A Short Primer On Digital Audio

Though we are concentrating on old-fashioned analog audio. we’re well aware (and occasional use) digital technology to modify, enhance sound or read or write to digital media. We also know about those raging battles between the analog-only aficionados and digital proponents. Let’s give you a short overview from a neutral perspective.

Audio / Sound is noting but a series of air pressure differences and displacements  that makes our tympanic membrane swing to create impulses that our brain is able to work with. In order to make this article more easily digestible, we concentrate on a single audio wave.

sound1In the analog world we try to preserve this wave as it is. But this is unfortunately not possible. Every piece of equipment adds, removes or modifies the wave. Microphones are limited as to what they are able to hear, mixers and amplifiers add noise, speakers ‘color’ the sound because of the limitations of their membranes. In other words: Though modern equipment is quite excellent, it still only reproduces an approximation of the original sound.

The digital technology converts the sound into numbers which allows computers to work with it. There are pretty much two ways to do that. PCM (Pulse Code Modulation) measures the value of the sound curve in a given number of intervals.

sound_pcmThe “sample rate” defines how often those measurements are taken, the bit rate defines how many values can be assigned to specify the curve. An old 8 Khz 8-Bit recording has only 254 values to describe the full  curve .. 0 is silence, -127 the full minus deflection and  127 the full positive deflection. We only have the numbers between -127 through 127 to describe the wave and we can’t convert anything above 4 KHz. This is a very rough measurement comparable only to a very early 1920’s mechanical recording. Modern CD’s have a sample rate of 44,100 Khz and a bit rate of 16 bit. This allows to convert a sound wave of max 22 Khz and to describe the wave with numbers between -32,768 through 32,767

The other method of converting sound is called DSD (Direct Stream Digital). In this method, the curve will not be measured in absolute terms, there are no values that would tell us the exact height of the curve at any given time.

sound_dsd1DSD records the “Delta” or the difference between the samples. If we compare PCM and DSD, we would ask: What is the value of the curve at sample number 1100: PCM would give us a number describing the absolute value of the curve allowing us to to plot a point at the exact position within the curve. DSD would give us the difference  xy values more (or less) in regard to the previous value and we would not be able to plot a point without knowing (and calculating) all the previous samples starting at the beginning – which makes it pretty much impossible to digitally edit the sound without converting it into PCM.

Now .. what is “better” ?

As explained above there is no such thing like a “perfect” recording. All elements from the microphones to the record player, tape machine or DSD player and finally the speakers introduce their very specific limitations and problems.

PCM measures the sound curve in a given interval. And regardless how often the curve is measured there will always be those parts that are not measured. PCM will never be able to get the whole sound – regardless of the sample rate. But in all fairness:  High resolution PCM recordings (192 Khz, 24bit) are virtually indistinguishable from analog or DSD streams. The unmeasured parts are just to small to be noticeable.

DSD is closer to the original analog. It doesn’t ‘force’ the curve into absolute numbers (though even measuring the difference is limited of course). But I doubt that anybody would be able to tell the difference between high resolution PCM, DSD or the original analog if high end equipment is used.

The listening experience is much more defined to the production process of a recording, the positioning of the microphones, how the sound has been mastered. Needless to say: If the sound has been digitally edited or converted into an inferior format  (like CD format) it is permanently reduced to this quality – no matter what will be used in the final conversion process.

If one is looking for the highest quality listening experience, one should look for analog only processing or direct to DSD / high resolution PCM recording.