YOUR LIBRARY. FINALLY ORGANIZED.

You need a USB-B cable for file transfer (USB Storage mode), and — for MIDI scanning, prelisten, and live category writes — a class-compliant USB-MIDI interface with DIN connectors. Wire interface OUT → GT IN, GT OUT → interface IN. Both directions are required.

About the Roland USB driver: Roland's official V-Synth GT USB driver shipped in the PowerPC / early-Intel macOS era and doesn't work on modern macOS. Rogue Librarian is built around the class-compliant DIN-MIDI route — that's the configuration to use.

The indicator reflects the MIDI connection, not USB Storage. The two transports do different jobs:

• USB Storage — moves files. Mount the GT as a drive on your Mac; the app reads and writes projects from it.
• MIDI (via class-compliant DIN interface) — live scanning, prelisten, writing categories back. This is what the indicator tracks.

For MIDI to work, make sure:

• The GT is not in USB Storage mode — it needs to be in normal play mode.
• Both directions are wired: interface OUT → GT IN, GT OUT → interface IN.
• On the GT, press System and go to the MIDI screen. Set: Device ID → 17, Receive Messages → System Ex → ON, Transmit Messages → Edit Data → ON.

That last one — Edit Data: ON — is the most common cause of silent scan failures. See the next entry.

Edit Data must be ON on the GT. It's a prerequisite — without it, the GT silently drops every SysEx and the app cannot detect this from the host side.

Press SYSTEM → MIDI on the GT and confirm Edit Data is highlighted:

V-Synth GT · System → MIDI · Edit Data ON

Not over USB — the GT does USB MIDI or USB Storage, not both. Mounting it as a drive disconnects USB MIDI; this is hardware behaviour, not a bug.

DIN MIDI is independent. With a class-compliant USB-MIDI DIN interface plugged into the GT's MIDI ports, MIDI stays alive while USB is in Storage mode. That's the setup if you want both transports working simultaneously.

No — Rogue Librarian is a library manager and project builder, not a patch editor. You can't design patches from within the app.

What it does instead:

• Organises and manages patches you've already created on the GT
• Lets you build custom setlists and projects from your existing sounds
• Gives you visibility into what's inside each patch — which tones, which wave files, which oscillators

So the workflow is: design patches on the GT as normal → import them into Rogue Librarian → organise, rate, tag, build projects → export back to the GT.

Wavelabs (new in v1.2) do let you import your own .wav samples and bundle them into a wave-only project for the GT — see the next entry.

A My Project is a curated patch collection (up to 512 patches with their tones and waves) — what most people think of as a GT project.

A Wavelab is wave-only: no patches, no tones, just a fresh wave bank assembled from your own .wav samples and any waves from your imported GT projects. The GT loads either as a regular .prj file via Storage mode.

Read more about wavelabs ↑

No — the export worked. The new project is on the GT's internal flash, but the GT's playback area (the temporary area) is still holding whatever was loaded before.

Power-cycle the GT. That is the only operation that loads the externally-written flash bytes into the temporary area. Load Project from the GT's menu does not do this for projects written by Rogue Librarian — only a power cycle does. The export wizard reminds you of this on the Finish step.

After the power cycle, your new project is live.

Save Project on the GT writes the wrong direction: temporary area → flash. After a Rogue Librarian export, your fresh project is in flash but the temp area still holds the old state. Hitting Save Project would overwrite your fresh project with stale data and silently undo the export.

The same trap applies after importing: don't touch Save Project until you've power-cycled the GT to load the new flash contents into the temp area.

The clean rule: power-cycle is the only commit operation after a Rogue Librarian write.

No. Rogue Librarian's wave pool is content-addressed by SHA256 hash — identical waves are stored once, regardless of how many projects reference them. Same for user-imported .wav files: re-importing the same audio is a no-op.

The Import wizard (new in v1.2) also detects duplicate projects by content hash before importing and warns you, so you can re-import deliberately or skip.

It analyses each sample's peak amplitude on import and computes a non-destructive gain so the peak hits −1 dBFS. Your original .wav file is never modified — gain is stored as metadata and applied at preview and export time. You can edit the gain later in the wave inspector at ±24 dB.

Hold the EXIT button while powering the GT on. This boots the synth with a clean, empty project — useful when you want a fast startup (no wave loading), or a blank slate to import just a handful of patches into.

Your actual default project on the GT isn't touched. Power-cycle normally next time and it comes back as it was. Once you're booted empty, you can go to MODE → Disk → Import Files and pull in individual patches, rather than loading an entire project.

Note: this is an owner-community tip, not something documented in Roland's official manuals, so behaviour may vary slightly by OS version. It works reliably on V-Synth GT OS 2.x.

Two tutorial videos are currently in production:

• The essential workflow — import, build, export
• Connection setup — USB Storage mode vs MIDI, including GT-side settings and interface wiring

They'll be published and linked from this page as soon as they're up. In the meantime, the in-app help bundle and the FAQ above cover the most common setup questions.