Behind the simple technology that keeps your clock ticking.
When we talk about getting the best possible sound from our digital audio systems, one topic that comes up is jitter. Jitter is a variation in clock timing that can lead to audio distortion.
Analogue-to-Digital (A-D) and Digital-to-Analogue (D-A) conversion relies on an analogue signal being represented as a series of digital words, most commonly 16 or 24 bits in length. However, those numbers can't represent a waveform accurately unless they are spaced apart in time very precisely — imagine trying to draw an accurate graph on a sheet of stretchy rubber rather than paper! Even tiny fluctuations in the clock that determines these time intervals can cause distortion in audio, whether on recording (A-D) or playback (D-A).
Naturally no clock is perfect, and the timing of pulses can vary a tiny bit from pulse to pulse — i.e. jitter. This gets worse if you have a lot of different digital devices that all have their own clocks, each with its own jitter, all running out of sync, although this part of the problem can be solved by forcing all devices in a studio to sync to one main clock. Then you only have one jittery clock to contend with rather than many.
One way to tighten the precision of a clock signal is with a phase-locked loop or PLL. The idea behind PLL is as follows: an input signal at a given frequency (like a clock) is compared to an output signal generated by a variable oscillator, by feeding back the output signal into a device that compares the phase of input and output. This causes a feedback loop that constantly adjusts and stabilises the frequency of the signal. (The filter in the diagram removes components of the signal that aren't desired in the output.)
In practice, doing this with clock signals had been quite difficult and expensive until the Spring of 2006, when a paper was presented to the Audio Engineering Society that described a new, cost-effective, highly accurate technology that could be applied almost anywhere in digital audio.
In simple terms, the new technology used two PLL circuits in a cascade, one analogue and one digital, whose unique behaviours complemented one another in a very neat way. Most critically, because of its low cost, this circuit could be placed inside any digital audio device, to smooth out even the tiniest variations in a main clock signal without introducing latency — so each device with this technology could be relied upon to have rock-solid clock performance.
This was a huge leap forward in clock accuracy, and it was called Jitter “Elimination" Technology or JetPLL™. (The patents and trademarks are held by developer Sonopsis, who put the quotation marks on “Elimination" to indicate that the process wasn't perfect – simply because no clock is perfect.) Equipment using JetPLL can reduce jitter to around one nanosecond (a billionth of a second!), far less than the sample rate of even the highest-quality audio. In other words, it helps make jitter practically nonexistent.
JetPLL started as part of a sync system based on the IEEE 1394 (FireWire) digital interface that was common at that time. It spread quickly to other types of communication, including various AES-approved digital audio standards running over cable or fibre optics, and then to networked audio systems such as Dante.
JetPLL is now everywhere in the world of digital audio, and Focusrite gear takes full advantage of its capabilities. For example, because JetPLL effectively removes jitter problems from audio networked over Dante, it's a fundamental part of the accurate, exceptional-quality sound of Focusrite RedNet.
Even without knowing the technical details of JetPLL — or even knowing that it's there — every musician and engineer can take advantage of JetPLL and what it does for audio. So, when you see JetPLL mentioned as a feature on audio gear, now you know what you're getting, and why it's important.
Words: Mike Metlay