Suno Studio tutorial: a first session from import to export

Most Suno Studio tutorials start by touring the interface. That is the slowest possible way in. A workstation only makes sense once you have pushed material through it, so this walkthrough is built around one finished job: take a generated track, break it apart, replace a part with something you played, and export stems and MIDI. Everything else in Studio is a variation on that loop.

You need Premier. Studio is not on Pro, and there is no separate purchase.

Before you open Studio: bring material, not ambition

The single biggest reason a first session dies is starting from an empty project. An empty project asks you to be a composer, an arranger and a mixing engineer at once, in an unfamiliar tool.

Bring one of three things instead:

• a track from your library that is 80 percent right and annoying in one specific way
• a stem set you already have
• a loop, a hum, a phone recording, anything six seconds long that you made

Notice what all three have in common. They give the session a problem to solve. "Make something good" is not a problem, it is a mood.

If the material is coming from a generation, spend the effort upstream first. Studio rearranges what the model gave you, it does not rewrite it, so a sharper Style string changes more than an hour of knob work. Pull a concrete style recipe from the catalogue if your prompt is running on adjectives.

Step 1: get the track into a session

Studio takes imports: audio files, stems you separated elsewhere, or a track you already made in Suno. You can also start with nothing and build from the chat bar, which is fun and, for a first session, a trap.

Set the project BPM early. Everything downstream, quantised playing, loop lengths, automation that lines up with bars, depends on the session agreeing with the material about tempo. Getting this wrong at the start is how people conclude that "the timeline feels off" three hours later.

Step 2: split it into stems, then listen with the vocal muted

Studio does advanced stem separation inside the session: vocals, drums, bass and the rest, split with more resolution than the two-way split available on Pro.

Then do the thing almost nobody does. Mute the vocal and listen to the backing alone.

A generated mix hides a lot behind the voice. Muted, you hear what the model actually built: whether the bass has a shape or just a note per bar, whether the drums change between sections or loop flat for three minutes, whether the "layered strings" in your prompt turned into one sustained pad. This one minute of listening tells you what to fix, and it usually contradicts the thing you thought was wrong.

Keep the stems. Even if the session goes nowhere, exported stems are the reusable asset.

Step 3: replace exactly one part

Not three. One.

You have two ways to play a new part. Plug in a MIDI controller and use the built-in synths. Or use musical typing, which turns your laptop keyboard into an instrument, with an arpeggiator and chord mode when your fingers are not up to the job. Chord mode is not cheating. It is how a producer with no keyboard skill gets a usable progression down before the idea evaporates.

Pick the part that annoyed you in the muted listen. Usually it is the bass, because generated basslines tend to follow the root and stop there. Play a line with actual movement under the existing drums, mute the original bass stem, and hear what a single human decision does to a track that sounded finished.

That is also the honest test of whether you need Studio at all. If replacing a part felt like relief, you are the target user. If it felt like homework, cancel Premier and put the money into prompting.

Step 4: mix, and let something move

Volume and pitch are per track. Effects include the standard workshop kit, sidechain compression and convolution reverb among them, and you can design an effect from scratch if the stock ones do not fit the vibe.

The move that separates a Studio mix from a generation is automation. Any parameter can be automated, either by drawing a curve or by recording the movement of a control while the track plays. Record mode is the faster route, because your hand hears the music and a drawn curve only knows about numbers.

Three automations worth the effort on a first pass:

1. A filter opening across the last chorus. Cheap, effective, instantly reads as intent.
2. Sidechain depth easing off in the bridge. The pump that drives the verses becomes exhausting if it never lets up.
3. Reverb send rising on the final vocal line. The song sounds like it is walking away from you.

None of these are new ideas. That is the point. Studio is where generative material gets treated like recorded material, and recorded material has had these tricks for forty years.

Step 5: export like you have a download budget, because you do

Studio exports full songs, stems, or a selected range. MIDI comes out too, which is the feature that makes Studio worth its price for anyone finishing in another workstation. Play the part in Studio, export the MIDI, rebuild it with your own instruments.

Watch the ceiling. Song downloads are capped monthly from 3 September 2026: 20 on Pro, 60 on Premier. Suno handles stem and multitrack downloads as a separate question in its Studio FAQ, so check the current wording before a heavy export week rather than discovering the rule at song 61.

Also check your credit position before a long session. Credits do not carry over from month to month, and once the monthly allowance is gone you drop to the 50-a-day allotment that free users get, which is not a rate you want to hit mid-project. The credit mechanics are worth knowing in advance.

A realistic first hour

• 5 minutes: import, set BPM
• 5 minutes: stem separation
• 5 minutes: listen with the vocal muted, write down the one thing that is wrong
• 20 minutes: replace that one part
• 15 minutes: three automations
• 10 minutes: export stems and MIDI

If that hour felt like production, you have found your tool. If it felt like fighting a browser, the honest answer is that a Pro subscription plus the workstation you already own covers most of this, and reviewers who produce for a living have said so repeatedly.

Where first sessions go wrong

Starting from a blank project. Covered above, and worth repeating because it is the number one killer.

Fixing everything at once. Four replaced parts and you no longer know which change helped.

Treating the chat bar as a producer. It builds sounds and handles busywork well. It does not have taste, and it does not know what your track needs.

Uploading material you do not own. Suno's matching system rejects uploads it recognises, and the error reads audio matches an existing recording. Working around a rights check is not a workflow tip.

Expecting a better song. Studio changes arrangement and mix. Writing happens in the prompt, before any of this.

FAQ

Do I need a MIDI controller? No. Musical typing plus chord mode covers a first session. A controller makes it faster, not possible.

Can I open old Studio projects in Studio 2.0? Suno answers this directly in the Studio FAQ on its site, and that is the source to trust as versions move.

Are there official tutorials? Yes, Suno publishes its own: an introduction to Studio 2.0 running about 21 minutes, a getting-started session near 38, plus shorter ones on MIDI, editing and mixing, instruments and effects, recording, and building plugins.

Can I mix a track that was not made in Suno? Yes, import it. The rights disclaimer applies.

Is Studio enough to release a track? For a demo, easily. For a commercial release, most people still finish and master elsewhere, and the export options exist because Suno expects exactly that.