What Happens Before the First Note?
Startup latency, prefix rendering, buffering and why playback quality begins before audio reaches the DAC.
Press Play.
For the listener, this is a single action. Inside a high-rate DSD player, a surprising amount of work may need to happen before the first sample reaches the DAC.
The Startup Pipeline
A PCM source may require:
- container parsing;
- decoding;
- sample-rate conversion;
- sigma-delta modulation;
- DSD or DoP framing;
- buffering;
- and finally submission to the audio device.
If every stage waits for the previous stage to finish a large block of work, startup quickly becomes unpleasant.
Prefix Rendering
One solution is to treat the beginning of playback differently from long-running continuation.
Instead of preparing an entire track, the player prepares a bounded prefix sufficient to start safely.
Playback can begin while later material continues to be generated.
Why Higher Rates Matter
At DSD64, many inefficient designs can still appear fast enough.
At DSD512 or DSD1024, those inefficiencies become visible.
The amount of output data grows dramatically, and apparently small scheduling decisions begin to dominate the user experience.
Startup latency is often a systems problem disguised as a performance number.
Buffering Is Not Just Insurance
Buffers are often described as protection against timing irregularities.
In a modern playback engine they can do much more: they define where work occurs, when expensive conversion is allowed to happen, and how far the player can look ahead.
Conclusion
The first note is the result of work performed before the listener hears anything.
Designing that invisible interval well is one of the reasons a high-rate audio player can feel immediate despite doing considerably more computation than an ordinary PCM player.