6
1 Configuration
Extended Bit Depth Audio
Traditionally, PC-based audio capture and playback has been restricted to
either 8 bit or 16 bit encoding, with the WAVE_FORMAT_PCM descriptor
used to define the audio parameters such as samples per second, bits per
sample and number of channels.
With the advent of 24 bit A/D and D/A converters, there was a need to
support extended bit depth. To achieve this, Microsoft introduced a new
data format descriptor called WAVE_FORMAT_EXTENSIBLE. This provides
additional information, including the number of valid bits per sample and
support for specifying channel placement in multi-channel streams.
The WAVE_FORMAT_EXTENSIBLE descriptor may be used in place of
the normal WAVE_FORMAT_PCM in setting up playback and recording
through either the Winmm (wave) API or the DirectSound API. For further
information refer to the Microsoft DirectX 8 documentation.
Some audio editing applications, such as Syntrillium’s Cooledit 2000 and
Adobe Audition, support extended bit depth audio and work well with the
Auricon 4.4 using 24 or 32 bit recording.
Hyperthreading & Audio
Some motherboard processors offer hyperthreading, a technology lying
midway between single core and fully independent multiple processor
cores. A hyperthreaded processor provides parallel execution paths for
multiple threads, however there remain some shared resources such as the
memory caches and the floating point processor.
For most applications this is fine and results in a substantial improvement
in performance for a given processor complexity, however for time-critical
audio applications a problem can arise, when a low-priority thread is using
a shared resource that the high-priority audio thread requires. Since the
processor has no concept of thread priorities, the high-priority thread is
blocked until that resource becomes free, resulting in gaps on playback or
skipping while recording.
If this proves to be a problem on a hyperthreaded system, the
hyperthreading can usually be disabled in one of the BIOS settings. Refer to
the motherboard or system documentation for details on how to do this.
Note
that this problem does not occur on true multicore processors.
Connecting the Auricon 4.4 to
Unbalanced Inputs and Outputs
Most professional audio equipment, including the Auricon sound card, uses
high level balanced audio interconnections. Commercial radio stations in
Australia typically use a nominal programme level of +8dBu, while the ABC,
SBS and many production houses use +4dBu. In the USA 0dBu is commonly
used.
By contrast, domestic audio equipment uses low level unbalanced
interconnections, typically at a level of about 100mV rms (about -18dBu).
The best way to interconnect between these two systems is with a
balancing amplifier. Failing that, it is still possible to achieve an acceptable
interconnection, particularly when going from a high level balanced output
to a low level unbalanced input.
In the diagram below, the transformer does the balanced to unbalanced
conversion and the resistors adjust the level. It is preferable to use a
1:1 transformer and the resistive attenuator rather than a stepdown
transformer as this will help minimise capacitive coupling of common mode
noise across the transformer windings.
Note
that you should
never
ground one side of a balanced output. Doing so
will most likely damage the output drivers.
