Important note: this is a fork of original RTPProxy tool focused exclusively on extractaudio tool.
These modifications were mostly vibe-coded mainly to quickly serve one task - extract audio from pcap files produced by sngrep utility with possibility to force the output codec.

A powerful Docker-based audio extraction tool for RTP streams from PCAP files, rtpproxy recordings, and RTP stream data.

The extractaudio-docker.sh tool provides enhanced audio extraction capabilities with:

• Automatic Docker image building and management
• Support for multiple audio formats and codecs
• Linux SLL format PCAP conversion
• True stereo extraction from dual RTP streams
• SRTP encrypted stream support (when compiled with libsrtp)

The tool automatically builds the required Docker image on first use. To manually build:

./extractaudio-docker.sh --build-image

Basic syntax:

./extractaudio-docker.sh [wrapper options] [extractaudio options] input_file output_file

• --build-image - Force rebuild of Docker image
• --show-info - Show Docker image information and codec support
• --shell - Open interactive shell in container
• --direct - Skip Linux SLL conversion (use original extractaudio)
• --true-stereo - Split RTP streams by SSRC for true stereo (default with -s)
• --mixed-stereo - Use single stream mixed to stereo (legacy mode)

• -d - Delete input files after processing

• -s - Enable stereo output (2 channels)

• -i - Set idle priority for processing

• -n - Disable synchronization (nosync mode)

• -e - Fail on decoder errors instead of continuing

• -S - Scan mode - analyze files without extracting audio

• -F FORMAT - Output file format (wav, aiff, flac, ogg, etc.)

• -D FORMAT - Output data format (pcm_16, pcm_24, float, etc.)

• -A FILE - Answer channel capture file

• -B FILE - Originate channel capture file

• --alice-crypto CSPEC - Crypto specification for Alice (answer) channel
• --bob-crypto CSPEC - Crypto specification for Bob (originate) channel

CSPEC format: suite:key[:salt]

Basic mono extraction:

./extractaudio-docker.sh -F wav call.pcap output.wav

True stereo from dual RTP streams:

./extractaudio-docker.sh -s -F wav call.pcap stereo_output.wav

Separate channel files:

./extractaudio-docker.sh -A answer.pcap -B originate.pcap -s output.wav

High-quality PCM output:

./extractaudio-docker.sh -F wav -D pcm_16 call.pcap hq_output.wav

FLAC lossless compression:

./extractaudio-docker.sh -F flac -D pcm_24 call.pcap lossless.flac

Scan mode (analyze without extraction):

./extractaudio-docker.sh -S call.pcap

• PCAP files (.pcap extension)
• Linux SLL format (automatically converted)
• rtpproxy recording directories (rdir.a.rtp, rdir.o.rtp)
• Individual RTP stream files

• G.711 μ-law (PCMU) - payload type 0
• G.711 A-law (PCMA) - payload type 8
• G.729 - payload type 18
• G.722 - payload type 9
• GSM - payload type 3
• Opus - dynamic payload types

• WAV, AIFF, AU, RAW, PAF, SVX, NIST, VOC
• IRCAM, W64, MAT4, MAT5, PVF, XI, HTK
• SDS, AVR, WAVEX, SD2, FLAC, CAF, WVE
• OGG, MPC2K, RF64

• Docker must be installed and accessible
• First run may take several minutes to build the image
• File paths are automatically mounted into the container

For detailed options and examples:

./extractaudio-docker.sh --help

For Docker image information:

./extractaudio-docker.sh --show-info