In 1995 there were no easy solutions to this problem but it was possible to build what we wanted.

Paul McGowan writes:

In my previous post Jittered, we used a water flow analogy to further our understanding of jitter and how to make it go away. You’ll remember that jitter is all about timing and making sure that the digital audio bits arrive on schedule without any slow downs or speed ups to their progress. What you hope for is the source equipment that’s sending out the digital audio hasn’t any jitter, but that’s a lot to hope for. Certainly a lot to hope for back in the mid 1990′s. At that time the CD player was relatively new and the only source of digital audio. These early machines had a lot of jitter.

In our quest for better sound from digital audio we knew the high levels of jitter had to be reduced or eliminated. But how? We couldn’t actually start from scratch building a CD player at that time, like we can now, so our only two choices were to reduce jitter at the DAC’s input or place something between the CD player and the DAC. And the latter is what came about as a new class of product called Jitter Boxes. There were a bunch of them and they all worked pretty much the same way. With a regulator of sorts.

For the record, the first de-jitter device was designed by our friend and engineer Doug Goldberg (who also invented our MultiWave for Power Plants). Goldberg’s first de-jitter device, The DTI (Digital Transmission Interface), was designed for Audio Alchemy owned by our friend Mark Schifter. This was circa 1990. It decoded S/PDIF, reclocked the data with a series of shift registers and two crystals (44/48). The DTI Pro, a later product from Audio Alchemy, added a DSP that did inerpolation and dither as well.

If you’ll recall our water plant analogy we had an uneven flow of water from our stream. Let’s relate that to the CD player outputting faster and slower bits of music. We could smooth out the irregular flow of bits by placing a regulator directly in the path of those bits. In water terms a regulator restricts some of the incoming flow to produce a somewhat steadier output: faster water can’t get through quickly and slow moving water is let through without restriction, thus the flow is steadier. In electronic terms our regulator is called a PLL (Phase Locked Loop) which does pretty much the same thing. Neither fix the problem of uneven flow, but both help. Band Aids to be sure, but if you’re bleeding, a Band Aid is a welcome thing to have. So too is a jitter reducer.

I have never been big on Band Aids or partially correcting problems when it’s possible to actually fix it perfectly. Just not my style. I am not alone. Let’s go back to our water plant manager, faced with the problem of uneven flow. His predecessor has installed flow regulators on every home and they helped, but customers still have problems. Our new manager finds those problems unacceptable and aims to fix them once and for all. He builds a water storage tank.

Our mythical engineer has figured out the only way to eliminate flow (timing) problems in the water supply is to have a large reserve tank full of water. During high demand and low supply periods, the tank has enough reserve in it to supply what’s needed and acts as a buffer. During high supply and low demand, the tank stores the excess and also acts as a buffer. And buffer is the key word here.

If only there was an easy way to add a storage tank to our digital audio stream, we could simply buffer the faster and slower bit rates and output a constant, jitter free stream of bits that would help our DAC sound its best. In 1995 there were no easy solutions to this problem but it was possible to build what we wanted.

Just not easy.