Batch Conversion & Recording

---

Batch Mode

DpdoEngine supports batch processing of multiple files via shell scripting. There is no dedicated --batch flag — a shell loop does the job easily.

Shell Batch Conversion

Bash example — batch FLAC → DSF:

#!/bin/bash
for f in *.flac; do
  DpdoEngine -i "$f" "${f%.flac}.dsf" --oversampling 128 --order 7
done

Recursive processing of subdirectories:

#!/bin/bash
find /music/pcm -name "*.flac" | while read f; do
  outdir="/music/dsd${f#/music/pcm}"
  mkdir -p "$(dirname "$outdir")"
  DpdoEngine -i "$f" "${outdir%.flac}.dsf" --oversampling 128 --order 7
done

Multi-Core Parallel Conversion

Use xargs or GNU parallel to fully utilize multi-core CPUs:

# xargs — 4 parallel jobs per batch
find . -name "*.flac" -print0 | xargs -0 -n1 -P4 -I{} sh -c \
  'f="{}"; DpdoEngine -i "$f" "${f%.flac}.dsf" --oversampling 128 --order 7 -opencl'

# GNU parallel (recommended)
find . -name "*.flac" | parallel DpdoEngine -i {} {.}.dsf --oversampling 128 --order 7

Tip: combined with --opencl, GPU acceleration is especially efficient in multi-task scenarios.

---

Pipe Mode

--pipe enables stdin/stdout piping so DpdoEngine can be combined with other command-line tools to build flexible processing pipelines.

Basic Usage

# Read PCM from stdin, write to stdout
cat input.wav | DpdoEngine -i pipe output.dsf --oversampling 64 --pipe

# Chain with ffmpeg
ffmpeg -i source.flac -f wav - | DpdoEngine -i pipe output.dsf --oversampling 128 --pipe

Pipeline Examples

Multi-stage chain:

# PCM source → loudness normalization → noise shaping → DSD output
ffmpeg -i track.flac -f wav - | \
  DpdoEngine -i pipe track.dsf --oversampling 128 --order 7 --lufs -16 --pipe

Real-time monitoring with DpdoPlayer:

# Convert and monitor through DpdoPlayer at the same time
DpdoEngine -i input.flac output.dsf --oversampling 64 --pipe | DpdoPlayer - -vol 80

---

Record Mode

DpdoEngine can record directly from compatible audio interfaces. The goal is lossless direct-to-DSD recording.

Basic Recording

# Start recording to a DSF file
DpdoEngine --record recording.dsf

Recording Options

Parameter	Description	Example
--record	Enable record mode (mono default, press q to stop)	--record
--record-device NAME	Recording device name (default: system default)	--record-device "USB Mic"
--record-samplerate HZ	Capture sample rate (0=auto)	--record-samplerate 44100
--record-channels N	Capture channels (0=auto)	--record-channels 2
--record-duration SEC	Auto-stop after SEC seconds (0=manual)	--record-duration 300
--record-buffer N	Frames per buffer (64–65536, default 512)	--record-buffer 512
--record-headroom DB	Headroom dB (default 3.0)	--record-headroom 3
--list-devices	List available audio devices	--list-devices

Output format is determined by the output file extension (.dsf / .dff).

Native DSD Recording

For audio interfaces with native DSD output, DpdoEngine records the DSD stream directly (DoP or ALSA raw DSD):

# DoP mode (PortAudio, broad compatibility)
DpdoEngine --record-dsd --dsd-mode dop live.dsf

# Native mode (ALSA raw DSD_U32_BE, requires device support)
DpdoEngine --record-dsd --dsd-mode native live.dsf

The recorded DSD stream never passes through a PCM intermediate format — the PDM bitstream goes straight from the ADC into a DSF file, keeping the recording chain maximally clean.

Record + Real-Time Monitoring

Record and monitor through DpdoPlayer simultaneously, zero latency:

DpdoEngine --record recording.dsf --pipe | DpdoPlayer - -vol 70

---

Statistics & Logging

Statistics

DpdoEngine -i input.wav output.dsf --oversampling 128 --stats

Reported statistics include: - Input/output sample rates - Processing time and utilization - Peak and RMS levels - Noise floor - Dynamic range

Verbose Logging

By default DpdoEngine prints progress to the terminal. To save a log:

DpdoEngine -i input.wav output.dsf --oversampling 128 2>&1 | tee conversion.log

---

Automation Script Reference

Full-Album Conversion (with metadata)

#!/bin/bash
# Batch-convert an entire album, adding metadata to each track
ALBUM="Symphony No. 9"
ARTIST="Beethoven"
YEAR=2024

for f in *.flac; do
  TITLE="${f%.flac}"
  DpdoEngine -i "$f" "${f%.flac}.dsf" \
    --oversampling 128 --order 7 \
    --artist "$ARTIST" \
    --album "$ALBUM" \
    --title "$TITLE" \
    --year "$YEAR"
done

Folder Watch + Auto-Convert

# Watch a folder with inotifywait, convert new files automatically
inotifywait -m /watch/folder -e close_write --format '%f' | while read f; do
  if [[ "$f" == *.flac || "$f" == *.wav ]]; then
    DpdoEngine -i "/watch/folder/$f" "/output/${f%.*}.dsf" --oversampling 64 --order 7
  fi
done

One-Shot Album Authoring with DpdoPackSACD

#!/bin/bash
# Batch convert → author SACD ISO
for f in *.flac; do
  DpdoEngine -i "$f" "${f%.flac}.dsf" --oversampling 128 --order 7 \
    --artist "Artist" --album "Album" --title "${f%.flac}"
done

# Author the SACD ISO
DpdoPackSACD --album "Album" --artist "Artist" -cover cover.jpg *.dsf