3-22 Operation
x HD (The output of the 5950’s HD audio path)
Synchronization determines if the sample rate appearing at the digital-channel output is synced to the 5950’s
internal clock, to an AES3 signal appearing at the 5950’s digital input, or to an AES11 signal appearing at the
5950’s sync input. Sync can be set separately for Digital Output 1 and Digital Output 2, allowing them to have
different sample rates.
The selections for each of the two AES outputs are Internal, Sync In, and Input. Input sets a given AES3 output
sample rate and synchronization to the same sample rates.
The selections for each of the two AES outputs are Internal, Sync In, and Input. Input sets a given AES3 output
sample rate and synchronization to the same sample rate present at the 5950’s AES3 (audio) input. Likewise,
Sync In uses the AES11 sync input’s sample rate and synchronization as the source. Internal synchronizes the
given AES3 output rate to the 5950’s internal clock and uses the Samp Rate setting to determine its output
sample rate.
For a given AES3 output, the output sample-rate selector (“Samp Rate”) has no effect in the Input and Sync In
modes unless sync is lost. Then the output reverts to internal sync at the sample rate that is preset by the
sample-rate selector for that output. Otherwise, the output sample rate follows the sample rate present at
the selected input, regardless of the setting of the output sample rate selector.
If no signal is provided to the 5950 Input or Sync In, set SR Sync to Internal and select the desired output
sample rate.
The 5950 Sample Rate can be set from 32kHz to 96kHz. Most systems will either use 44.1 or 48kHz.
Word Length sets the word length (in bits) emitted from the digital-channel output.
The largest valid word length in the 5950 is 24 bits. The 5950 can also truncate its output word length to 20,
18, 16, or 14 bits. The 5950 can also add dither, which we recommend.
Dither turns on or off addition of “high-pass” dither before any truncation of the output wordlength.
The amount of dither automatically tracks the setting of the Word Length control. This first-order noise
shaped dither adds considerably less noise in the midrange than does white PDF dither. However, unlike
extreme noise shaping, first-order noise shaped dither adds a maximum of 3 dB of excess total noise power
when compared to white PDF dither. It is thus a good compromise between white PDF dither and extreme
noise shaping.
In many cases, the source material has already been correctly dithered so you will not need to add dither and
can set this control to Out. However, particularly if you use the Noise Reduction feature, the processing can
sometimes attenuate input dither to a point where it is insufficient to dither the output correctly. In this case,
you should add dither within the 5950 by turning this control on.
Format determines if the digital-channel output follows the professional AES3 or consumer SPDIF standard.
We expect that AES will be appropriate for almost all users, but some consumer sound cards may require
SPDIF.
Output 2, just below the Output 1 controls, operates in the same manner as Output 1.