Pre/Post-Alert Signaling

EAS Station™ can prepend or append a short attention signal to every outgoing SAME broadcast — auto-forwarded CAP/IPAWS alerts, OTA relays from upstream EAS sources, and operator-authored manual broadcasts. The signal plays:

• Pre-alert signal: before the very first SAME header burst.
• Post-alert signal: after the final EOM (End-of-Message) burst.

Signals are purely cosmetic — they are not part of the SAME / FSK signaling and decoders such as ENDECs and SAME-aware receivers will simply treat them as ambient audio. They are useful for station fingerprinting, paging-style call selection, and operator/listener attention before the official tones begin.

Important: signals are inserted at the broadcast boundaries. They never replace, mask, or modify the SAME headers, attention tone, narration, or EOM.

Where to Configure

The signal settings live in the Admin → Broadcast → EAS Encoder Settings → Pre/Post-Alert Signaling card. Settings persist in the eas_settings table.

The pre-alert and post-alert profiles are configured independently, so you can e.g. play a single bell before each broadcast and leave the post-alert position disabled.

Available Signal Profiles

none

No signal is generated. This is the default for both positions.

bell

A single 880 Hz sine tone with an exponential amplitude decay (≈ 5 % of peak amplitude after 2 s). Sounds like a struck bell or chime. Duration controlled by the signal-duration field.

beep

A sustained 1000 Hz sine tone for the full configured duration. Useful as a simple "stand by" attention beep.

three_tone

Three ascending sine tones (440 Hz, 880 Hz, 1320 Hz), each occupying roughly one third of the duration. Reminiscent of pager / paging-system call tones.

qc2 — Motorola Quick Call II

Two-tone paging used by fire, EMS, and dispatch radio systems.

• Tone A plays for 1 s.
• Tone B plays for 4 s.
• Total duration is 5 s (duration field is ignored).
• Set Tone A and Tone B frequencies to the values published by the receiving agency / department (typical range 288–3000 Hz).
• Optionally enable the Long Tone to append a sustained tone at Tone B frequency after the A+B sequence. Configure its duration (1–120 s) with the Long Tone Duration field.

Note for existing deployments: Prior to this change, Tone B was 3 s (4 s total). Existing QC-II configurations will automatically use the new 4 s Tone B (5 s total) without any settings change.

dtmf — Dual-Tone Multi-Frequency

Plays each character of the configured DTMF Sequence as its standard ITU-T Q.23 / Q.24 tone pair (low-group + high-group sine sum).

• Allowed characters: 0–9, A–D, *, #. Lower-case is accepted and upper-cased. Any other characters are stripped on save.
• Maximum length: 32 valid digits (extra digits are dropped).
• Timing is fixed at 100 ms tone / 50 ms gap per digit (the signal-duration field is ignored for DTMF). Total length therefore depends only on the number of digits.
• The two tones are summed and the combined waveform is scaled to half amplitude so the peak signal level matches single-tone profiles.

mdc1200 — Motorola Selective Calling

A 1200-baud FFSK packet (mark = 1200 Hz, space = 1800 Hz) carrying this station's Unit ID and an op-code that tells receiving Motorola subscribers what kind of call this is (PTT-ID, Emergency, Request to Talk, Call Alert, Selective Call, etc.).

• Packet duration is fixed by the protocol — the signal-duration field is ignored:

  • Single-packet ops (PTT-ID, Emergency, Request to Talk, Remote Monitor, Custom): 26 bytes / 208 bits → ~173 ms.
  • Double-packet ops (Call Alert, Selective Call): 40 bytes / 320 bits → ~267 ms. A second 14-byte info block carrying the target subscriber's ID is appended after a 4-byte inter-packet re-sync preamble.

• Unit ID is the 16-bit source identifier — the EAS station's own ID that receiving radios display as the caller (1–65535). Accepts both decimal (1234) and hex (0x04D2) input — Motorola CPS commonly displays unit IDs as 4-digit hex.

• Target Unit ID is the 16-bit destination ID for double-packet ops. When call_alert or selective_call is selected and a target ID is configured, the encoder produces the canonical 40-byte double frame; receiving radios programmed with that ID alert (Call Alert → beep + display) or unmute their speaker (Selective Call → audio path opens). Leaving the field blank falls back to a single-packet frame whose source ID also acts as the target — some receivers tolerate this shortcut, but real Motorola CPS-programmed subscribers usually expect the double form.

• Op-Code preset chooses what packet to emit:

  | Preset | op | arg | Frame | Meaning | |---|---|---|---|---| | ptt_id_pre | 0x01 | 0x80 | single | PTT-ID at the start of a transmission (default) | | ptt_id_post | 0x00 | 0x80 | single | PTT-ID at the end of a transmission | | emergency | 0x40 | 0x80 | single | Emergency alarm | | request_to_talk | 0x35 | 0x89 | single | Request-to-talk paging | | remote_monitor | 0x11 | 0x80 | single | Remote-monitor command | | call_alert | 0x63 | 0x85 | double | Page a target subscriber (target ID required for canonical form) | | selective_call | 0x35 | 0x80 | double | Voice Selective Call — unmute target subscriber | | custom | any | any | single | Operator-supplied raw op/arg bytes (decimal or 0x.. hex) |

• Smart pre/post pairing. When both the pre and post signals are set to mdc1200 and the op-code preset is ptt_id_pre (the default), EAS Station automatically substitutes ptt_id_post on the post-alert position so receiving Motorola radios see a complete bookend pair and close the call cleanly. All other presets — including call_alert and selective_call — pass through unchanged on both sides.

Verifying generated packets. Render any test alert and decode the generated WAV with a dedicated MDC1200 decoder such as mdc-decoder, which reads a WAV directly and reports the unit ID and op-code of each packet (e.g. PTT-ID Pre = op 0x01 / arg 0x80). Use this on the bench before going on the air with a real fleet. Note: mainline multimon-ng has no MDC demodulator (MDC is not a valid -a option) and cannot decode these packets.

Two-way LMR forwarding (the canonical use case)

The driving design goal of MDC1200 support is forwarding EAS audio over an existing Motorola-style two-way radio system. With MDC1200 signals enabled and the audio output wired into a base-station radio's auxiliary input (PTT keying handled by an external GPIO / VOX / COR loop), each broadcast goes on-air as:

[radio PTT keys up by GPIO/VOX]
  Pre  MDC1200 PTT-ID-Pre   ← "from unit 1234"
  1 s silence
  SAME header burst × 3     ← decoded by ENDECs and SAME-aware radios
  Attention tone (8 s)
  Voice narration
  SAME EOM × 3
  1 s silence
  Post MDC1200 PTT-ID-Post  ← "1234 clear"
[PTT releases]

Receiving subscribers display the calling unit ID, optionally selectively unmute for that ID, log the call in their dispatch console, trigger any configured alarms, and cleanly close the call on the post-ID. Common fleets where this fits naturally: county EM dispatch repeaters, volunteer fire / EMS / Skywarn / RACES nets, school and utility radio systems, industrial site-wide LMR. The full technical reference, including frame format, FEC, and op-code table, lives in docs/reference/protocols/MDC1200.md.

PTT keying is not part of this feature. EAS Station™ produces audio. Putting the radio into transmit is the responsibility of the radio interface (GPIO output, COR loop, VOX, VOIP gateway, …) and must be configured separately, keyed off the same trigger that starts EAS Station audio playback.

Ordering Inside the Composite Broadcast

[ pre-alert signal ]   ← system-level signal, configured in EAS settings
   1 s silence
[ SAME header burst ] × 3
   ...                ← attention tone (1050 Hz), narration, end-of-message
[ EOM burst ] × 3
   1 s silence
[ post-alert signal ]  ← system-level signal, configured in EAS settings

For manual broadcasts the existing per-broadcast pre-alert audio / post-alert audio uploads continue to bracket the narration segment (between the attention tone and the EOM). The system-level signals always sit outside the SAME signalling, so the two features are complementary:

[ pre-alert SIGNAL ]
[ SAME header ]
[ attention tone ]
[ pre-alert AUDIO upload ]    ← per-broadcast, optional
[ TTS narration ]
[ post-alert AUDIO upload ]   ← per-broadcast, optional
[ EOM ]
[ post-alert SIGNAL ]

Programmatic / Environment Overrides

For headless deployments, the following environment variables override the database values when set:

Internally the signal is rendered by app_utils.eas._generate_chime(); profiles are listed in app_utils.eas.ALERT_CHIME_PROFILES. Adding new profiles is a matter of extending that helper and the validation set in webapp/admin/maintenance.py.

REST API

The signal settings are exposed via the standard EAS settings endpoint:

GET /admin/eas_settings
PUT /admin/eas_settings    Content-Type: application/json

PUT payload fields (all optional):

{
  "pre_alert_chime": "bell",
  "post_alert_chime": "none",
  "pre_alert_chime_duration": 1.5,
  "post_alert_chime_duration": 2.0,
  "qc2_tone_a_freq": 1387.5,
  "qc2_tone_b_freq": 1530.0,
  "qc2_long_tone_enabled": true,
  "qc2_long_tone_seconds": 10.0,
  "dtmf_sequence": "*911#",
  "mdc1200_unit_id": "0x04D2",
  "mdc1200_op_code": "ptt_id_pre",
  "mdc1200_op_code_raw": null,
  "mdc1200_arg_raw": null,
  "mdc1200_target_unit_id": null
}

Server-side validation:

• *_chime must be one of none, bell, beep, three_tone, qc2, dtmf, mdc1200 (case-insensitive). Unknown values are silently rejected.
• *_chime_duration must be a number in [0.1, 10.0].
• qc2_tone_*_freq must be a number in [50.0, 4000.0] Hz.
• qc2_long_tone_enabled must be a boolean.
• qc2_long_tone_seconds must be a number in [1.0, 120.0].
• dtmf_sequence is filtered to 0-9 A-D * # and truncated to 32 characters.
• mdc1200_unit_id must be an integer in [1, 65535]. Strings are parsed with int(value, 0), so "1234" and "0x04D2" are both accepted.
• mdc1200_op_code must be one of ptt_id_pre, ptt_id_post, emergency, request_to_talk, remote_monitor, call_alert, selective_call, custom (case-insensitive).
• mdc1200_op_code_raw and mdc1200_arg_raw must each be an integer in [0, 255] (or null/"" to clear). Strings are parsed with int(value, 0), so both decimal and 0x.. hex are accepted.
• mdc1200_target_unit_id must be an integer in [1, 65535], or null/""/0 to clear (which forces single-packet emission). Strings are parsed with int(value, 0), so both decimal and 0x.. hex are accepted. Used only when mdc1200_op_code is call_alert or selective_call; ignored for other presets.

Database Schema

Schema additions are applied by Alembic migrations 20260501_add_alert_chime_to_eas_settings, 20260505_add_qc2_long_tone_to_eas_settings, 20260505_add_mdc1200_to_eas_settings, and 20260505_add_mdc1200_target_unit_id_to_eas_settings:

The migration uses ADD COLUMN IF NOT EXISTS and the application also self-heals these columns on first read via _PENDING_MIGRATIONS in webapp/admin/maintenance.py, so existing deployments will not require a manual schema fix-up.