Rev-Logger Quick Start Guide

This page explains what to launch first, which files you need, how the speech-to-text path works, and how to use the Rev-Logger interface during an event.

Start here Launch the FastAPI backend first.

The browser page sends short WAV chunks to http://127.0.0.1:8000/transcribe. If the backend is not already running, STT buttons in the page will not produce transcript text.

1) Files you need

Backend

server_patched_stable.py — local FastAPI transcription server
requirements_patched.txt — Python package list

Frontend

RevLog_Alpha_02_2026-03-22.html — Rev-Logger page with stable STT integration (note - there may be more recent editions, but this is the minimum version for this ‘help me’ file

2) What each backend file does

File	Purpose	Notes
`server_patched_stable.py`	Hosts the local API endpoints, loads `small.en`, receives uploaded audio chunks, transcribes them, and returns text as JSON.	Run this with Uvicorn before opening Rev-Logger.
`requirements_patched.txt`	Installs FastAPI, Uvicorn, multipart upload support, and faster-whisper.	You install this once per Python environment.

3) One-time setup

Install Python 3.10 or 3.11
If you already have a working Python for your Whisper test harness, you can keep using it.

Create a project folder
Put these files together in one folder:

server_patched_stable.py requirements_patched.txt RevLog_Alpha_01_ui_panels_synced_v16_stt_fixed.html

Create a virtual environment

python3 -m venv .venv

Activate it

source .venv/bin/activate

On Windows Command Prompt: .venv\Scripts\activate
On Windows PowerShell: .venv\Scripts\Activate.ps1

Install dependencies

pip install -r requirements_patched.txt

4) Start the backend each time you use Rev-Logger

Open a terminal in the folder and run:

uvicorn server_patched_stable:app --reload

You want to see output showing that Uvicorn is running on http://127.0.0.1:8000.

The first startup can take longer because faster-whisper may need to initialize the model.

5) Open the Rev-Logger page

Open RevLog_Alpha_02_2026-03-22.html in Chrome.

Allow microphone access when prompted.

Use the STT controls in the transcript area to start and stop transcription.

Watch the level meter. Aim for strong but not clipping audio.

6) Normal operating sequence

Load the page and confirm the interface is responsive.

Start STT in the transcript panel.

Watch the transcript window. Incoming speech should appear there automatically.

Select useful text from the transcript and move it into your logging fields using your existing Rev-Logger workflow.

Complete message traffic / callsign fields and log the entry.

Stop STT when you are done.

7) Test mode and client log

The client log is intentionally hidden during normal use.

Enable the Test checkbox to expose the Client Log button.
Open that modal only when you need troubleshooting details.
Use it to verify chunk size, upload timing, MIME/format behavior, and returned transcript text.

8) Recommended audio settings

Why

Stable decode path
Good latency for near-real-time operation
Avoids fragile MediaRecorder container fragments

9) Practical level guidance

Normal speech should usually live around -18 to -10 dBFS.
Brief peaks near -6 dBFS are fine.
If the meter frequently shows HOT or CLIP, reduce the input level.
For radio audio, cleaner and slightly conservative is usually better than loud.

10) Troubleshooting

Problem	Likely cause	What to check
No transcript appears	Backend not running, mic not allowed, or wrong file open	Confirm Uvicorn is running; reload the page; re-allow mic access
Buttons do nothing	Wrong or older HTML file	Ansure the latest version is in use (Alpha_02_2026-03-22 or newer)
Server returns errors	Dependency issue or backend crash	Read the terminal running Uvicorn for the traceback
Transcript quality is poor	Input too hot, noisy room, radio audio distorted	Use the level meter; reduce drive; keep audio clean

11) Fast startup checklist

1. Open terminal 2. cd into the Rev-Logger folder 3. source .venv/bin/activate 4. uvicorn server_patched_stable:app --reload 5. Open RevLog_Alpha_01_ui_panels_synced_v16_stt_fixed.html in Chrome 6. Allow mic access 7. Click Start STT