X-Fi USB DAC

INTRO: I’ve been testing portable reasonably priced USB partly in the hope of finding one that’s a good match for pairing with my 02 headphone amp. The C-Media based inexpensive USB DACs from FiiO, Turtle Beach and Syba were all disappointing. What happens if you spend a bit more money ($35) for a similar DAC from one of the biggest names in PC audio? Creative offers the X-Fi Go Pro as their “thumb drive” portable DAC. How does it measure up to the C-Media DACs already reviewed?

CREATIVE SOUND BLASTER X-FI GO! PRO THX TRUESTUDIO SB1290: For all the money Creative spends on marketing, and after Apple elegantly slaughtered them in the portable player market, one might think Creative could at least adopt user-friendly product names. I think they only left out “Turbo”, “Audiophile” and “Ultra”. Fortunately the DAC is simpler than the convoluted marketing mess. There’s a whole range of X-Fi products, but I’m just going to call this one the “X-Fi” for simplicity. It’s a lot like the Syba USB DAC with an integrated USB plug, 3.5mm headphone/line out, microphone jack, LED, and nothing else. It does, however, come with a USB extension cable which is handy for accessing recessed USB ports or where the DAC might block other ports. And it has a cover for the USB plug making it look a lot like a Zippo cigarette lighter—more clever Creative marketing targeting old school smokers.

OOOOH THX! Once upon a time THX was a rigorous certification process reserved for some of the best home theater gear. THX amplifiers, for example, were required to deliver certain levels of power, over a wide bandwidth, at low distortion, with all channels driven. Now, sadly, it seems THX is just a logo and/or a software license that manufactures buy to slap on most anything. THX certified wine opener anyone? For the X-Fi it’s about the supplied software rather than the hardware. THX has put their name on a suite of me-too digital audio processing that’s not very different than similar products from SRS Labs, Dolby, and many others including the stuff Creative did on their own years ago. They’re just DSP routines that alter your music in assorted artificial ways. Some might find it useful—especially for gaming or movies with headphones. These sorts of DSP effects, to my ears, nearly always sound cheesy and artificial much like the “cathedral” mode on your home A/V receiver. The THX licensed DSP is is part of the bloatware Creative wants you to install on your PC and has nothing to do with the DAC hardware itself.

BLOATWARE SKIPPED: The last time I installed Creative software on my PC it wrecked all sorts of havoc. It’s even worse than iTunes. Creative’s massive install took over file associations, became the default media player, installed various video CODECs that broke other CODECs, tried to index my music collection and promptly locked up, etc. With Creative having already burned that bridge with me I simply used the X-Fi with native Windows drivers.

WINDOWS INSTALLATION & 24 Bit? The X-Fi installed smoothly in both XP and Windows 7 (minus THX DSP, 3D effects, etc.) without needing any third party drivers. Windows 7 reported it as a “Sound Blaster X-Fi Go! Pro”. Only one sample rate is available, 44100, but interestingly there’s a choice for 16 or 24 bits. Is the X-Fi really capable of 24 bit operation? If so that would be a significant plus as it allows using the PC (or player software) volume control and still, in theory, maintain close to 16 bit resolution to the DAC.

SUBJECTIVE SOUND QUALITY: Running the X-Fi into my 02 headphone amp the sound quality was good with no obvious problems. There was, however, some audible hiss. Connecting headphones directly the results depended on which headphones. My DT770 Pro 80s sounded fairly good, there was not enough power for my HD650s, and my UE SuperFi 5s sounded pretty bad and also revealed audible hiss.

MEASUREMENT SUMMARY: The overall results are substantially better than the C-Media offerings and generally respectable for a $35 DAC. The weak areas are noise, low level linearity, low frequency roll off into 32 ohms, and a somewhat high output impedance. I would not use the X-Fi with any headphones under 60 ohms if you want the best sound quality. It also lacks enough power for a lot of high impedance headphones. I ran several tests with both the 16 and 24 bit modes and could not detect any difference. The letter grades range from A to F where A is excellent and F is Fail (unacceptable):

Measurement	X-Fi Go	FiiO D5	CM119	UCA202	TB Micro II
Freq. Resp. 10K	+/- 0.4 dB A	+/-1.5 dB C	+/-1.0 dB B	+/- 0.1 dB A	+/- 1.0 dB B
Freq. Resp. 33 ohms	+/-5.0 dB D	+/-1.5 dB C	+/-6.0 dB F	N/A	+/- 1.8 dB C
Output Imp Ohms	7.8 C	0.72 A	5.9 C	47 F	0.95 A
Max Output 10K	1.0V	1.5V B	0.95V C	1.12V B	1.34V B
Max Out 33 Ohms	0.75V C	1.4V B	0.68V C	N/A	1.26V B
Max Power 32 Ohms	18 mW C	61 mW B	14 mW C	N/A	50 mW B
THD+N 0 dBFS 10K	0.007% A	0.24% C	0.035% B	0.008% A	0.14% C
THD+N 100hz 10K	0.007% A	0.08% C	0.035% B	0.007% A	0.025% B
THD+N 1Khz 10K	0.007% A	0.08% C	0.035% B	0.007% A	0.02% B
THD+N 1K 33ohms	0.009% A	0.08% C	0.095% C	N/A	0.12% D
THD+N 10Khz 10K	0.009% A	0.04% B	0.090% C	0.009% A	0.11% C
IMD CCIF 10K/33	0.004% A	011% D	0.028% D	0.005% A	0.028% D
IMD SMPTE 10K	0.0005% A	0.80% D	0.012% B	0.002% A	0.02% B
Noise A-Wtd dBu	88.9 C	-90.0 C	-89.0 C	-88.8 C	-93.8 B
-90 dBFS Linearity	1.5 dB B	0.7 dB A	0.9 dB A	3.8 dB C	0.8 dB A
USB Jitter Jtest	VG B	Poor D	Poor D	VG B	Poor D

 

FIRST CLASS:

• Low distortion even at high frequencies.
• Flat frequency response into 10K
• Well behaved DAC filtering
• Relatively low jitter
• Highly portable
• Reasonably priced ($30 – $40)

ECONOMY:

• 24 bit option appears to be useless
• Marginal 7+ ohm headphone output impedance
• Rolls off deep bass into headphones below 60 ohms
• Marginal noise performance and dynamic range
• Marginal low level linearity
• Only 1 Vrms maximum output into 10K

O2 COMPANION: Used with an amp like the O2 with its own volume control, the X-Fi is a reasonable DAC with very flat frequency response, relatively low distortion and very low jitter. If you leave the PC/software volume control at maximum it has enough dynamic range to be quiet enough in most applications. If you want to use software volume controls, however, you might hear some noise in some circumstances.

BOTTOM LINE: The X-Fi Go performs vastly better than any of the three C-Media DACs I’ve tested. It gives the UCA202 a good challenge with similar distortion and even lower jitter while being significantly more portable. The downsides are higher noise, some low level linearity error, and not being suitable for many headphones under 60 ohms or any balanced armature IEMs. Used to drive a headphone amp, or powered speakers, the X-Fi makes a decent DAC—especially if a bit of noise under some conditions doesn’t bother you. Just don’t be fooled into thinking it supports 24 bit operation.

---

TECH SECTION

---

TECH INFO: I haven’t researched it, but I’m guessing Creative uses their own proprietary DAC chip in the X-Fi Go. Windows 7 thinking it’s capable of 24 bit operation is one clue it’s probably not a typical off-the-shelf part. If someone knows more, please leave a comment?

FREQUENCY RESPONSE: The frequency response with a 10K load (such as a headphone amp) was excellent being down only about 0.4 dB at 20 Khz with none of the filter-related ripple seen in the C-Media DACs. With a 33 ohm load, typical of portable headphones, it’s not so good. The X-Fi is –3 dB at 30 hz which is likely audible. With 16 ohm headphones it would be even worse. This indicates a capacitor coupled output with a cap that’s too small and is unfortunate. But used with a decent headphone amp (or to feed a powered audio system or speakers) the X-Fi is looking great so far:

THD+N vs OUTPUT 1 Khz 16 & 24 BIT MODES: This test starts at 10 mV (around – 40 dBFS) where noise dominates the measurement. The yellow line is with a 10K load and the X-Fi hits almost 1 volt with distortion at only 0.007%. This is a bit lower output than many USB DACs and well below the Redbook standard of 2 Vrms. In theory, 24 bit operation should lower noise and distortion at low levels. Repeating the test with the X-Fi supposedly running at 24 bits as shown by the red trace you can see it’s identical. So, at least for 1 khz THD+N vs Output level, there’s no benefit to 24 bits with the X-Fi. Into 33 ohms distortion is slightly higher but still impressively low. But maximum output is only about 750 mV due to the output impedance (more on that later). The second graph shows the UCA202 and Micro II for comparison (note the horizontal scales are different):

 

THD+N 100 hz 0 dBFS & OUTPUT IMPEDANCE: With the PC volume at maximum, and a 0 dBFS input, the X-Fi produces about 1 Vrms at very low distortion (shown in yellow). The harmonics are below the magic –80 dB threshold. Even the ultrasonic components are below –80 dB. With a 15 ohm load, the output drops to 0.65 Vrms indicating an output impedance of 7.8 ohms at 100hz. Part of that is the output capacitor. At 1 khz it’s slightly better at 7.1 ohms. This is a marginally high output impedance and is similar to the iPod Touch 3G. Following the “1/8th Rule” the X-Fi should ideally only be used with headphones that are 60 ohms or higher. It’s an especially poor match for Balanced Armature IEMs such as Etymotic, Shure, Ultimate Ears, Westones, etc. The output impedance, and capacitor coupled output, explains why my Ultimate Ears didn’t sound quite right. But, just like the frequency response above, used with a headphone amp the X-Fi will work reasonably well:

THD+N vs FREQUENCY: Here’s the THD+Noise plotted from 20 hz to 20 Khz into 10K (yellow) at 775 mV (0 dBu) and 33 ohms (blue). The 10K plot is impressively flat and entirely below 0.01% which is very good performance for such a small inexpensive DAC. It’s similar to the UCA202 up to 5Khz, and notably better above 5 Khz. Into 33 ohms you can see the effect of the electrolytic output capacitor increasing distortion at low frequencies and the amp apparently doesn’t like to drive low impedances above 13 Khz (it was likely clipping). So, once again, the X-Fi does fairly well if it’s driving an amp, but not so well driving headphones. The second graph shows the UCA202 and Micro II for comparison:

SMPTE IMD 10K: The X-Fi does very well here with everything around –90 dB or lower (please ignore the errors in the caption):

 

SMPTE IMD 33 OHMS: Even into 33 ohms, the X-Fi still does fairly well. The output capacitor is causing some trouble as can be seen by the 2nd harmonic at 120 hz just above –70 dB but that’s relatively benign THD. The IMD is still extremely low:

CCIF IMD 44 Khz 10K: The X-Fi also does well here for an inexpensive DAC into 10K. It’s particularly amazing the side bands to the high frequency signals are below –100 dB! The overall result, due to the higher difference component at 1 Khz, is not as good as the UCA202 but they’re very close. They’re both massively better than any of the C-Media DACs such as the FiiO D5 shown in the third graph for comparison:

 

CCIF IMD 44 Khz 33 Ohms: The X-Fi also does well here driving 33 ohms and very similar to the result above into 10K:

NOISE & LINEARITY: The X-Fi is a bit disappointing for noise at about –89 dBu A-Weighted and 50 uV of raw noise into most any load. This is no better than the $12 Syba DAC. It’s also slightly worse than average for linearity with a 1.5 dB error but that’s not likely to be audible and is still better than the UCA202’s error. I repeated this test operating (supposedly) at 24 bits and got the same result. Consistent with the distortion sweep, it’s apparent the supposed “24 bit” mode is just just creative marketing—pardon the pun. The X-Fi, in either mode, falls well short of of the theoretical 96 dB total dynamic range for 16 bits. The raw dynamic range is 1V/50uV or only 86 dB which yields an Effective Number Of Bits (ENOB) of only 14 bits. This is a bit noisier than even the UCA202 which already isn’t especially quiet (shown in the second graph for comparison). If you plan to use the X-Fi by itself, or use a software volume control with an external amp, you will hear some hiss under some circumstances:

 

JITTER: Here’s the spectrum from the dScope’s J-Test for jitter. The two things to look for are the number and level of symmetrical sidebands and the “spread” at the base of the 11025 hz signal. The second graph is UCA202’s jitter for comparison. The X-Fi does a significantly better job all the way around. The sidebands are lower in level and the spread is significantly less. This is excellent jitter performance for a cheap USB DAC:

 

24 BIT OPERATION: The X-Fi reports to Windows supporting 24 bit operation but no matter what I tried I couldn’t get anything but around 14-15 bit performance from it. Perhaps the driver is sending 24 bits over USB but the X-Fi Go might be simply discarding the 8 least significant bits or otherwise is unable to do anything with them. That’s a disappointing result and some might argue deceptive. If the X-Fi really is receiving 24 bits all the way to the DAC it might be the DAC and analog performance are poor enough they’re masking any benefit. But even if that’s the case, I would still expect to see lower THD at low input levels from reduced quantization distortion and I did not.

RESAMPLING: Creative has a reputation for re-sampling 44 Khz to something else—either on the PC side or in their hardware. I didn’t see any signs of that with the X-Fi Go. It only supports 44 Khz and if that wasn’t its native sampling frequency you would expect whatever that is (i.e. 48 Khz) would be reported to Windows. In addition you can see a 44 khz component in the spectrum graphs indicating that’s what the DAC is really running at.

TECH COMMENTS: Used to drive an amp like the O2, powered speakers, or other gear with a line input, the X-Fi Go! is a decent DAC with the exception of mediocre signal-to-noise numbers. If you leave the software volume at maximum and control the volume from the amp/speakers/etc. it should have enough dynamic range for most applications. I put it slightly ahead of the Behringer UCA202 in most regards. The maximum output is a bit limited but most amps should still have enough gain. I would not recommend it to drive headphones under 60 ohms and especially not balanced armature IEMs. It also does not have enough output for a lot of higher impedance headphones including the popular Sennheiser HD600/650.