GP-SDR

General Purpose Software Designed Radio is an app-first, receive-only SDR workstation for scanning, tuning, logging, P25 trunk following, wideband channel mixing, and later review. The native macOS app contains the complete interface and local receiver service. The same interface can also run headlessly in a web browser and is designed for phones and tablets.

GP-SDR 1.1 is a public release. On-air performance still depends on the receiver, antenna, local signals, USB link, and gain settings.

What is included

• Native macOS app for Apple Silicon, Intel, and Universal Macs
• Windows 10/11 package and Debian/Ubuntu packages for amd64 and arm64
• Bundled SDRTrunk v0.6.1 P25 Phase 1/2 trunk-following stack and JMBE Creator
• Native HackRF and RTL-SDR discovery and live IQ input
• Concurrent receivers with control, voice, survey, tuner, and channel-bank roles
• Built-in AM, narrow FM, and broadcast FM DSP, squelch, recording, and live audio
• Real tuner spectrum and scrolling waterfall with peak hold, channel markers, cursor frequency/power, click-to-tune, recent frequencies, and estimated SNR
• Separate hardware center and software Listen VFO controls; GP-SDR offsets a new VFO away from receiver DC by default and can move it inside a locked HackRF/RTL-SDR passband without restarting the device
• Unified channel controls plus per-talkgroup mute, solo, identity, and activity
• Unattended activity logging, signal grouping, recordings, and later review
• Indexed Timeline search across transcripts, callsigns, protocols, labels, systems, and frequencies; configurable recording/IQ retention
• Independent Mapper jobs per receiver: split ranges across SDRs, run parallel Discovery sweeps, or run Discovery and Identify simultaneously. Each job has live frequency/progress/ETA, 5-second-to-7-day per-channel listening, and receiver/job provenance in the combined result table and CSV export
• Peak activity hours, expandable identification evidence, optional location/transcription, filters, saved job controls, and Sheets upload
• Dedicated pages for Analog, P25, DSD-FME, rtl_433, dump1090, multimon-ng, acarsdec, and AIS-catcher
• Normalized live decoder bridges for digital voice, sensors, ADS-B/Mode S, paging/signaling, ACARS, and AIS; analog CTCSS detection
• Built-in decoder scan profiles for conventional digital voice, common 315/345/433.92/868/915 MHz sensors, 1090 MHz ADS-B/Mode S, paging/signaling, North American ACARS, and both marine AIS channels
• Built-in GMRS, NOAA Weather Radio, MURS, CB, broadcast FM/AM, civil air, marine VHF, 2 m, 70 cm, public-safety discovery, and custom-range profiles
• Bundled San Ramon/East Bay conventional and P25 profiles, all 84 California GMRS repeaters from the local archive, and sanitized handheld/travel banks
• Profile export plus bulk CHIRP CSV/TSV import; select multiple programming files and turn every file into a complete channel bank in one step
• Automatic Google Sheets range sync with read-only built-in profiles, manual refresh, scheduled updates, validation, and an offline cache
• RadioReference ZIP/location import with 5/10/25/50/100-mile and custom 1–100-mile ranges
• Optional local whisper.cpp transcription
• Included Universal macOS SoapySDR bridge for other installed SDRs and remote sources
• Authenticated, responsive web console for headless and mobile use

Encrypted P25 calls are identified and skipped. GP-SDR does not transmit and does not attempt to defeat encryption.

Mapper keeps RF evidence separate from protocol proof. Activity on a known decoder target is labeled as a candidate and records the matching decoder and its installation state; a specific protocol is not claimed merely because a frequency falls inside a known band.

Install and run

For the complete first-run checklist, receiver-driver steps, P25 setup, LAN access, and troubleshooting, see the installation guide. The GP-SDR Wiki provides a detailed guide to every main page, receiver workflow, decoder, integration, and setup option. The 1.1 feature status separates live-hardware evidence from implemented features and external requirements.

macOS

Download the Universal DMG for the easiest choice, drag GP-SDR.app to Applications, and open it. Separate arm64 and x86_64 ZIP packages are also available. These preview bundles are ad-hoc signed rather than Apple-notarized, so macOS may require Control-click → Open the first time.

While the native app is open, GP-SDR prevents idle system sleep so unattended scans and recordings continue. Display sleep remains available, and the sleep assertion is released when the app quits.

The complete P25 trunking engine is already inside the app. The P25 page can create the JMBE voice codec locally on first use. For the built-in analog tuner and scanner, open Hardware and use each component's Install or How to button. On macOS, HackRF and RTL-SDR use user-space host tools rather than a kernel extension.

Windows 10/11

Extract the ZIP and run GP-SDR.exe -open. The included P25 engine can open HackRF and RTL-SDR devices. Windows may still need the correct WinUSB device driver; the Hardware page explains the device-specific setup. Other vendor drivers must come from their vendor.

Debian or Ubuntu

sudo apt install ./gp-sdr_1.1.1_amd64.deb
sudo systemctl enable --now gp-sdr

Use the arm64 package on 64-bit ARM. Open http://127.0.0.1:8073/ locally.

Test without a radio

gp-sdr -demo -open

Demo activity is clearly labeled and never presented as received RF.

App workflow

1. Connect one or more SDRs.
2. Open Hardware, press Refresh, and follow any shown setup action.
3. Use Tuner for direct listening and the real spectrum/waterfall.
4. Choose a built-in profile or create/import one under Profiles.
5. Assign receiver roles. For P25, use one wideband radio or separate control and voice radios. For GMRS or similar banks, use the channel-bank role.
6. Press Start on Live.
7. Control every active channel or P25 talkgroup from the mixer; inspect saved transmissions under Activity.

Bundled regional data, public sources, local-file sanitization, and accepted bulk-import columns are documented in Docs/CHANNEL_DATA.md. Google Sheets setup and accepted range columns are documented in Docs/GOOGLE_SHEETS_SYNC.md. Mapper CSV exports and reviewed writes to a master Additions Queue are documented in Docs/MAPPER_SHEET_WRITE.md.

GP-SDR chooses the standard user configuration directory. -data /path selects another location.

P25 setup

The packaged app contains SDRTrunk; no separate trunking app is required. Create or import a profile containing the system's control channels and optional talkgroup labels, then assign available receivers. A single wideband receiver can host the control and fitting traffic channels. Multiple radios increase the number of simultaneous traffic channels SDRTrunk can follow.

The example two-receiver P25 profile contains placeholder values, not a real radio system. RadioReference import can build location-specific conventional and P25 profiles when authorized account credentials are available.

Optional components

The Hardware and Settings pages keep setup inside the app wherever an automatic installation is safe. Where platform security or vendor licensing prevents it, the How to action gives exact steps instead.

On macOS, a decoder card's Install action can run the bundled, revision-pinned decoder-suite installer for DSD-FME, rtl_433, dump1090, multimon-ng, acarsdec, and AIS-catcher. Receiver host tools and offline transcription also have in-app Homebrew actions. Windows driver replacement and Linux system packages continue to use explicit How to steps because those require administrator control.

For source builds on macOS (the release app already includes the stream bridge):

brew install hackrf librtlsdr soapysdr cmake
Scripts/build_optional_components.sh build/helpers/darwin-arm64

For offline transcription, use Install on the in-app Transcription card. GP-SDR installs whisper.cpp and downloads its checksum-pinned English base model without an API key. Source builds can still override either path with:

export GPSDR_WHISPER_EXECUTABLE=/path/to/whisper-cli
export GPSDR_WHISPER_MODEL=/path/to/ggml-small.en.bin

For authorized RadioReference import:

export GPSDR_RR_USERNAME='your account'
export GPSDR_RR_PASSWORD='your password'
export GPSDR_RR_APP_KEY='approved developer key'

Credentials are never included in shared profiles or event logs. A Premium subscription and developer API access are separate RadioReference requirements.

For an offline library without API access, open a RadioReference county or state Downloads page while signed in and save its official CSV files into a folder. In GP-SDR, open Settings → Local database → Choose folder. GP-SDR recursively imports .csv, .tsv, and GP-SDR .json files; large statewide CSVs are split into stable 4,000-channel banks so rescans update rather than duplicate them. RadioReference provides an official All Identified Frequencies in California CSV on the California Downloads page, but does not document a single whole-US archive. GP-SDR does not scrape or mirror the RadioReference database.

Headless and mobile use

Local-only mode is the default:

gp-sdr -listen 127.0.0.1 -port 8073

To use GP-SDR from a phone on a trusted LAN:

gp-sdr -listen 0.0.0.0 -port 8073

The service prints a URL with a random access token when listening beyond localhost. Keep that token private. For use outside a trusted LAN, place GP-SDR behind a VPN or authenticated HTTPS reverse proxy; the built-in service is HTTP.

Build from source

Requirements: Go 1.24 or newer. macOS desktop packaging additionally needs the Xcode command-line tools. Packaging downloads official SDRTrunk v0.6.1 and JMBE Creator v1.0.9 assets and includes their corresponding source archives as documented in THIRD_PARTY.md.

cd server
go test ./...
go build -o gp-sdr .
./gp-sdr -demo -open

Build all release packages on macOS:

chmod +x Scripts/build_release.sh Scripts/fetch_p25_stack.sh
Scripts/build_release.sh 1.1.1

Outputs are written to dist/ with SHA256SUMS.txt. The script creates macOS arm64, x86_64, and Universal apps; Linux amd64/arm64 DEBs; and a Windows amd64 ZIP. Public notarization requires an Apple Developer ID and notarization credentials, which are intentionally not stored in this repository.

Command-line options

-listen ADDRESS   Listen address (default 127.0.0.1)
-port PORT        Interface port (default 8073)
-data PATH        Profiles, recordings, and event data directory
-token TOKEN      Access token; generated automatically for LAN binds
-demo             Generate clearly marked simulated activity
-open             Open the interface in the default browser

Verification scope

Automated tests cover profile validation, API authorization, tuner input, spectrum generation, SDRTrunk playlist generation, concurrent Mapper jobs, indexed event search, storage boundaries, optional-decoder handoffs, and the web application. A live HackRF acceptance capture through GP-SDR decoded EBRCS NAC 0x1F5, WACN 0xBEE00, system 0x1F1, Phase 1/2 grants, traffic-channel frequencies, and encrypted-call state with IMBE/AMBE loaded. Live RTL-SDR P25, other systems, and other packaged operating systems remain separate hardware acceptance checks; a passing build alone is not presented as RF proof.

See Architecture, release notes, and third-party credits.

Responsible use and license

Receive, record, transcribe, and share only communications you are legally permitted to handle in your jurisdiction. GP-SDR's original code is released under the MIT License. Bundled and optional components retain their own licenses; all credits and redistribution details are recorded in THIRD_PARTY.md.