Panopticon — the dashboard

Now-playing card with album art (auto-fetched from iTunes when automation doesn’t supply it). Every encoder shows its current HLS segment, byte count, segment number, and on-air status. Every mount shows live connection state with the CDN. Service controls let you restart audio capture, automation parsing, or delivery independently — no full reboot needed.

Audio capture

Capture from any WASAPI device on Windows — your sound card, a USB interface, a virtual cable from another app. Optional WASAPI loopback for capturing system audio. Optional incoming RTP for receiving feed from a remote codec or another StreaminDemon instance acting as a relay.

The Level Devil — four-band audio processor

Studio-grade processing chain runs before the encoder. Wideband AGC (target RMS, attack/release/max-gain), four-band compressor (Bass, Mid, Presence, Air with per-band threshold/ratio/attack/release/makeup/gate), brick-wall limiter with lookahead, stereo enhancer for width adjustment, declipper, and pre-encode DSP per codec — automatic TPDF dither, high-shelf trim tuned for AAC vs MP3 vs HE-AAC, optional M/S width reduction. The Torment Level meter grades your overall loudness from sloth (under-compressed) to gluttony (over-cooked).

Codecs

aac (LC) heaac mp3 flac opus ogg vorbis

libfdk_aac available for AAC-LC where licensed. MP3 high-quality mode. Multiple bitrates per codec, multiple codecs simultaneously — one PCM source, parallel encoder pipelines.

Repetitive Tortures — delivery

Icecast SHOUTcast Futuri CDNStream1 AIS HLS → S3

One source, many destinations. Each mount is independent — separate codec, bitrate, credentials, retry behavior. Multiple Futuri ingest nodes (East, West, North, LAX) per mount. ID3v2.3 metadata embedded in every AAC segment so any HLS player gets in-stream now-playing data without polling a sidecar API.

Crystal Ball — automation integration

WideOrbit Enco DAD RCS NexGen RCS Zetta Rivendell RadioBoss iMediaTouch AudioVault Maestro FAAI Custom listener

Each system’s wire format is parsed natively — no middleware, no XML translators to maintain. The custom listener handles anything else: TCP server, TCP client, UDP, or file-watch; XML, JSON, delimited, or auto-detected. AI-assisted field mapping turns unknown formats into properly-tagged tracks.

SoothSettings — config and license

License activation and machine binding (one key, one machine, deactivate to move). Audio input selection — WASAPI device, RTP listener, RTP relay output for sending the clean processed feed to a peer. FFmpeg path. Monitor output device + level. Ad-insertion category configuration: pick the cart-prefix categories that should trigger the ADTS originality/copy bit flip in your encoded stream so downstream ad-replacement systems know when to splice.

Operational extras

• Token-delay alignment — millisecond-accurate alignment of metadata to audio, even with processor latency.
• GPIO talk-trigger — serial-port GPIO drives a “talk mode” flag for live segments. Latching, momentary, or flipped.
• Per-mount pre-encode DSP — different EQ and width settings per output, tuned for each codec.
• Live operator UI — dashboard, automation config, mount-point management on localhost:5100. SignalR pushes audio levels and now-playing in real time.
• Diagnostic sniffer — capture and replay automation traffic to debug integration issues without bothering on-air.
• Cloud settings backup — export and restore your full configuration from the cloud, key-tied to your license.