Objective Reviews & Commentary - An Engineer's Perspective

February 28, 2011

Behringer UCA202 Review

behringer uca202 testINTRO: The little Behringer UCA202 has a rather loyal following and I’m not aware of anything that comes close for its bargain $29 price. I also was hoping it might make at least a decent USB headphone DAC. I love finding bargains that perform far better than their price would suggest!

BEHRINGER TRIVIA: Behringer is an interesting company headquartered in Germany. People tend to be rather polarized in their love, or dislike, of Behringer. Most of the criticism seems related to Behringer charging much less for similar products than some of their competitors. They’ve been accused of copying more expensive designs, some say their gear is “junk”, and some are very happy with their Behringer products.

For better or worse, Behringer has changed an entire industry. Before they came along, companies like Mackie had plenty of customers willing to pay premium prices for their gear. But then Behringer changed the status quo with similar products for a small fraction of the price. Mackie, for example was was forced to move their manufacturing to Asia, update their products, and radically drop their prices to stay in the game. The same has been true of other competitors. Today, not surprisingly, Behringer has a lot more reasonably priced competition.

behringer a500AN INTERESTING BEHRINGER STORY: Many years ago Behringer came out with what looked like a clone of a popular power amplifier used in recording studios—The Alesis RA300. And, at the time, the street price ($200) of the Behringer A500 was half the price of the $400 RA300. And similar amps by highly regarded companies like QSC, and Crown cost even more. Looking at the pictures here of both it’s easy to understand the clone accusations. So the Behringer A500 amp didn’t get much respect and was frequently attacked in the forums as sounding awful, being a cheap copy, and even a “joke”. alesis ra300Interestingly, however, the actual amplifier circuitry of the Behringer was very different than the Alesis so the clone accusations beyond the front panel were myth rather than fact.

behringer-blind1BEHRINGER VINDICATED: Here’s this supposedly awful sounding, made in China, bargain basement, pro amp considered “junk” by many. It’s about the last thing a true card-carrying audiophile would ever consider for their own system. So why not have a sense of humor and throw it under a bed sheet and use it in a $12,000+ system in a blind comparison against a seriously high-end audiophile approved wonder amp? And while you’re at it, you might as well drive it with a cheapo bottom-of-the-line Sony CD player—cheap DAC and all--and stack both on a rickety old chair instead of the usual $1000 sand filled audiophile equipment stand with floor spikes. And, just for good measure, use a $4 extra long tin-plated thin cable between the CD player and Behringer. Then have some volunteers stand behind the speakers and swap the cable plugs back and forth at the request of the card-carrying audiophiles doing the listening.

HOW DID BEHRINGER DO? As many of you may have probably guessed, the golden-eared audiophiles listening to the high end speakers and both sets of electronics couldn’t tell the Behringer/Sony combo apart from the megabuck gear. There’s a lot to be be learned from blind tests, but at the least, Behringer likely got bit more respect after this one. Here’s a link to the full story: Matrix HiFi's Behringer vs High End Gear

behringer-blind2WHY IT MATTERS: Behringer has shown across their product line that you don’t have to spend lots of money to get great performance. Tests of Behringer, and similar gear  from other manufactures, have also proven supposed high end companies are often selling mostly hype rather than genuinely better sound or performance. In a few areas (such as speakers) spending more gets you better sound. But in most product areas, once you reach a certain point of diminishing returns, spending more just gets a you a higher-end brand name, sometimes nicer looks, and a lot less money to spend on other things.

And if you don’t like using a pro-sound company for this analogy, there are plenty of home audio ones as well—such as Emotiva. Their gear can stand up to just about anything comparable at any price in terms of both measurements and blind listening tests. You really can have your cake and eat it too sometimes.

HOW CAN BEHRINGER DO IT? Behringer’s large size and high volume gives them a big advantage. The DAC chip in the UCA202 sells for about $6 just by itself—and that’s if you buy 2000 at a time. That’s what some little boutique audiophile USB DAC manufacture would have to pay. But Behringer likely buys a zillion of them and pays much less. They have in-house specialized design talent shared across their diverse products for everything from circuit board layout to injection molded enclosures. And they have strong manufacturing capability in China. So their total costs are a fraction of what a specialty audiophile manufacture would pay to make the exact same thing.

IS THE UCA202 JUNK? I’m sure some of the more esoteric forums that discuss DACs (with some costing thousands of dollars) would largely dismiss the UCA202 as “junk”--just like those who dismissed the Behringer A500 amp above. It might be cheaper than two 3D movie tickets, but is it really junk? It’s USB powered, features an optical output, RCA Line inputs/outputs, a 3.5mm headphone jack and even a headphone volume control.

WHAT ABOUT THE UCA222? (revised 2/28/11) Behringer has a newer version of the UCA202 called the UCA222. I don’t have one, but as far as I can tell, it’s the 202 with some DSP effects likely added to a proprietary driver. This could be bad news if it’s no longer true “plug-n-play” like the UCA202 is. And the proprietary driver may not be “bit accurate” (totally transparent) even with all the DSP functions disabled. On the other hand, if you want to mess around with sound processing and effects, then it might be a cheap way to experiment before spending big bucks on a more capable effects processor. And if it will work without the Behringer driver it may well perform similar to the UCA202.

THE UCA202 AND WINDOWS: I’ve plugged my UCA202 into several PC’s of various vintages and configurations. They include an ancient laptop running XP, my Core 2 Duo lab system, a Vista machine, and a 64 bit Win 7 Core i7 system. It always auto-installed without so much as a single mouse click. No buggy drivers required. It shows up in Windows as a generic “USB Audio CODEC”. If you run Linux or OS X, your results may vary.

SUBJECTIVE LISTENING: First I hooked it into my “big rig” via the line outputs and played some favorite tracks using Foobar 2K and listened over my reference speakers and assorted headphones on my Benchmark DAC1 Pre. The UCA202 sounded good. In fact, the line outputs sounded rather similar to the DAC1 Pre doing the D/A work. The UCA202 was quiet and never had any obvious distortion or sonic flaws. Moving to the headphone output, things were not as simple. With 16 ohm headphones it sounded noticeably different than the Benchmark and it wouldn’t get very loud. But with higher impedance efficient headphones it still sounded decent. With my uber-efficient UE SuperFi 5 Pro’s there was only readily audible hiss at the top 20% or so of the volume control’s range but the frequency response had obvious problems due to impedance issues. With more typical headphones the hiss was inaudible, or barely audible, at any setting.

MEASUREMENT SUMMARY: In most ways, especially via the line outputs, the UCA202 performs respectably well. The headphone output, however, has a 50 ohm output impedance. Behringer likely intended using professional large headphones of the sort often used for studio and monitoring work (they are, after all, a pro-sound company). Most of these headphones are 80 – 600 ohms and would work reasonably well with the Behringer depending on how loud you want to listen and their sensitivity. But with typical 16 to 32 ohm portable headphones, it’s a different story. Here the UCA202 struggles. And with most balanced armature style IEMs, the frequency response will no longer be even close to flat due to the impedance interaction (see: Headphone & Amp Impedance).

BOTTOM LINE: If you your PC has marginal line outputs to feed an outboard sound system, amp, use for measurements, etc., the UCA202 is a bargain and offers significantly better performance than most built-in sound hardware. It also makes a good second interface (Windows 7 lets you assign different software/sounds to different interfaces). It’s especially useful for laptops that often have inferior sound hardware and often no line inputs or outputs at all. So if you’re not looking for a high-end solution, the UCA202 might fill the need and save you some serious money. If you want a headphone amp, and happen to have some 80 – 600 ohm fairly efficient headphones, the UCA202 is also worth considering. The headphone amp mostly measures very well, offers excellent channel balance, and has reasonably low noise. The only real downsides are the output impedance and power. If these are not a problem for your application, it might be an ideal solution. You could also use the line outputs to drive a dedicated analog headphone amp.

TECHNICAL STUFF (non-geeks might want to skip this section):

uca202 top pcbUCA202 INTERNALS: The UCA202 uses a highly integrated Burr Brown/TI PCM29xx CODEC chip. As mentioned earlier, they’re nearly $6 each in 2000 piece quantities. With most everything integrated into a single chip, it makes the implementation harder to screw up. The chip has a delta-sigma DAC, oversampling digital filter, and is rated for 93 dB of dynamic range and 0.005% THD+N. The PCM2902 Datasheet has more info.  A classic 4558 op-amp is the headphone amp and another pair of op-amps handle line in/out duties. The power supply looks to be reasonably well filtered.

The pictures show it’s not a made-in-California HRT Music Streamer DAC, but it’s a clean layout and you can buy 5 of these for the $150 price of HRT’s least expensive model. Nobody will think the plastic case is “high-end” but as long as it doesn’t fall apart, what matters most is how it performs and how it sounds. And here, so far, things look OK.

uca202-pcb-bottomLINE OUT THD: Starting with the line outputs, here’s the THD and THD+N into a 100K load on the dScope with a 1 Khz at 0 dBFS 44.1 Khz digital signal via USB. The distortion measurements include all frequencies up to 22 Khz. This is the maximum output possible. The line outputs are fixed so the volume setting doesn’t matter (click for full size):

Behringer UCA202 1 Khz 0 dBFS 44.1 Khz USB Line Out 100K

A 0.008% THD measurement is very respectable while playing back the “loudest” digital signal possible (0 dBFS) and pushing the line output amps to their limit. The noise profile is interesting. Starting at around 25 Khz (well above what’s audible) the noise climbs steadily up to about –80 dB at 70 Khz. This is something I’d never seen in RMAA measurements of the UCA202 because they rarely show anything past 20 Khz.

NOISE SHAPING: It may look bad, but that big ultrasonic bump is likely what’s known as Noise Shaping. Designers intentionally push the noise from the D/A conversion above the audible band to improve the audible signal-to-noise ratio and that’s likely what’s happened here. It’s worth noting many devices have much louder distortion products in the audible range and this is well above the audible range. I think it’s safe to say the ultrasonic bump is inaudible and better than more noise and distortion below 20 Khz where you can hear things. Behringer could have filtered more of this out, but the filter itself might have had a negative effect on performance in the audible range so this is likely a good trade-off.

LINE OUT & USB POWER NOISE: Here’s the noise analysis under the same conditions as above but with a barely visible –115 dBFS signal to give the DAC something to do (many DACs mute and shut off with no signal giving an unrealistic noise measurement and making the specs look better—but no cheating allowed here):

Behringer UCA202 1 Khz -115 dBFS 44.1 Khz USB Line Out 100K Noise

The un-weighted noise up to 22 Khz is about –88 dB from the reference level in the previous screenshot. And the A-weighted noise (the “dBA” spec you often see published) accounts for human hearing sensitivity and is about –92 dB. These are not amazing numbers but certainly respectable—especially for a device running from famously noisy USB power. And as indicated in the subjective comments earlier, the resulting slight hiss can be audible under some conditions but typically isn’t a problem. OK, so the line outs look good, but what about the headphone outputs?

HEADPHONE THD:  Here’s the UCA202 via the headphone output into a typical load:

Behringer UCA202 1 Khz 0 dBFS 44.1 Khz USB 150 Ohms Reference Level (~400 mV) THD and THD N

Amazingly enough, the THD+N here is actually lower than the Line Outs! That’s likely because the 1.1 volt RMS (3.1 v p-p) line output is stressing the line-out amplifier a bit while the 0.4 volt signal here is a piece of cake on a 5 volt power supply. Also note the load is 150 ohms, not the more typical 15 ohms I have used testing portable players. More on that in a bit. The volume control on the UCA202 was adjusted here to deliver a reasonable reference level for 150 ohms. With the typical studio headphones Behringer intended, 400 mV RMS of output would be a decent listening level. The noise floor in the audible band is impressively clean.

HEADPHONE & USB POWER NOISE:  The headphone noise is pretty much identical to the Line Out noise above:

Behringer UCA202 1 Khz -115 dBFS 44.1 Khz USB Headphone 15 Ohm Noise (ref ~ 180 mV)

MAXIMUM OUTPUT: Here’s the max headphone output into the same 150 ohm load:

Behringer UCA202 11025 hz 0 dBFS 44.1 Khz USB 150 Ohms Bal Max Output Clip (ref ~400 mV)

The UCA202 is on the edge of clipping (just under 1% THD) at about 660 mV output into 150 ohms. This works out to about 3 mW into 150 ohms which doesn’t sound like much, but the Sansa Clip+ can only manage about 430 mV into the same load which is only 1.2 mW. So it very much depends on the impedance “match”. Here’s what happens to the UCA202 with the same 15 Ohm load the Clip+ was tested with:

Behringer UCA202 1 Khz 0 dBFS 44.1 Khz USB Headphone 15 Ohm Clipping Point (ref ~ 180 mV)

It’s actually a bit over 1% THD+N here but it’s close enough. I used this level because it’s the same 1 mW/32 ohm reference voltage I’ve used for testing other devices. Into 16 ohm headphones, the UCA202 will only manage about 179 mV or 2 mW. The Clip+ is much better at about 11.5 mW. This is a significant difference and it’s a big reason why the UCA202 isn’t an ideal headphone amp for typical low impedance portable headphones. As I reported in the subjective tests, the UCA202 didn’t get loud enough with some of my headphones. I also reported it sounded odd with other types.  This is related to the output impedance of the UCA202.

OUTPUT IMPEDANCE: Here’s the No Load output referenced to the same 400 mV at 150 ohms level used earlier:

Behringer UCA202 1 Khz 0 dBFS 44.1 Khz USB No Load (ref ~400 mV) THD and THD N

The no load voltage is 530 mV and with 150 ohms it was 403 mV. Do the math and you get an output impedance of 47 ohms—close to the Behringer spec of 50 ohms. The good news is this will work OK with many full size efficient 80 - 600 ohm studio-type cans. But, as already indicated, the UCA202 isn’t a good match for low impedance loads. That’s why most of the tests here are into 150 ohms.

THD vs FREQUENCY: While some devices have low THD at 1 Khz, the distortion nearly always rises at high frequencies and sometimes low frequencies as well. Here’s how the UCA202 fares driving a load with the measurement bandwidth set to 22 Khz:

Behringer UCA202 Swept THD N -3 dBFS 44.1 Khz USB 150 Ohms (ref ~400 mV)

The UCA202 does impressively well staying well around 0.015% up to nearly 10 Khz when the THD+N starts rising but it’s still at a very respectable 0.1% even at 20 Khz. Note the distortion is calculated here “on the fly” with the frequency sweep versus the spectrum 1 Khz plots above that are done with a more detailed (and slower) FFT result--that’s why the 1 Khz number here is slightly higher. And the “stair steps” above 10 Khz are due to the way the 22 Khz cut-off frequency interacts with the distortion calculation.

IMD DISTORTION: Here’s the excellent IMD result, run at –2 dBFS which is just under the clipping level of the combined signals:

Behringer UCA202 -2 dBFS 44.1 Khz USB Dual Tone SMPTE IMD Spectrum

IMD of 0.0009% is hard for anyone to argue with but it’s often interesting to get a closer look around the 7 Khz signal. The 60 hz tone creates “sidebands” here at multiples of 60 hz. Here’s the closeup view of the above spectrum:

Behringer UCA202 -2 dBFS 44.1 Khz USB Dual Tone SMPTE IMD Spectrum Closeup 7 Khz

The sidebands are limited to just a couple spikes at –100 dB or lower. Overall this is very respectable performance for any DAC let alone the loaded headphone output of one costing $29.

FREQUENCY RESPONSE: Built-in sound hardware and low-end PC sound interfaces often suffer from audible frequency response problems. Some use coupling capacitors that roll off low frequencies, and some have rather severe high frequency performance from poor D/A output filtering or other problems. Here’s how the UCA202 does with both channels shown in 2 colors and divisions of just 0.5 dB into 150 ohms:

Behringer UCA202 Swept Freq -3 dBFS 44.1 Khz USB 150 Ohms (ref ~400 mV) Ch A (yellow) Ch B (blue)

Note it’s dead flat all the way down to 10 hz, the channels are very closely matched, and the response is about –0.25 dB even at 20 Khz. Again, this is very good performance but in this case it’s helped out by the relatively easy 150 ohm load. There is some low frequency roll off into lower impedances.

DAC LINEARITY: As mentioned in the Sansa Clip+ Review, a key difference between cheap DACs (or a bad implementations of good ones) is often their low level linearity. So it’s a good test to feed in a low signal near the noise floor and see how close the DAC comes to reproducing it at the correct level. Burr Brown and the UCA202 here, impressively, nail it:

Behringer UCA202 1 Khz -90 dBFS 44.1 Khz USB Line Out 100K

JITTER SPECTRUM & PITCH ACCURACY: Another measurable difference between DACs, and especially USB DACs, is their jitter performance. There’s a lot of debate about this topic, but I can personally attest to having heard plainly audible jitter in a listening test. So don’t let anyone tell you it’s always inaudible. You can read more about it, and how it’s typically measured, in my Jitter Post. The USB interface and clock design can really make a significant difference. Here’s the UCA202 on the same standard jitter test I use for all digital devices:

Behringer UCA202 11025 hz -6 dBFS 44.1 Khz 150 Ohms Jitter Spectrum (ref ~400mV) with comments

This is a fairly impressive result—especially via USB. There are only two obvious sidebands, and they’re at about –120 dB which is quite low. There is some mild-to-moderate “spread” (likely low frequency jitter) but I’ve seen much worse. Overall, this is a good result for any USB dac and an amazing result for a $29 one. For comparison, here is the Benchmark DAC1 Pre on the exact same test playing the same file via USB:

Benchmark DAC1 Pre 11025 hz -6 dBFS Jitter Spectrum 15 Ohms (Ref 400 mV)

With the Benchmark there is less spread, virtually no sidebands, and the DAC1, as expected, has a significantly lower noise floor as well. I’ve also included a pitch measurement as part of this test. The dScope has a very accurate internal timebase so the 11025 hz input signal should (and nearly always will be) reproduced at 11025 hz plus or minus a small fraction of a hertz. The Clip+ was a notable exception with a 0.25% pitch error (that’s corrected by the Rockbox firmware). The UCA202 is fine.

SQUARE WAVE TEST: As revealed in the Clip+ Review, square waves can be very telling about several different key parameters. Here’s the UCA202:

Behringer UCA202 1 Khz 500 mV p-p Square Wave 150 Ohms

There’s nothing much to complain about here. This is a fairly typical result that compares well to even much higher-end products.

CHANNEL SEPARATION: Small devices tend to, out of necessity, physically locate the right and left channel circuitry close together. And the two channels often share IC chips to save more space (and money). Both of these tend to reduce channel separation. So I wasn’t expecting the UCA202 to do well here into a typical headphone load:

Behringer UCA202 Channel Separation vs Frequency 150 Ohms (ref ~400 mV)

The separation above 100 hz (below that frequencies are largely non-directional anyway) ranges from a worst case –50 dB up to nearly –75 dB around 3 Khz. This is far better performance than I expected from such a small inexpensive device and more than enough for most any application. But, like the frequency response, it’s helped out by the relatively easy 150 ohm load.

VOLUME CONTROL AND CHANNEL BALANCE: Analog volume controls, like the cheap little potentiometer in the UCA202, have isolated sections for each stereo audio channel. And it’s difficult, and expensive, to make these two sections match perfectly. So as you adjust the control, the actual settings for the right and left channels can be different from each other. How much tracking error you get depends not only on the control itself, but also its relative resistance to the rest of the circuit. A good circuit design helps minimize the errors.

Generally errors of less than 1 dB are considered acceptable while higher errors are usually audible. Due to the way volume controls work, any error in the control itself is usually most obvious at the lowest volume settings. Some devices use various kinds of digital or electronic stepped analog volume controls to avoid this problem. With a cheap looking “pot” I expected the UCA202 to do poorly here. So I made several measurements at various volume settings lower than the ~400 mV reference level:

Behringer UCA202 11025 hz 0 dBFS 44.1 Khz USB 150 Ohms Volume -4 dB (ref ~400 mV)

At the reference level (scroll back up to the start of the measurement section) the channel balance was within a very impressive 0.12 dB. Turning the volume down to –4 dB increases the error slightly to 0.45 dB as shown above. This is is still comfortably below the threshold of audibility. The “Balance” measurement above is simply the two left-most measurements subtracted from each other. Here’s the result at much lower –20 dB:

Behringer UCA202 11025 hz 0 dBFS 44.1 Khz USB 150 Ohms Volume -20 dB (ref ~400 mV)

Even at a fairly soft –20 dB, the channel balance is still under 1 dB (but just barely). How about even lower:

Behringer UCA202 11025 hz 0 dBFS 44.1 Khz USB 150 Ohms Volume -30 dB (ref ~400 mV)

A level of  -30 dB is getting really quiet. And here the tracking is still under 1 dB. How about extremely low levels:

Behringer UCA202 11025 hz 0 dBFS 44.1 Khz USB 150 Ohms Volume -45 dB (ref ~400 mV)

If you want to know how soft a 45 dB drop is, try adjusting the level on a piece of audio gear that has a calibrated volume control to a comfortable level and then lower it 45 dB from there. If you have a portable player with Rockbox, you can do this with Settings > Sound > Volume. The above is likely lower than most people would listen at but it depends on the efficiency of your headphones. If your headphones are 10 dB more efficient than average, –45 dB becomes –35 dB. Here the UCA202 has a slightly audible error of 2 dB but this is still an impressive result for what has to be a really inexpensive volume control in a $29 USB DAC.

MEASUREMENT SUMMARY: As stated in the Bottom Line section earlier, the Behringer UCA202 line outputs measure very well even for a more expensive DAC. So if you need line outputs, or want to use them to drive a headphone amp, the UCA202 might be a great deal.

The headphone output, on the other hand, works best with fairly efficient high impedance professional or studio-type headphones (with an impedance in the 80 – 600 ohm range). For typical 16 – 32 ohm headphones you probably won’t be that happy unless they’re really efficient or you like to listen at low levels. There may also be some significant frequency response variations due to the impedance—especially with balanced armature IEMs. Still, the UCA202 is so cheap, it doesn’t cost much to find out if it will meet your needs?

UCA202 MODIFICATIONS: (revised 3/15/11) The only thing keeping the UCA202 from being a respectable headphone DAC is the ~50 ohm output impedance and limited level into low impedances. By replacing a few parts the UCA202 can be upgraded to a much higher quality headphone output for less than $5 worth of parts. See how that experiment turned out in the Modified DAC article.

February 24, 2011

Sansa Fuze (mini review)

sansa fuze packageINTRO: I liked the Sansa Clip+ so much I was curious about its bigger brother: The Fuze. The word on the web is the Fuze has similar performance to the Clip+ and it has several advantages.  And the newer Fuze+ is getting some negative reviews, so I wanted to grab an original Fuze before they're all gone. I purchased a black 4 GB version for $50.

FUZE ADVANTAGES vs CLIP+: Some reasons why someone might choose a Fuze over the Clip+:

  • The screen is bigger, easier to read, in color, and can show more information like more of the artist/song information, album art, etc.  The bigger screen is especially an advantage with Rockbox as it can display all sorts of cool things.
  • The controls are larger and easier to use. The buttons are more spread out on the Fuze and it has a rubberized rotating "wheel" which makes navigating the menus, adjusting the volume, etc. quicker.
  • The Fuze has a larger battery and will play music longer than the Clip+
  • You can watch videos in several formats including MP4, MOV, AVI, DiVX, WMV
  • You can take your pictures with you and display them on the Fuze.
  • The Fuze looks and feels more substantial. The back is “rubberized” which makes it feel better in your hand. It’s much thinner than the Clip+ which makes it seem more sleek and expensive.
  • I’m not sure many take advantage of it, but the “dock-style” connector at the base of the Clip+ allows it to be used in Sansa-compatible devices like portable speakers, etc. There are line outputs available in the connector.


  • It’s bigger, heavier, and especially wider. It’s more noticeable in a pocket.
  • You can’t clip it directly to your shirt sleeve, pocket, etc.
  • You may need to buy an armband if you want to use it at the gym or jogging
  • Some find the wheel imprecise and it also makes accidentally changing the volume easier
  • It uses a proprietary Sansa “dock” connector (the same as the e200 series) vs a standard mini USB jack on the Clip+. This means you need a special cable for charging and syncing the Fuze but not the Clip.
  • It’s been discontinued and will eventually be hard to find at stores.

sansa line upIMPRESSIONS: It’s a nice player—certainly one that seems like it should cost more than $50. It’s sleeker than the much thicker and harder to hold Clip+. There are some extra screen animations and graphics, plus the wheel, but the user interface, menus, etc. are mostly very similar to the ones on the Clip+. The Clip+ has volume up/down buttons on the side and the Fuze uses the wheel instead. Both players also run Rockbox firmware in similar ways but the bigger screen on the Fuze is a huge advantage here. Listening to the Fuze it sounds the same, to my ears, as the Clip+.

MEASUREMENTS: The Fuze measures nearly identically to the Clip+ right down to the 0.25% pitch error with the Sansa firmware. I checked all the same things I measured with the Clip+ and they’re all essentially the same. It’s very likely Sandisk is using the same audio design in both. It has the same ruler flat frequency response, low output impedance and respectable distortion levels. The numbers and graphs are so similar it’s a waste of time to publish them here—just see the Clip+ Measurements. Here’s a quick dScope capture showing several Fuze measurements (click for full size):

Sansa Fuze 0 dBFS 1 Khz ~180 mV 15 Ohms Multiple Measurements

SANDISK GOING BACKWARDS: I haven’t personally used or tested one, but judging by the reviews on Amazon, the forums, and elsewhere, the “new and improved” Sansa Fuze+ is mostly a step backwards. They updated the shape, look and controls to try and chase Apple, but if what I’ve read is true, it’s harder to use, slower, and doesn’t have as impressive of audio performance. Several people have said they got rid of their Fuze+ in favor of the older original. I also fear this may happen when they replace or update the Clip+ as well but perhaps they’ll learn from their mistakes with the Fuze+?

sansa fuse testBOTTOM LINE: If some of the advantages of the Fuze over the Clip+  are appealing, you might want to consider getting one soon before they’re sold out. I wouldn’t be surprised, if like the previous 5th generation iPod Nano, the Fuze ends up selling for more on eBay than they sold for new in the stores. Apple, in their quest for ever smaller/thinner/sleeker body jewelry, once again chose form over function and removed the video camera from the 6th generation Nano. That’s created a strong secondary market for the previous generation with factory sealed Nano 5G’s bringing higher prices than they did new. If you already have a Fuze and love it, you might want to consider buying a spare.

MISC DETAILS (edit 2/24/11): My Fuze is running V02.03.33A firmware. And I’ve been informed by one of the Rockbox developers on Anything But iPod the Fuze V2 (i.e. like the one I tested) uses different hardware than the Fuze V1. So your mileage may vary if you have the earlier version.

February 23, 2011

Cowon i9 Review

cowon i9 and ipod
INTRO: I wanted a new player with great sound I could comfortably wear on my arm at the gym. Cowon players have a strong reputation for better sound than most players. The Cowon iAudio 9 (aka "i9") is similar in size to the previous generation iPod Nano and promised impressive audio performance. So I bought one and figured some objective testing was in order given all the praise for these players.
WHO’s COWON? Unlike say Apple, Creative, Sony, or even Sandisk/Sansa, a lot of people have never heard of Cowon. They’ve been making players for many years and seem to have a fairly small but loyal following in the USA. Their headquarters are in Korea and they’re represented by JetAudio in the USA. I couldn’t find an address or even phone number for JetAudio. I could only find a contact form on their website which isn’t very encouraging if you need support.
THE iAudio 9: The $120 - $180 Cowon i9 has been around since 2009 and is still a current Cowon model. It can be seen in the photo to the right of my iPod Touch 3G with the protective plastic still stuck to it and some test files displayed. It's impressively thin (far slimmer than the Sansa e200 series), and the display is bright and clear. There’s an unusual touch pad on the top and 4 controls on the sides, including volume up/down, but unfortunately they’re located right where you're most tempted to wrap your palm around the i9 when holding it making the buttons more awkward to use.

THE USER INTERFACE: I'm not usually too concerned with the user interface on a device as long as it's half way intuitive. The Cowon, however, oddly combines a touchscreen-like interface with a regular screen and separate touch pad. Anyone with a touch phone, iPod Touch, or similar device, may find the i9 rather odd. Navigating menus and settings is partly done with a diagonal swipe between the lower left corner and upper right corner of the pad. Some swipes "wrap around" and some stop when you “hit the end”. And there are no labeled "soft keys" to speak of—just forward and back arrows and a few dots that are backlit on the touch pad (see the pic below). Basically there’s the diagonal slide bar, 2 soft touch “pads”, 3 real buttons on the sides, and an on/off/hold switch.  Many functions you have to memorize and, the 6 controls behave differently depending on what screen you’re on or where you are in the menu structure. Plus, you have to remember what a “short press” or "long press" do for each of the controls in those various situations. Got all that? Here’s a link to the manual at JetAudio: i9 PDF Manualcowon i9 closeup
FORM OVER FUNCTION: The user interface follows a popular theme in many Asian-designed products where "cute", "flashy" or “different” are often favored over ease-of-use. Of course Apple products are designed in California and they're often guilty of putting form-over-function as well. But Apple does such a good job with certain things, many don't seem to mind when they hold their iPhone 4 wrong it drops their calls. But, regardless, design compromises in the interest of fashion, or just to be different, can be frustrating and that includes the i9 user interface. The Sansa Clip+, for example, makes do with a standard 4-way “D” pad and 2 other buttons quite nicely. The iPods are even more elegant. Just about any player I’ve encountered is easier for a stranger to just pick up and use compared to the i9.

FLAC SUPPORT: Another reason for buying the i9 was FLAC support. Perhaps 75% of my library is now in FLAC. So it's a big plus to have it natively supported on a portable player. To my knowledge, none of the current iPods support FLAC although the Clip+ (and any Rockbox player) does.

NO iTUNES REQUIRED: Lots of people dislike iTunes--especially on Windows. It's rather invasive and some even consider it "malware" as it can do things like re-tag your music collection with Apple proprietary tags and is often doing things you don't want in the background or trying to sell you something via the iTunes Store. The whole sync process can also be rather slow and more involved than it need be for just throwing a few new songs on your player. Plenty of angry prose has already been written about iTunes so I'll leave it at that. But it's nice the Cowon players don't need it and work with simple "drag and drop" if you want.

PROPRIETARY USB JACK: Apple iPods require a special Apple licensed "dock" connector to charge them because Apple gets a ransom (err… royalty) from everyone who makes anything that mates to an iPod/iPhone using their connector—even a lowly USB cable. And if you already have other devices that use the standard USB mini/micro jacks (like say your non-Apple phone), it's a hassle to have to use another charger/adapter/cable to also charge your Apple device. But at least iPod's are so popular virtually everyone offers various cables and adapters that fit them. But Cowon chose to use a tiny proprietary weird USB jack that's roughly the same size as the widely used micro USB jack. If you lose the cable that comes with it, your player will soon be dead until you special order another custom cable from one of the very few sources even selling them. To me, that's more than a little obnoxious. I’m betting there will be lots of dead Cowon’s out in the wild from people who forgot their cable at home, can’t find it, etc.

LACK OF ACCESSORIES: One downside of Cowon players is because of their smaller market share there are relatively few accessories available—especially for the i9. So if you want a really nice neoprene sports armband that accommodates all the controls, for example, you’re probably out of luck.

OTHER DETAILS: The i9 has several features you won't find on an iPod. With this review, I mainly wanted to evaluate its audio performance. You can find other reviews on the web that go into the features not covered here.

A FEW WORDS ABOUT EQ: It's generally accepted the only fair way to compare two different players is with the EQ off. EQ can be nearly infinitely adjustable, the same settings/options are not usually available on different players, and everyone's taste in EQ is different--all of which make comparisons all but impossible. So the only fair way to compare sound quality is with the EQ turned off so it’s more apples-to-apples instead of bass-boosted to vocal-emphasis. The player that's most accurate with the EQ off should respond even better to cranking in whatever EQ a particular user might enjoy. The idea is to find out how solid the basic foundation is before you start modifying things.

COWON'S EQ: The i9 has a wealth of EQ options including several licensed from BBE. The BBE technology involves specific EQ curves and sometimes crossfeed or other "soundstage" enhancements. More advanced BBE choices may include dynamic processing. Some of the choices do sound good on certain kinds of music. Some, to my ears, are rather overdone or irritating. But, on the whole, there are more useful EQ selections on the Cowon than the iPod (but see Third Party EQ below).

ADJUSTABLE EQ (edit 2/24/11): I also chose the i9 as, I was under the impression, it had parametric EQ. Normal adjustable EQ has fixed frequency "bands" that you can boost or cut. The i9 was advertised as having bands that were adjustable in frequency like a parametric equalizer. In reality this feature is disappointing. Each band has only a few fixed frequency and width choices and the lowest frequency possible isn't all that low. For example, you're stuck with boomy (rather than deep) if you want to dial in some bass boost.
THIRD PARTY EQ CLOSES THE GAP: It's worth mentioning there are some very nice third party EQ options on some players. The popular Rockbox firmware supports dozens of different players (including the inexpensive Clip+ but not the i9) and has, among other things, true parametric EQ that's extremely flexible and can be used in conjunction with the equally flexible crossfeed options to, with some experimentation, recreate most any of the fixed settings on the i9. There's also EQu for the iPod Touch 3G and 4G (and iPhone 4). It's a $3 app with very flexible EQ options and goes a long way towards closing the gap between the iPod’s EQ and the i9’s.

WHAT ABOUT THE SOUND (subjective impressions)? Given the less-than-friendly user interface, weird USB jack, etc. there must be some other reason people pay $120+ for these, right? I’ve read it's the sound quality. Before measuring the i9 I spent some time listening to some of my favorite music with a variety of headphones and I was reasonably pleased with the sound. With most of the headphones I tried, it played loud enough for my tastes and didn’t have any obvious sonic flaws.

COMPARISONS: Without any EQ (the only fair way to directly compare the two), the i9 didn't sound obviously better than my iPod Touch. Getting the iPod's clean sound in a smaller/cheaper package isn't a bad thing but what about the Sansa Clip+? It’s even smaller and cheaper than the i9. Using very high-end headphones, and some of my favorite highly revealing acoustic tracks of music, I wasn’t sure I could hear any obvious differences with the Clip either. That's not to say there are none, but they don't stand out in a casual comparison at normal listening levels.

AUDIBLE HISS: Playing back a really low signal with my most efficient headphones (UE SuperFi's) the Cowon i9 had some audible hiss but it wasn't objectionable. It's about the same as the Clip+. It was, however, more obvious than the Touch 3G's noise level. With more typical headphones the hiss will likely be inaudible.

THE MEASUREMENTS: More detailed results are presented below, but here's the brief summary:

  • Frequency Response: With a typical load, the i9 starts rolling off in the bass at 100 hz until it's about 4.2 dB down at 20 hz. That's a significant roll off and is much worse than the iPod or Clip+ both of which stay ruler flat all the way down to 10 hz. With good headphones and the right music, this is likely audible and would make the Cowon sound a bit weaker in the deep bass. Above 100 hz the response is fine.
  • Distortion: At typical listening levels, THD+N distortion was roughly 0.02% from 20hz to 1 Khz, but rose to about 0.4% at 10 Khz. This is decent but not great performance. It’s fairly similar to the Sansa Clip+ overall with the i9 doing better at 1 Khz and the Clip doing better at 10 Khz. 
  • Maximum Output: The i9 has higher than average power output--about 2 dB higher than what the iPod can manage into the same load. It delivers roughly 24 mW into 16 ohms. This gives the i9 a bit more "headroom" for inefficient headphones or those who like their music really loud.
  • Output Impedance: The Cowon is better than average here. The output impedance was about 1.5 ohms which should keep frequency variations with even some of the more wild headphones to within +/- 1.5 dB or so.
  • DAC Performance: The square response was clean but the DAC linearity was a bit off at low levels and the i9 might have higher than average jitter. So the DAC results are kind of mixed.


NO HEADPHONE AMP REQUIRED: Unlike many portable players, the i9 should drive most reasonably priced portable headphones, and perhaps even some of the more difficult ones, to reasonable levels without audible problems. The relatively high power and low impedance should mean it's compatible with a wider range of headphones than most portable players. In fact, using a headphone amp with the i9 might make things worse instead of better by creating noise and/or volume tracking problems. See: Headphone Amps & DACs 

i9 vs CLIP+: If a person doesn’t need the higher output of the i9 I think the Clip+ is a better player overall. The Clip+ has much better bass response, less high frequency distortion, and otherwise very similar performance for about 1/3  or 1/4 the price. The Clip also runs the excellent Rockbox firmware, is smaller, and doesn’t require buying an armband to wear on your arm (shirtsleeve). If that extra few dB of output power is important to you, or you’re just in love with Cowon’s particular choices of EQ or some other unique i9 feature, than I’d say go for the i9.

i9 vs iPOD TOUCH 3G: The iPod has much lower distortion across the spectrum and also lower noise and arguably better DAC performance. And with the $3 EQu app, it offers impressive EQ. The Touch of course also runs iOS apps, plays games, has bluetooth, WiFi, a much bigger screen, etc. But, the flipside is it’s bigger, heavier, more expensive, generally requires iTunes, won’t play FLAC files, is missing several of the i9’s features, and has a higher  output impedance which makes it a bit more fussy about what headphones you pair it with. With typical dynamic headphones, if audio quality was the main focus, I’d likely pick the iPod over the i9. For higher-end balanced armature headphones, however, the Cowon or Clip+ would be my pick.

BOTTOM LINE: The Cowon i9 delivers fairly respectable audio performance but it's by no means at the top of it's class. Even the smaller and cheaper Sansa Clip+ outdoes it in some significant ways. And I'm not crazy about the user interface or proprietary USB jack.  To me, it doesn’t do enough better than the Clip+ to justify its higher price, higher hassles, and potential performance problems.

TEST RESULTS (for those who are curious about detailed audio measurements, or just like to geek out on the numbers and graphs):

FREQUENCY RESPONSE: In my opinion, this is the i9's weakest measurement. It's generally considered players should manage +/- 1 dB from 20 hz to 20 Khz to be considered "flat enough" and "accurate". I can even live with +/- 1 dB from 30 to 15 Khz. But the Cowon is down 1 dB way up at 50 hz and it's down over 4 dB at 20 hz. That's a significant roll off in the bass and one that's likely audible under some conditions. It's worth noting this graph in yellow below is roughly what you'll get with most 16 ohm headphones. Higher impedance headphones won't create as dramatic of roll off. This is likely because the output of the Cowon's headphone amp is capacitor coupled and Cowon simply didn't use a big enough capacitor for either space or cost reasons. It's a relatively inexcusable design error in my opinion. Here's the graph:

The red trace above is the frequency response with no load. You can see the Cowon behaves much better here and is virtually flat. That's why I believe the roll off is caused by a poorly chosen output capacitor (or the presence of one at all--the iPod and Clip+ apparently have none which is even better). Here, to show how RMAA can mess things up, is the RMAA result with the exact same load and volume setting:

While RMAA confirms about the same low frequency roll off, note how RMAA shows a high frequency roll off when the dScope result above it does not. It starts around 10 Khz and falls off a cliff right before 20 Khz. This is likely a result of RMAA's multi-tone analysis being flawed. But I have verified the i9 does not roll off the way RMAA claims it does in the high frequencies. RMAA is also supposed to set its own relative 0 dB level which is generally done at 1 Khz. But it fails to do so accurately sometimes (like above). All this stuff is "automatic" in RMAA and, unlike with a real audio analyzer, the RMAA user has little control over any of it. It either delivers good results or it doesn't. And without anything to compare RMAA results to, you never know if the results are correct or not!

DISTORTION: The i9's distortion was fairly typical for a decent player at 1 Khz. Here's the THD+N and spectrum:

THD+N of 0.01% is very likely inaudible and the spectrum doesn't show any ugly surprises. Distortion usually rises with frequency and the Cowon rises a bit more than some other players (like the Clip+). Here's the dScope result at 10 Khz at the same typical listening level of ~ 180mV RMS:

THD+N of 0.4% isn't awful but it's way than the Clip+ and much higher than the Touch 3G. Here's the RMAA plot of the i9 versus frequency:

Note RMAA shows 3 times the THD at 1 Khz and slightly less THD at 10 Khz compared to the dScope but it does generally confirm similar results. I’m not sure what’s going on above 10 Khz but it’s likely an RMAA problem, not the i9.
IMD: I didn't run the dScope IMD test, but RMAA shows around 0.08% in the results table, and the spectrum looks pretty good:

And here's the swept IMD result showing somewhat worse, but similar, numbers:

DISTORTION SUMMARY: Overall, the i9 distortion results are only average. The i9 is similar to the much cheaper Sansa Clip+. And it's vastly worse across the board compared to the iPod Touch 3G. See my comments in the distortion section of the Sansa Clip+ review for more background on distortion measurements. It’s likely the i9 doesn’t have any seriously audible distortion problems.

MAXIMUM OUTPUT: Here the i9 does really well and delivered an impressive 624 mV RMS into 15 ohms at about 0.2% THD+N. This is about 24 mW into 16 Ohms which is about 2 dB louder than either the Clip+ or Touch 3G can play into typical 16 ohm headphones. 2 dB isn't a lot but it's noticeable. Here's the i9 at it's maximum output below 1% THD+N:

DAC LINEARITY: See the Sansa Clip+ review for some background on why this matters, but it's one common measure of the quality and implementation of the DAC in a portable player. The i9 didn't do as well as the Clip+ or iPod here. The -90 dBFS test signal measured -95 dBFS on the i9. This may cause some dynamic audible distortions when listening at low levels but they're not likely to be very objectionable.

NOISE: The noise performance of the i9 is very similar to the Clip+ both on the dScope and RMAA. You can see in the graph above the noise floor is right at the same -120 dB as in the Clip+ review. The iPod does a few dB better than both of them.
JITTER SPECTRUM: See my article on Jitter and the Sansa Clip+ review for background on how and why I make Jitter measurements. Here the i9 is a mixed bag. It shows lower low frequency jitter ("spread" at the base of the signal below) than either the iPod or the Clip+. But it shows some fairly significant sidebands that are likely jitter related. While it's hard to know which one might be more audible (if either), the one thing I can safely say is it has very different jitter performance than either of the other 2 players (which are very similar to each other). Here's the i9 plot:
Compare the above to the iPod and Clip+:
SQUARE WAVE PERFORMANCE: Here the i9 does really well and better than even the next best Clip+. This is a sign of a decent DAC, well designed digital filter, and a stable headphone amp circuit. If you're curious to know more, see the Sansa Clip+ review for some background on Square Wave testing. Note this test is at a higher level than in the Clip/iPod review but, in this case, it makes very little difference. I also didn't test it into headphones, just the 15 ohm load. Here's the excellent result:
CHANNEL BALANCE & SEPARATION: The channel balance was within 0.3 dB which is generally considered "good enough" but ever so slightly worse than the Clip+ and iPod. The channel separation (stereo crosstalk) was about 2dB better than the Clip+ at about 52 dB but it was 8 dB worse than the iPod's 60 dB. Both tests are with a 15 ohm load at ~180 mV for all players.
RMAA RESULTS: Here are the RMAA results of the i9 vs the Clip+ They're mostly similar except for the i9 having slightly lower distortion here (but it has much higher distortion at high frequencies as shown above). RMAA isn't doing a very good job here of representing the true low frequency response. Neither frequency response result in the table even matches RMAA's own graphs from the same test run or the dScope results but the rest is more consistent:
TEST SUMMARY: The i9 delivers mostly respectable performance and even excels in a few areas like maximum output and the square wave performance. But it has weak low frequency response into a typical load and the DAC linearity isn't great. Overall is it worth several times the price of the Clip+? I personally don't think it is. But, for some, it might still be a good choice.

TEST DETAILS: The i9 was running the latest V1.14 firmware from May 2010.

Jitter Does it Matter?

Intel IDT 92HD73E G45 Motherboard JitterINTRO: Some have asked for more details on jitter. It's a controversial topic and there are lots of myths associated with it. Here’s what I know.

TYPICAL JITTER MYTH: I read in a well regarded high-end audiophile magazine that portable players have less jitter because they use flash memory compared to a PC using a rotating magnetic hard drive. The explanation given: Because a hard drive is spinning, it must have jitter, while solid state flash memory does not. Like so many things in high-end audiophile magazines, this has absolutely no basis in fact. Jitter doesn't come from how the digital audio file is stored. It comes from how it was created, how it is sent in real-time (files are not real-time) over interfaces, and how it is re-constructed.

THE FLAC JITTER MYTH: Some argue that FLAC sounds worse than a WAVE file because the compression/de-compression process and/or CPU loading introduces jitter. This is completely false. The DAC receives the identical bit stream for either file type.

THE AES3 VS COAX VS TOSLINK MYTH: While it’s true jitter can be induced by interconnecting digital equipment, all three of the most popular standards AES3 (also called AES/EBU), S/PDIF coaxial, and S/PDIF TOSLINK optical, use essentially the same bitstream format and all combine the clock with the data itself. One format isn’t necessarily better than the other two as it very much depends on other variables. 3 feet of optical TOSLINK might easily outperform 6 feet of coax or 12 feet of AES3. It depends on the cable quality, transmitting hardware, receiving hardware, potential sources of interference, and more. The 3 standards are all prone to similar kinds of problems with AES3 being a bit more robust. Contrary to what some believe, the extra pin in AES3 is not a dedicated clock.

A REAL LISTENING TEST: I participated in a listening test at an Audio Engineering Society conference where they played test tracks with varying known levels of different kinds of jitter. It wasn't a rigid ABX or double blind test, but it was very enlightening. We, the listeners, didn't know which tracks were which until they were explained later. The most obvious audible effect was high amounts of jitter at certain frequencies sounds much like the old analog "wow and flutter" that is produced by vinyl turntables and tape recorders. Both devices have tiny deviations in speed--they might be slow variations (wow) like a vinyl record with the hole punched off-center. Or they might be fast (flutter) caused by the motor not rotating smoothly or bad bearings. Digital jitter, interestingly, often produces a rather similar end result.

PIANOS AND CYMBALS ARE CHALLENGING: Acoustic piano music seems to be an especially sensitive indicator of jitter (and wow and flutter). As the jitter increases, the notes take on a more brittle quality--that expensive Steinway Grand now sounds more like a garage-sale upright piano. And when the jitter is really high at certain frequencies, you can hear a "warble" to sustained piano notes--just like with low quality tape players or turntables. Some also say cymbals are good at revealing jitter.

WHAT IS JITTER ITSELF? Imagine a ticking clock that keeps perfect time over the long haul, but the second hand might move every 0.9, 1.0, or 1.1 seconds. If the variations average out evenly, the clock still keeps perfect time. But if you try to measure just 1 second, it might be off by 10%. That's jitter. The 44.1 Khz clock used for CD audio averages out to the right rate but the individual “ticks” may not be so accurate. These variations introduce a form of time distortion into the music. Excessive jitter is typically caused by poor design, cost savings, poor interfaces, or other shortcuts.

JITTER IN SPECTRUM VIEW: If you look at the audio spectrum both wow & flutter and jitter manifest themselves as symmetrical sidebands to a fundamental pure frequency. And very low frequency jitter causes the base of the pure tone to “spread”. The height and extent of the spread indicates the magnitude of low frequency jitter. The height and number of the side bands indicates the magnitude of higher frequency jitter.

SAMPLING ACCURACY: To properly digitize and re-construct digital audio it's relatively important the sampling intervals be as accurate as possible between the recording process and when it's played back. Random or periodic variation to these intervals along the way is jitter. It gets complicated (and controversial) to describe all the possible sources, and interactions, of jitter but some things are fairly well understood and agreed upon.

JITTER SOURCES: Clock circuits are everywhere. Pretty much anything with a CPU—right down to the little keyfob that unlocks your car’s doors—has a clock in it. For some applications, like a wristwatch, the important spec is how accurate the clock is when it’s manufactured, over time, and with changes in temperature. If it averages above or below the designed frequency the watch won’t keep accurate time. But jitter in this application virtually doesn’t matter. It could be relatively huge and the watch will still work exactly as expected.

CLOCK PHASE NOISE: Another name for clock jitter is “phase noise”. The clock for digital audio equipment has to be both accurate in frequency (otherwise the audio will play fast or slow) and have low jitter (to avoid significant sampling error). The lower the jitter the more expensive the clock circuitry becomes. All clocks have some inherent jitter even if they’re operated in a perfect isolated environment, from a perfect power supply, etc. And when you put them in a less than perfect environment—typical consumer electronics—the jitter will usually get worse because there’s noise on the power supply, ground noise, and likely some EMI (electromagnetic noise) from other parts of the circuitry. Just changing the PC board layout can have a significant impact on jitter performance even using the exact same parts.

JITTER REDUCTION: Some audiophile manufactures making jitter reduction claims are likely just touting the features built into whatever chip(s) they’re using and going off the datasheet for the part(s). Some, I know for a fact, don’t even have equipment to even properly measure jitter. So it’s not surprising some of the products from these manufactures have lousy jitter performance despite the high-end parts used, and their marketing claims. You can’t take jitter performance for granted, design only by ear, etc. You have to measure it.

JITTER REDUCTION METHODS: Various methods are used for jitter reduction. Some use a PLL, a double PLL, Asynchronous Sample Rate Conversion (ASRC) techniques (as Benchmark does), and now some USB products use a different form of the USB Audio interface specification known as Asynchronous. Each company will often argue their way is the best way. But, in reality, each has its own strengths and weaknesses. The best way I know to compare them is to evaluate the jitter spectrum using a J-Test.

jitter 350CABLE JITTER: Cables can “smear” digital signals by attenuating the highest frequencies. The diagram to the right is an example. The top blue waveform illustrates a perfect digital S/PDIF or AES3 bitstream. The bottom red waveform is what you might get out the other end of a long cable. The hardware receiving the signal uses the “zero crossings”—where the signal transitions an imaginary line drawn horizontally through the middle of the red waveform—to extract the clock. As you can see by the red arrows, that isn’t always correct due to the waveform distortion. The gaps between the red arrows are jitter. The amount of “smearing” depends on the bitstream itself so as the audio signal changes so does the clock timing creating jitter related to the audio itself.

RECORDING JITTER IS A MORE SERIOUS PROBLEM: If the device doing the A/D conversion during recording has a poor clock, or there are other clocking issues (such as a multi-track with multiple A/D devices recording simultaneously), the music will not be reliably sampled with low jitter. And, contrary to what many manufactures and the audiophile media would like you to believe, you cannot correct for recorded jitter during playback. Even those uber-expensive "jitter correction" devices like the Genesis Digital Lens can't fix recorded jitter. This is partly why I invested nearly $2000 in the Benchmark ADC1. I wanted an A/D that had reference quality jitter performance as you're stuck with whatever the A/D hardware gives you.

FURTHER INFORMATION: Here are 3 links including Benchmark's excellent description of jitter and how they deal with both measuring it and eliminating it, a generic description of jitter in general (not just audio jitter) on Wikipedia, and a rather wordy and technical PDF of an Audio Critic issue which had Jitter as their feature article:

Jitter Effects & Measurement (by John Siau, Benchmark Media)

Sound on Sound Benchmark Review With Jitter Discussion (an independent verification of Benchmark’s method)

Jitter on Wikipedia

Audio Critic Jitter Issue (pdf)

MEASURING JITTER FROM AN ANALOG SIGNAL: There are various ways to measure or evaluate jitter depending on what you're testing. If you're forced to test in the analog domain (i.e. gear with only an analog output), you generally use a particular FFT spectrum analysis of certain test signals and visually interpret the results. The question is what signal to use?

J-TEST: Julian Dunn at Prism Sound (makers of the dScope) developed the Jtest as a way to provoke “worst case” jitter for the purposes of measurements. His work was published by the Audio Engineering Society and has become something of a benchmark for Jitter measurements. The Jtest signal creates a jitter “torture test” at exactly 1/4 the sampling rate combined with toggling the lowest bit in a way that exposes jitter. I use the same Jtest signal with a similar spectrum analysis and my results should compare well those in Stereophile and elsewhere.

SIDEBANDS: As mentioned earlier, periodic higher frequency jitter is revealed as symmetrical pairs of “sidebands” around the main signal—this is analogous to the old mechanically induced flutter. The “height” or peak level of these sidebands indicates the relative amplitude (in picoseconds or nanoseconds) of the jitter components. While “spread” at the base of the main 11,205 hz signal indicates random low frequency jitter—this is analogous to the old mechanically induced wow. The greater the spread, and higher in amplitude it reaches, the greater the low frequency jitter amplitude.

MEASUREMENT VALIDATION: Variations of this method has been well documented and verified by many. It relies on the fact that nearby symmetrical sidebands to the high frequency signal are not likely to be from anything but jitter. The difference in frequency is much less than a single sample period of the digital signal. So “distortion” components this close, in matched pairs, are extremely likely to be jitter related. Here's a description that's less technical than most and also serves to help validate the Benchmark link above: Measuring Jitter

PC SOUND OUTPUT: I've used the above technique and can say it works. Some devices have minimal jitter, and some have fairly obvious jitter. Here, for example, is the Jtest jitter spectrum of a typical PC’s audio output. The high amount of noise on the motherboard likely corrupts the digital audio clock and/or bitstreams creating more sources for high frequency periodic jitter than usual. There’s also significant “spread” indicating low frequency jitter:

Intel IDT 92HD73E G45 Motherboard Jitter


M-AUDIO TRANSIT USB INTERFACE: The M-Audio shows much more “spread” all the way up to about –88 dB and fewer sidebands slightly lower in level than above:

m-audio transit jtest jitter


FiiO E7 USB DAC: This is an improvement over both of the above showing less spread and a rather different distribution of jitter sidebands:

FiiO E7 JTest Jitter Spectrum


SO HOW MUCH IS TOO MUCH? Based on all the research I’ve done, if the total sum of all the jitter components is less than –100 dBFS within the audio band, the jitter will be inaudible. I’ve further tried to clarify this by developing a “limit guide line” with individual components being under –110 dBFS with a bit more leniency close to the fundamental test signal but still under the –100 dBFS limit. Here’s an example of the DAC1 with the limit line shown in green. You can see the $1600 DAC performs very well on jitter:

DAC1 Jitter 11025hz -3 dBFS J-Test & Phase Both Ch Ref 2 Vrms 24-44


MEASURING A DIGITAL SIGNAL: In the digital domain, if you're say looking at a digital signal coming out of an A/D converter, CD/DVD/Media player with a digital output, etc. you can directly measure the jitter with the right instrument. The Prism Sound dScope has a low jitter internal time base clock. This allows the dScope to compare the clock in the digital signal to its own quality time base. From the timing differences it can calculate an actual jitter value in the time domain.

OSCILLOSCOPE MEASUREMENTS: Some think if you look at a digital signal on a scope with an infinite persistence feature the “smear” you get is jitter. But it’s not. What you’re really seeing is variations in the scope’s trigger point. Because the scope is entirely triggering on the incoming waveform, it’s automatically trying to filter out any jitter. The only way to measure jitter is to compare the signal in question to a known highly stable source. And scopes are not, by themselves, generally set up to do that.

SUMMARY: My thoughts on jitter are pretty simple:

  • It doesn't matter how a digital audio file is stored or if lossless compression is used. A file played back on a PC, network file server, portable flash-based player, WAVE, FLAC, etc. are all the same with respect to jitter at the file level. It's what happens later in the playback process that matters.
  • Many devices exist which effectively reduce playback induced jitter to very likely inaudible levels. If you're worried about jitter it's easy enough to choose one of these devices--especially ones that back up their marketing claims with some actual valid test results (like Benchmark, Anedio, etc.).
  • If you’re worried about jitter, choose a device that has been independently tested for jitter with similar jitter spectrums to those shown in this article. Look for sidebands and “spread” under –110 dBFS.
  • Jitter present on recordings, from the A/D process, is a more serious problem and cannot be removed by any playback hardware.