Separate names with a comma.
Sound Cards item created by , May 5, 2010
Pros - Sound quality, low noise, built well, decent components.
Cons - Asus Xonar driver can be buggy, switching output sources can be better.
This is a decent card for the price. This card comes with good but not great components. If you want better you will have to dish out more $$ but for the $67 I spent for this I think the quality is more than adequate. This was a breadth of fresh air coming from the Creative camp which has been problematic ever since direct sound was dropped from Windows. This is not to say that the D1 had its fair share of audio issues in games like Fallout 3 but then again I had these same issues with the Creative Fatality Pro which leaves me to believe that this may be an issue with the game engine not supporting add on sound cards. Holding the D1 in your hand will give you an idea of the workmanship that went into this device. While the D1 is capable of accelerating audio in hardware I personally found the Xonar Unified driver in Low DPS Latency mode to be more responsive than the Asus driver. Low latency mode removes hardware acceleration on the Xonar AV 100 chip on this card but with most modern games off loading sound processing to the CPU, hardware acceleration is not a must. In my experience I have few issues with game audio in Low DPS Latency mode. If you have a processor released over the last five years, you would not have to worry about loosing significant frames due to offloading of sound to the CPU. Whats important to look at from a sound perspective is the DAC and Op-Amp which are responsible for the sound quality on the analog outputs. The D1 boasts a front channel Audio DAC (Cirrus Logic CS4398) with a signal to noise ratio of 120db which is nice and a JRC NJM5532 Op-Amp which is good at this price point. I am looking at upgrading from the trusty D1 to the Xonar STX to power my Sennheiser HD 380's and Senn HD 598 as the STX has a headphone amp capabable of driving cans up to 600ohms. A low budget alternative to the STX would be the Xonar DG priced at $30. Bear in mind that the DG while boasting a headphone amp will not have enough power to drive very high end cans like the Sennheiser HD 650 or 700 (The Xonar D1 does NOT have a headphone amp). Other alternatives to the Asus line up are the Omega series but while I have herd good thing I don't have any first hand experience with them. I have seen some folks talk about distortion/noise on the front panel analog out, I had this issues with the D1. However I isolated the issue to not the D1 but an uninsulated cable in my case, when I upgraded my case to an Antec 600SE the first thing I noticed was the front audio out wire was thick and well insulated, thereafter the noise/distortion was not present.
Pros - Excellent DAC for the price
Cons - Too high output impedance for many headphones
This review focuses mainly on the performance of the card from a technical point of view. It most likely also applies to the Xonar DX, which is basically the same card, but with PCIe interface instead of PCI.
The strongest point of these cards is the (CS4398 based) DAC, which supports sample rates up to 192 kHz, and 24 bit resolution (I tested mainly at 44.1 to 96 kHz). The specs claim 116 dB A-weighted signal to noise ratio (112 dB for the input) relative to the 2 Vrms full scale output, and 0.00056% harmonic distortion at 1 kHz/-3 dB. As far as it can be determined with loopback tests, the card can come fairly close to those values in an electrically "quiet" PC. The distortion is also very low with a variety of more difficult test signals, and the DAC can handle a 0 dBFS signal at full volume (the volume control is apparently digital) without any problems. It has some additional headroom for signals that need more than 0 dBFS level in continuous time. Standard 44.1 kHz CD sample rate is also handled well, without significant artifacts. The DAC uses a minimum phase lowpass filter. In the 20 Hz to 20 kHz audio range, the frequency response is flat within less than 0.1 dB relative to 1 kHz. The output signal is not inverted (polarity is preserved).
To keep this review short enough, I do not get into further details, but so far I did not find anything to complain about the DAC performance, especially for a <$70 card. Of course, the higher Xonar cards should be even better.
The analog input / ADC is not as good as the DAC, but still quite decent for the low price, and also supports 192 kHz/24 bit. Unlike the DAC, it does not seem to handle a full scale signal perfectly (the distortion increases somewhat already 1-2 dB below clipping level), and on the card I have the left channel has a small amount of signal dependent noise that is not on the right channel. The input has an impedance of about 3.8 kΩ, preserves polarity, and does not capture DC.
The main disadvantage is the headphone output. While it has quite low distortion even when driving low impedance loads at the maximum level, indicating that at least some attempt was made at amplification, the output impedance is about 100.5Ω according to my tests. This is too much for most popular low impedance headphones. With the 2 Vrms maximum output voltage (actually, I measured about 3% less), the power into a 100 Ω headphone load is limited to 9-10 mW, and about half the maximum power into 16 Ω and 600 Ω. The front channel output is coupled by 220 μF capacitors, this slightly rolls off the bass with low impedance loads (about 0.5 dB at 20 Hz/32 Ω).
The limitations related to driving headphones are the main reason why the value rating is not 5.
The other analog outputs are of slightly lower quality than the front channel, and use CS4362A DAC chips. Interestingly, these have much lower output impedance (only a few Ω), but are nevertheless not well suited to driving headphones because of the very small (22 μF) coupling capacitors that audibly roll off the bass even with 250 Ω headphones. Also, the distortion of these channels does increase significantly with low impedance loads at high output levels.
The card supports the use of the front panel, although this is not recommended if the noise level is to be kept as low as possible.
UPDATE: detailed measurements are available here.