There are several reasons why one may encounter unwanted, sometimes random, pops and cracks when playing back recorded audio from either a PC or from digital audio equipment. Although generally this problem is certainly more commonly associated with computer based recording, it can also frequently occur whenever two or more digital audio hardware products are connected.
The source of pops and cracks within a digital audio stream can normally be attributed to one of four main factors, namely incorrect Word Clock settings, digital distortion, sound card or audio interface buffer size, or an incorrectly installed sound card or audio interface. Note that all four factors come into play with regards to PC recording, whereas only the first two apply to stand alone digital audio hardware products.
Please refer to blog covering Synchronisation for an explanation of Word Clock. In short, the effect of passing digital audio between two digital audio products, e.g. a Yamaha 01V96 and say a Fostex D2424LV, without correct Word Clock synchronisation will produce randomly appearing pops and cracks within the audio stream.
This effect is the result of the digital clocks within the two products, which drive the respective AD DA converters, sampling at different points in time. This problem can simply be corrected by ensuring that one product is running on its internal clock, for example the Fostex D2424LV, which in this scenario is the master, and that the other product (Yamaha 01V96) slaves to the Word Clock source supplied by the word clock master.
The same rule applies if connecting, say, a Tascam DM3200 digital audio console to a computer based recording system. In this instance, the PC is the Word Clock master and the DM3200 the slave, taking it’s Word Clock source from either a digital audio input (e.g. ADAT, SPDIF, AES/EBU or TDIF) or from its optional FireWire port. Also refer to blog covering Digital Formats.
Unlike analogue distortion, which if subtle can introduce a little warmth, digital distortion sounds dreadful. The result within recorded audio produces sharp cracks, and unlike the other effects generally described within this blog that may be corrected pending the issue, once digital distortion is recorded it cannot be removed generally rendering the recorded material useless.
Note that although Word Clock and buffer size issues may appear or manifest as random pops and cracks they may not necessarily be recorded within the audio stream, a point that depends upon the signal path. As a result, when the related issue is corrected the signal path can actually resolve itself.
However, digital distortion cannot be removed post event. Therefore it is critical within the process of recording digital audio to pay particular attention to ensure that distortion or ‘clipping’ of audio never occurs. Compression, pre AD conversion, proves to be an extremely useful and highly effective technique, particularly on dynamic signals such as vocals to prevent distortion from occurring. There are several products designed for this purpose, in particular Channel Strips, e.g. Focusrite Voice Master and Trak Master are ideally suited to these applications.
By adjusting the buffer settings of any sound card or audio interface one can effect or adjust the associated latency time, or delay of a PC with regards to passing an audio signal. Note that the lower the buffer size the lower the latency, therefore it is natural to try and set buffers as low as possible. However, there is a significant cost to be paid for doing this.
By simply halving the buffer size the user will actually double the workload or demand upon the CPU of any PC or Mac. Most audio interfaces, e.g. Presonus Firepod or Saffire Pro 26, will install with an automatic buffer setting of 1048K, although options to reduce this setting to as low a 128K may be present. The reductions follow the path of 1048K, 512K, 256K, 128K, therefore the PC will be running 8 times faster to handle the same amount of audio or work as at 1048K if buffer sizes are reduced to 128K.
As the old saying goes ‘you cannot have your cake and eat it’, same here I’m afraid! By running extremely low buffer settings the CPU of the host PC will be rather stressed and as a result may struggle to perform as required without overloading. Whenever the CPU of a PC overloads the system generally stutters, causing random pops and cracks. The issue can obviously be resolved by either increasing buffer settings, which in turn reduces overall workload or by limiting the number of tracks, effects, EQ and soft synth usage if wishing to maintain low buffer settings.
Incorrect Installation or Conflict
Pops and cracks may also result from either conflicting hardware settings, e.g. an IRQ conflict between a soundcard and say graphics card, or from an incorrect installation (refer to blog covering Hardware Drivers & Soundcard Drivers). If the issue is a result of an incorrect installation, this can generally be resolved by reinstalling either the soundcard or audio interface whilst adhering 100% to the manufacturers’ instructions.
Note that conflicts are somewhat more complex and as a result may or may not be resolved pending the issue, note that IRQ conflicts can generally be overcome by repositioning the soundcard in another PCI slot. However, certain soundcards or audio interfaces are simply incompatible with certain motherboard chip sets – an issue that cannot be resolved. Although certain products are definitely more tolerant, e.g. Tascam US144, EMU 0404 USB, Presonus Firepod and the RME Fireface series, than others that are currently available on the market place.