Technical Overview
A Fully In-House DSD Processing Solution
Every layer of Dpdo — from core algorithms to the user interface — is developed in-house, with no third-party DSP libraries or commercial audio frameworks. This gives us exact control over every stage of the digital-to-analog conversion chain, delivering a truly end-to-end DSD processing experience.
64-Bit Double-Precision Computation
Most audio software uses 32-bit single-precision floating point internally. DpdoEngine uses 64-bit double-precision floating point throughout, pushing cumulative rounding error far below the threshold of audibility to theoretically unmeasurable levels. This matters most in high-order noise shaping and long-tap FIR filtering — every extra processing stage introduces new error, and double precision keeps the accumulation of these errors inaudible.
15th-Order IIR Noise Shaping
Noise shaping is the most critical stage in PCM-to-DSD conversion. DpdoEngine supports IIR noise-shaping filters from 3rd to 15th order — the higher the order, the better the in-band signal-to-noise ratio, and the higher the demands on computational precision and stability. 15th-order IIR noise shaping is among the highest orders publicly implemented in the industry.
DST Lossless Compression
A DSD audio file is roughly twice the size of CD-quality PCM. DST compression reduces DSD streams losslessly by 30–50% while remaining bit-exact on decode. DpdoEngine supports DST encoding natively — no external compression tool required.
A Complete Chain from Recording to Playback
The Dpdo suite covers the entire DSD production pipeline:
- Record — DpdoEngine records native DSD streams directly from compatible ADCs
- Convert — convert PCM masters or recordings to high-spec DSD
- Author — package tracks into SACD ISO images for archiving or distribution
- Play — DpdoPlayer outputs native DSD streams for true high-fidelity playback