did spend a fair amount of time listening to music with various configurations. These more casual tests were conducted with mid-range Sennheiser HD 555 headphones and a selection of tracks from Neil Young, Radiohead, and The Heavy.
First, I compared the unamplified front-channel output to the amplified headphone jack with the stock OP-amp installed. These two were easy to tell apart; the unamplified out sounded dull and muddled, as if it were working with a limited frequency range. With the OP-amp lending a hand, there was more separation between the various elements in each track, and the sound quality improved noticeably overall. Radiohead was more poignant, The Heavy was more soulful, and Neil Young was more engaging. The amplifier added a crisp liveliness and improved clarity, seemingly without any drawbacks.
Encouraged by these initial results, I grabbed Asus' Xonar DSX sound card off the shelf. This card was the favorite in our last round of blind listening tests, and it sells for only $60. It's also our recommended upgrade for folks seeking superior sound quality to typical integrated audio.
The Xonar and the Sniper's amplified out sounded more closely matched than the two onboard ports from the first comparison. That said, the low end of the spectrum definitely kicked harder and deeper on the Xonar. This bassy grunt added a balanced fullness that was absent on the Sniper. Higher up the spectrum, the clarity I noticed on the Sniper in the first round of tests felt artificial and overly sharpened next to the Xonar's more natural output.
For an encore, I started swapping OP-amps to see if I could notice their impact on the Sniper's sound quality. This task required shutting down the system to switch the chips. Perhaps because of the extra time and effort involved, it wasn't as easy to pinpoint minute differences between the various options. Most of them sounded very similar, but the Linear Technology OP-amp stood out; it seemed to have brighter vocals and more muted bass than the others. None of the OP-amps made the Sniper sound as balanced and natural as the Xonar.
We also tested analog output quality objectively with RightMark Audio analyzer. All the configs from our listening tests were plugged into a separate Xonar Phoebus sound card, which captured their output of a 24-bit, 96kHz test track. RMAA grades analog signal quality on a scale between "very poor" and "excellent." We've translated those values to a numerical scale that starts at low of one and peaks at six. Higher values are better.
The Xonar scores the best overall. The unamplified line out doesn't look too bad according to these coarse measurements, but there are more obvious differences between it and the others if we look at the accompanying graphs. Click on the buttons below each one to switch between the line out, which is displayed by default, and the alternatives.
Total harmonic distortion
These graphs are a little indulgent, but they nicely highlight the differences between the Xonar DSX and the rest of the configs. The discrete card has lower noise levels, less distortion, and a broader frequency response than the Sniper's onboard audio.
For the most part, the OP-amps look like an improvement over the unamplified line out. However, they have shallower frequency responses at the low end of the spectrum, which explains the punch missing in our listening tests. The frequency response of the Linear Technology LT13581 amplifier falls off at the higher end of the spectrum, too.
I didn't notice any issues with the integrated audio in our listening or signal quality tests, which were conducted with the system idling at the Windows 8.1 desktop. Firing up a graphics load produced a noticeable buzzing sound, though. This buzzing was audible at normal volume levels with not only the Sennheiser headphones, but also cheap earbuds.
Battlefield 4 and the Unigine Nature benchmark reliably produce the buzzing noise. It's only apparent on the Sniper's amplified output, and it's not just my imagination. RMAA captures the behavior nicely. The following graphs come from "loopback" tests that route the motherboard's audio through the onboard line input. (They aren't directly comparable to the graphs above as a result.) The Nature benchmark provided the graphics load.
The amplified output has more distortion and higher noise levels with the graphics load running. The audible buzzing is apparent with hot-clocked GeForce GTX 680 graphics cards and also with a low-end Radeon R7 250. Further investigation with the motherboard installed inside a Corsair Obsidian Series 650D enclosure yielded similar results, this time with the addition of fainter feedback during web browsing and even when moving the mouse rapidly over icons. I've reproduced the buzzing with different system components and with the rig connected to a separate wall socket.
This behavior is the opposite of what we'd expect from onboard audio that's supposed to be isolated from interference. Gigabyte has been working with us to pinpoint the issue, but we haven't narrowed it down yet. We actually sent our test system to the company after it was unable to reproduce the problem in its labs. Gigabyte tells us it has replicated the buzzing with our hardware, but only with a GeForce GTX 680 installed. We've also noticed that a Newegg user review mentions similar sound interference. The investigation continues, and we'll update this article as we learn more.
|Nvidia's GeForce GTX 1080 graphics card reviewed||206|
|Sunglasses Day Shortbread||4|
|Color TV Day Shortbread||65|
|Oculus removes hardware check DRM from Rift exclusives||17|
|Only one month to go before the "second-10th" TR BBQ||8|
|Deals of the week: an affordable Core i7-6700K and gaming gear||20|
|3DMark is getting a full-featured DirectX 12 benchmark||30|
|Swim-a-Lap Day Shortbread||19|
|Steam Summer Picnic sale is all about tasty games||42|