Direct Stream Digital (DSD)¶

Direct Stream Digital (DSD) is used for digitally encoding audio signals on Super Audio CDs (SACD). It uses pulse-density modulation (PDM) encoding.

lib_xua supports DSD playback from the host via “DSD over PCM” (DoP) and a “Native” implementation which is, while USB specification based, proprietary to XMOS.

DSD is enabled with by setting the following define to a non-zero value:

Table 26 DSD defines¶
Define	Description	Default
`DSD_CHANS_DAC`	Number of DSD channels	`0` (Disabled)

Typically this would be set to 2 for stereo output.

By default both “Native” and DoP functionality are enabled when DSD is enabled. The Native DSD implementation uses an alternative streaming interface such that the host can inform the device that DSD data is being streamed. See: Audio stream formats for details.

If only DoP functionality is desired the Native implementation can be disabled with the define in Table 27.

Table 27 Native DSD defines¶
Define	Description	Default
`NATIVE_DSD`	Enable/Disable “Native” DSD implementation	`1` (Enabled)

DSD over PCM (DoP)¶

DoP support follows the method described in the DoP Open Standard 1.1.

While Native DSD support is available in Windows through a driver, macOS incorporates a USB driver that only supports PCM, this is also true of the central audio engine, CoreAudio. It is therefore not possible to use the “Native” scheme defined above using the built in driver of macOS.

Since the macOS only allows a PCM path a method of transporting DSD audio data over PCM frames has been developed.

Standard DSD has a sample size of 1 bit and a sample rate of 2.8224MHz - this is 64x the speed of a compact disc (CD). This equates to the same data-rate as a 16 bit PCM stream at 176.4kHz.

In order to clearly identify when this PCM stream contains DSD and when it contains PCM some header bits are added to the sample. A 24-bit PCM stream is therefore used, with the most significant byte being used for a DSD marker (alternating 0x05 and 0xFA values).

When enabled, if USB audio design detects a un-interrupted run of these samples (above a defined threshold) it switches to DSD mode, using the lower 16-bits as DSD sample data. When this check for DSD headers fails the design falls back to PCM mode. DoP detection and switching is done completely in the Audio/I2S thread (xua_audiohub.xc). All other code handles the audio samples as PCM.

The design supports higher DSD/DoP rates (i.e. DSD128) by simply raising the underlying PCM sample rate e.g. from 176.4kHz to 352.8kHz. The marker byte scheme remains exactly the same regardless of rate.

Note

DoP requires bit-perfect transmission - therefore any audio/volume processing will currupt the stream.

“Native” vs DoP¶

Since the DoP specification requires header bytes this eats into the data bandwidth. The “Native” implementation has no such overhead and can therefore transfer the same DSD rate at half the effective PCM rate of DoP. Such a property may be desired when supporting DSD128 without exposing a 352.8kHz PCM rate, for example.

Ports¶

The codebase expects 1-bit ports to be defined in the application XN file for the DSD data and clock lines for example:

<Port Location="XS1_PORT_1M"  Name="PORT_DSD_DAC0"/>
<port Location="XS1_PORT_1N"  Name="PORT_DSD_DAC1"/>
<Port Location="XS1_PORT_1G"  Name="PORT_DSD_CLK"/>

Note

The DSD ports may or may not overlap the I2S ports - the codebase will reconfigure the ports as appropriate when switching between PCM and DSD modes.