What Are USTs and How Do They Work in UTAU?

A UST (UTAU Sequence Text) is the sheet music of UTAU: a project file that tells a voicebank which samples to sing, at what pitch, and with what timing and expression. If you want a UTAU singer to perform a song, you load or build a UST, point it at a voicebank, and render the vocals. This explainer covers what a UST contains, how it drives a voicebank, and what to check when a UST doesn't sound right.

The core idea: UST as sheet music

UTAU works like a small recording session. You prepare the "studio" (the UTAU program), you prepare the "sheet music" (the UST), and you bring in the "singer" (a voicebank). The voicebank supplies the raw audio; the UST supplies the instructions for how that audio should be arranged and shaped into a performance.

That division matters when troubleshooting. A wrong note or bad timing is usually a UST problem. A muffled, off-pitch, or missing sound is usually a voicebank or oto.ini problem. Knowing which side of the session you're editing saves a lot of guessing.

What a UST actually contains

According to utauguides.carrd.co, a UST holds information such as:

• BPM — beats per minute, the tempo of the song
• Notes — the pitches and their positions on the timeline
• Aliases — the sample names the UST uses to "command" the singer
• Pitchbends — how the pitch glides between or within notes
• Vibrato — the periodic pitch movement applied to sustained notes

The guide notes that USTs can carry other properties too, but these are the ones that matter most for getting a usable performance.

BPM sets the grid everything else sits on

BPM (beats per minute) represents the tempo. It defines how long each beat lasts, so every note length, pitchbend, and vibrato is measured against it. If you import a UST and the whole thing feels too fast or too slow, check the BPM before you start nudging individual notes — a single tempo value can shift the entire arrangement.

Aliases are the bridge to the voicebank

An alias is a name attached to a cut audio sample. When the UST places an alias on a note, it's effectively telling the singer "summon this sample here." This is why a UST can sound perfect with one voicebank and broken with another: if the second voicebank doesn't have a sample registered under that alias, the note has nothing to sing.

Pitchbends and vibrato shape the performance

Notes alone give you a robotic, stepped melody. Pitchbends let the pitch slide into, out of, or between notes, and vibrato adds the wavering sustain that makes held notes sound sung rather than sampled. These are the parameters you'll spend most of your editing time on once the notes themselves are correct.

How a UST and voicebank combine at render time

The workflow, in order:

1. Prepare the studio — open the UTAU program.
2. Load the sheet music — open or create the UST.
3. Bring in the singer — select a voicebank so its samples are available to the UST's aliases.
4. Refine — adjust notes, pitchbends, and vibrato in the UST; adjust the voicebank's configuration if samples are misaligned.
5. Render — tell the program to "record," and it renders the singing of the entire UST into audio you can mix into a track.

The expected result of step 5 is a vocal audio file covering the whole sequence. If you get silence, missing syllables, or a track that's wildly out of time, the cause is almost always in steps 2–4, not the render itself.

Common problems and where to look

That last row is worth understanding: frequency files (.frq) store pitch information from the original recording and help UTAU "flatten" the pitch so the singer lands on tune. A voicebank with poor or missing .frq data will struggle to sing accurately no matter how clean your UST is.

What to do next

If you're new to UTAU, start by loading an existing UST with a voicebank you've already installed, render it once, and listen. That single pass shows you which of the two halves — sheet music or singer — needs your attention. From there, edit one parameter at a time (BPM first, then aliases, then pitchbends and vibrato) so you can hear exactly what each change does.

For a structured path, utauguides.carrd.co splits its material by skill level — Overview for concepts, Basic for using UTAU, and Advanced for improvement — which is a reasonable order to follow if you're building up from "what is a voicebank" to full cover production.