Orchestra modoki β Conductor

🎵
Song
📱
Devices
🎯
Tune
🎚️
Assign
▶️
Play

🎵 Song

Choose the song to perform. Sample songs work as-is.

📱 Devices

Scan this QR with a phone to join as a performer device.

Scanning this QR on a performer device (iPhone / Android / PC, etc.) opens the performance screen and auto-registers it to this conductor.

No devices connected yet

🎯 Tune

Automatically aligns timing differences between devices. Optional (skippable). Run it if alignment bothers you.

⚠️ On iPhone, mic calibration while a Bluetooth speaker is connected can make the connection unstable. Calibrating from a separate device is recommended.

🎚️ Assign

Auto-assigns connected devices to the song's parts. You can change each device's part below.

▶️ Play

🎵 —

—

🔊 Volume

🎼 To perform with your own songs, add them from the detailed screen.

⭐ Adding songs requires signing up for an account.

👋 Welcome to the conductor screen

A conductor's podium for distributing one song into CH parts across multiple performer devices (iPhone / Android / PC, etc.) and triggering them at millisecond precision.

What this screen does

Register performer devices and send play / stop as a single command
Split one song by CH (parts) and assign them to each device to ensemble
Upload your own songs directly per CH or auto-separate with Demucs
Match per-device output latency in milliseconds via calibration
Use scheduled play with "Now" / specific time

Quick start

① Each performer scans the "📱 Add performer device" QR on their phone
　 → the performer screen opens and they're registered with this conductor automatically
　 * For devices that can't read QR, send the URL via "📋 Copy invite URL"
② Pick a song in "🎵 Song selection"
　・Pick a sample track (♪◯) as-is
　・To use your own song, go to "Add song":
　　・Direct upload per CH: assign mp3 / m4a / wma already split into parts to each CH
　　・Auto-separate: upload one mixed song → Demucs auto-splits into Vocals / Drums / Bass / Other, etc.
③ Assign CH (part) to each device in "🔗 Connected devices"
　・Grab the ☰ to reorder rows (also controls the order for real-device volume matching, etc.)
　・Optionally set the physical layout (left/right, front/back) of each device by drag in "🗺️ Layout"
④ Run two-stage correction in "🎚 Calibration"
　・"🔊 Real-device volume matching": pink noise plays per device in turn — each performer matches loudness with the device volume buttons
　・"🎤 Auto-calibration": the conductor's mic auto-measures each device's output timing → in the result modal use "Apply correction to all"
　 * Fine-tune by ear with each row's ±10ms / ±100ms buttons
⑤ Set the start time in "🎻 Play" and trigger all devices simultaneously with "▶ Play for everyone"

Playing music on multiple Teslas

This is the overall flow from the conductor side for playing music on multiple Teslas at once.

⚠ The conductor must operate from a separate device that is not also used as a performer.
During the 12-second recording of auto calibration the microphone is occupied, so using that device as a performer can drop the boombox output or get it stuck in call mode. Always reserve a separate device for the conductor (Android, another iPhone / PC, etc.) that does not take part in the performance.

Preparation (each Tesla / each performer)
・Put each Tesla in P (Park) (the boombox only works while parked)
・Each performer connects their phone to the Tesla via Bluetooth and switches media output to the Tesla
・On each performer phone, scan the QR in "Add a performer device" above to open the performance screen
・Follow step ⓪ on the performance screen to launch Toybox → Boombox on the Tesla and turn external speaker output ON

Conductor steps
① Register each performer with "Add a performer device"
　・Recommended: have performers scan the QR on the left (auto-register)
　・Or: read the QR at the top-right of the performance screen one by one with "QR scan to add" on the right
② Choose a track and adjust volume in "Music"
③ Assign a CH (part) per Tesla on each row of "Connected devices"
④ Auto-measure the output delay of all cars with "🎤 Auto calibration" in "Calibration" → "Apply to everyone" in the result modal
　 ※ For fine-tuning, use the "±10ms / ±100ms" buttons on each device row
⑤ In "Play", choose "Now" or a start time and press "▶ Play all"
⑥ Playback starts on each phone → Bluetooth → Tesla audio → external speakers

How sync works
Each performer device syncs with server time via NTP and starts playback together with millisecond-precision scheduling via AudioContext.

Notes
・A car that starts moving (P released) automatically stops boombox output
・Because it goes through Bluetooth, watch both the Tesla media volume and each phone volume
・Devices on unstable Wi-Fi/mobile connections drift more easily, so choose devices with a good connection
・⚠ Especially, do not share the Boombox iPhone and the conductor device (running "Auto calibration" on the same iPhone switches the iOS Bluetooth profile to call mode HFP due to microphone use, and it may not return to A2DP even after Bluetooth OFF/ON or restarting Safari; as a result "Play current media" cannot be pressed on the Tesla. Use an Android or a second iPhone for the conductor)

🎵 Song selection

🎼 Add song

Log in or sign up (free) to upload and use your own songs. After logging in you will return to this screen to add them. You can keep using the sample songs as a guest.

📱 Add a performer device

📱 Have performers scan this QR Recommended

Scanning this QR on a performer device (iPhone / Android / PC, etc.) opens the performance screen and auto-registers it to this conductor.

Also open the performance screen on this device ↗

🔍 Scan a performer QR

If the QR is visible but not recognized, move the phone back a little to focus. If it is dark, turn the light ON.

Reads the QR at the top-right of the calibration box on the performance screen.

🔗 Connected devices

🎚 Calibration

Reset all device corrections to 0 before the next measurement (one time only; auto-off after running)

How to calibrate

Calibration is done in three stages from top to bottom: "① Position check → ② Volume matching → ③ Auto calibration". Going in this order before playing lets multiple devices form an ensemble.

① 📍 Device position check (check where each device is placed)
Pressing "📍 Device position check" plays a sine wave of a different pitch from each device in registration order (reorderable with ☰). You can listen to confirm whether the stereo left/right and front/back placement is as intended. It loops, so you can run it as many times as you like until you stop it.

② 🔊 Device volume matching (match each device volume)
Pressing "🔊 Device volume matching" plays the same pink noise from each device in turn, in registration order. Each performer should adjust the loudness with the device physical volume (phone/PC hardware buttons or OS volume) when it is their turn. It loops, so run it as many times as needed.
・↻ button: resets every device in-app volume slider to 5 (preparation before starting volume matching).

③ 🎤 Auto calibration (align the timing of the sound)
Using this device (the conductor phone) microphone, it records your reference tone (G4 = 392Hz) and each performer sound (C5–G7, a spread C major chord) at the same time and automatically measures the offset in ms. After measuring, the Apply correction button in the modal adds the delta to each performer output delay.
① Each performer keeps the volume matched in ① in a quiet environment.
② Pressing "🎤 Auto calibration" starts about a 12-second measurement; when it finishes, the offsets appear in a modal.
③ Apply per row with "Apply correction", or to everyone with "Apply to everyone" at the bottom.
④ Run it again and confirm delta ≈ 0 (within ±20–30ms) to finish.

④ Fine tuning (by ear)
For devices that auto calibration cannot fully align, use the ±10ms / ±100ms buttons on each row of "🔗 Connected devices" to fine-tune by ear. Pressing "Start calibration" plays a click on that device so you can compare it with the others.

↻ button / reset before measurement (when you want to redo corrections)
Repeated corrections can leave past values stacked on top, causing the offset to accumulate. The ↻ button of auto calibration resets every device correction to 0. If you check reset before measurement, the next run only will reset all device corrections to 0 before measuring. The checkbox clears automatically after running, so running auto calibration again lets you confirm the corrected state by ear (check it again if you want to start over).

🗺️ Layout

Grab a device icon and place it on the canvas. Lift them up from the holding row below. The placement is remembered on this device.

CH to device

↑ Lift up from here to place

🎻 Play --:--:--

■ All stopped

0:000:00

Bulk timer control

① Choose "Now" or a start time from the slots above.
② "▶ Play all" sends to every performer device at once. With "Now", playback starts together 2 seconds after each device finishes downloading the track (orange = scheduled); music, volume and each device CH assignment are delivered together.
③ "■ Stop all" cancels both playing and scheduled playback.
④ "⏸ Pause all" pauses while keeping the current position.
⑤ "▶▶ Resume all" resumes together from the paused position with a 2.5-second lead time.
Each device starts simultaneously with millisecond precision based on its own AudioContext and NTP sync.

📋 Status log