Practical guides for
AI music creators

Why Suno Tracks Often Need Mastering

Suno generates music using neural networks trained on a wide range of recordings. The output quality has improved significantly, but there is a consistent pattern: many Suno tracks are generated at relatively high loudness levels and may already contain significant dynamic processing. Spotify commonly normalizes playback around -14 LUFS, although playback behavior may vary depending on platform settings and listening environment — so when the platform adjusts your track, it can sometimes expose compression artifacts that were not noticeable at full volume.

Mastering does not fix a bad recording. What it can do is bring the track's integrated loudness to an appropriate target, reduce true peaks to prevent digital clipping in the streaming signal chain, and apply light EQ corrections if the frequency balance is off. The result is not a different song — it is a more consistent version of the same song, prepared to sit well alongside professional releases.

What WaveStudio Does (and Doesn't Do)

WaveStudio processes your track through a multi-stage signal chain:

• —Analysis — measures integrated loudness (LUFS-I), true peak level, loudness range (LRA), and stereo width before applying any processing.
• —EQ — a gentle broadband correction is applied if certain frequency ranges appear imbalanced. The correction is intentionally light; it aims to reduce harshness or muddiness without changing the overall character of the mix.
• —Dynamic processing — a transient-aware limiter brings the track to the target loudness while keeping a ceiling on true peaks. The processing is designed to focus primarily on controlling peak levels while preserving musical dynamics where possible.
• —Output — the processed audio is written to 24-bit WAV (or 320 kbps MP3 if selected) at the target level.

If the input track is already well-balanced and close to the target loudness, the processing will be minimal. The algorithm tries to do as little as necessary to reach the target — not as much as possible.

Step-by-Step: Mastering a Suno Track

• —Step 1 — Export from Suno. Download the MP3 from Suno's export button. WAV is not available from Suno directly.
• —Step 2 — Open WaveStudio. Go to wavedisco.com/studio. No account, no installation — runs entirely in your browser.
• —Step 3 — Upload your file. Click "Upload & Auto-Master" or drag your file into the drop zone. Accepts MP3, WAV, and FLAC.
• —Step 4 — Check the analysis. Before processing, the analyzer displays your track's current LUFS and peak level. A track at -10 LUFS will need more limiting than one at -14 LUFS — this reading helps set expectations.
• —Step 5 — Choose a loudness profile. Spotify / YouTube (-14 LUFS) is a safe default for most content. Apple Music (-16 LUFS) if that's your primary platform. Radio (-11 LUFS) for broadcast contexts. Club (-9.5 LUFS) for DJ use.
• —Step 6 — Process. Click the master button to begin. Processing time depends on file size, track length, and server load.
• —Step 7 — A/B compare. Switch between original and master using the A/B player. Listen for loudness consistency, high-frequency balance, and low-end clarity.
• —Step 8 — Download. Download as WAV for distribution platforms, or MP3 for direct sharing. For Spotify and Apple Music via DistroKid or similar, WAV is recommended.

A Note on Expectations

Mastering works best when the source material is already reasonable. If a Suno track has a specific problem — a very harsh snare, an overdriven bass, phasing artifacts — mastering will not remove it, and in some cases may make it slightly more apparent at a consistent loudness level. Results vary by track, genre, and starting loudness.

Why Loudness Matters More Than You Think

When you upload a track to Spotify, the platform does not play it at whatever volume it was exported at. It adjusts playback so that all songs in a playlist sound roughly consistent to the listener. This adjustment is based on a measurement called LUFS — and understanding it can help you make better decisions before distribution.

What LUFS Means

LUFS stands for Loudness Units relative to Full Scale. It is a measurement standard developed by the International Telecommunication Union (ITU) to reflect how humans perceive volume over time — not just the peak level of individual samples, but the integrated loudness of a track as a whole.

• —LUFS-I (Integrated) — average loudness measured over the entire duration of the track. This is the number streaming platforms primarily reference for normalization.
• —True Peak (TP) — the actual maximum signal level, including inter-sample peaks that can cause distortion in the digital-to-analog conversion chain. Most platforms recommend a true peak ceiling of -1 dBFS or lower.
• —LRA (Loudness Range) — the difference between quieter and louder sections, measured in LU. Higher LRA means more dynamic contrast; lower LRA means a more consistently loud track.

Streaming Platform Targets

Many major streaming platforms use loudness normalization, although implementation details and playback behavior may change over time.

*Playback behavior may vary depending on platform settings and listening environment.

What Happens If Your Track Is Too Loud

A track exported at -8 LUFS on Spotify may be turned down by roughly 6 dB depending on platform behavior and listener settings. That reduction is applied to the whole signal — but it does not undo the dynamic compression or limiting used to achieve that loudness in the first place. The result can be a track that sounds over-processed compared to less aggressively mastered content playing at the same perceived volume.

AI-generated music from Suno or Udio may come out at relatively high loudness levels. Bringing these tracks closer to the streaming target before upload may reduce the amount of normalization applied during playback — and the track retains more of its original character.

What Happens If Your Track Is Too Quiet

If your track is at -22 LUFS and the platform targets -14 LUFS, it may be turned up slightly — though upward normalization behavior varies by platform and is sometimes capped or not applied at all. A track that sits close to the platform's target tends to behave most predictably in a playlist context.

How to Check Your Track's LUFS

WaveStudio displays integrated loudness and true peak before any processing begins. Upload your file to wavedisco.com/studio and the analyzer shows your current LUFS-I reading in the meter area. No processing is required just to take a measurement.

Choosing the Right Target

• —-14 LUFS is a reasonable starting point for most content — matches the common Spotify and YouTube playback reference.
• —-16 LUFS may be preferable if Apple Music is your primary platform.

If you are preparing audio for a DJ set, podcast, or video sync rather than streaming distribution, different loudness conventions apply and the targets above may not be relevant.

A Note on Dynamic Range

Loudness normalization does not reward over-compression. A track mastered at -14 LUFS with natural dynamics will generally sound more open after normalization than a track that was brickwall-limited to the same number with all dynamics removed. The goal is not to hit a number at any cost — it is to reach a reasonable target while preserving as much of the original character as possible.

Can You Upload AI-Generated Music to Streaming Platforms?

Most major distribution services accept AI-generated music, provided you own or have the appropriate rights to the content and the platform's terms of service are met. Suno and Udio provide licensing terms that may allow distribution of tracks created on their platforms, depending on subscription level, usage rights, and current terms of service. Check the terms of service for each tool before distributing, as policies in this space continue to evolve.

Step 1 — Choose a Distribution Service

You cannot upload directly to Spotify or Apple Music as an independent creator. A distribution service acts as the intermediary. Common options include:

• —DistroKid — flat annual fee, offers plans that allow artists to retain most or all royalties, depending on selected service options.
• —TuneCore — per-release fee model, good for single releases.
• —CD Baby — one-time per-release fee, takes a percentage of royalties.
• —Amuse — free tier available with limitations.

Delivery times vary by distributor, platform, and review processes.

Step 2 — Prepare Your Audio File

Most distribution services require:

• —Format: WAV or FLAC (MP3 accepted by some, WAV recommended)
• —Bit depth: 16-bit or 24-bit (24-bit preferred)
• —Sample rate: 44.1 kHz
• —Channels: Stereo

Tracks commonly targeting -14 LUFS tend to play back more consistently on platforms like Spotify and YouTube. If you are exporting from WaveStudio, the 24-bit WAV output is formatted to meet standard distribution requirements.

Step 3 — Prepare Cover Art

• —Format: JPG or PNG
• —Size: 3000×3000 pixels minimum
• —Color space: RGB

Some distributors accept AI-generated artwork, though requirements and policies may vary. Verify with your chosen service before submitting.

Step 4 — Fill in Metadata

• —Track title — as you want it to appear publicly
• —Artist name — consistent across releases
• —Genre — select the closest match available
• —Release date — most distributors allow future scheduling
• —ISRC code — most distributors generate this automatically
• —Explicit content flag — required if the track contains explicit language

Genre metadata may influence how platforms categorize and recommend content. Choose the closest match rather than leaving it blank.

Step 5 — Set Your Release Date

Scheduling a release 1–2 weeks in advance gives you time to pitch the track to Spotify editorial playlists (via Spotify for Artists, where available) before the release date. Editorial playlist placement is competitive and not guaranteed for independent releases. Algorithmic playlists (Discover Weekly, Release Radar) are driven by listener engagement data and do not require pitching.

A Note on AI Music and Platform Policies

Streaming platform policies on AI-generated content are still developing. Some platforms have introduced disclosure requirements for AI-generated or AI-assisted music. Check the current policies of both your distribution service and the platforms you are targeting before each release — this is an area where guidelines may change.

Why This Comparison Matters

Suno and Udio are two widely used AI music generation tools. Both produce complete tracks from text prompts, but their audio output has different characteristics — and those differences affect how much mastering needs to do, and what the result sounds like. Individual results vary significantly by genre, prompt, and generation — treat the observations below as general tendencies, not guarantees.

Loudness and Dynamic Range

Both platforms generate tracks with dynamic processing already applied. The output is not "unmastered" in the traditional sense — it arrives with compression, limiting, and level balancing built into the generation model.

Suno tracks commonly come out at relatively high integrated loudness levels. Many fall in a range that sits above typical streaming normalization targets. The dynamic range can vary significantly between genres — electronic and hip-hop outputs tend to be more compressed; acoustic and ambient outputs sometimes retain more dynamic contrast.

Udio tracks tend to have somewhat more variation in output loudness across generations. Some outputs land closer to streaming targets; others come out louder or quieter than expected for the genre. Checking integrated LUFS before processing gives you a useful baseline in both cases.

Frequency Characteristics

Some Suno outputs may exhibit additional energy in upper-mid and high-frequency regions, depending on genre and generation. This can produce presence and clarity that works well for many genres, but in some outputs it manifests as brightness that becomes more apparent after loudness normalization.

Some Udio generations may present a fuller low-mid balance depending on genre and arrangement. The low end can occasionally feel heavier or less defined than a traditionally mixed track.

Neither platform produces a "neutral" mix in the way a human engineer targeting specific frequency targets would. The generation model has its own spectral tendencies that vary by genre and prompt phrasing.

How Each Responds to Mastering

The core mastering operations — loudness normalization, peak limiting, and broadband EQ — apply to both platforms in the same way. The difference is in how much correction is needed and where.

Tracks with high initial loudness levels may reveal compression artifacts more clearly after playback normalization or additional processing. Tracks that arrive closer to the target tend to respond more cleanly. In both cases, mastering works most predictably when the source material is internally consistent — similar energy levels throughout, no sudden peaks significantly louder than the rest of the track.

What to Listen For in the A/B Compare

• —Loudness consistency — does the mastered version sit at a more stable level relative to other tracks?
• —High-frequency balance — is the top end cleaner, harsher, or unchanged?
• —Low-end definition — does the bass feel tighter or looser after processing?
• —Dynamic feel — does the track feel more compressed, or roughly similar to the original?

Which Platform Produces Better Results?

This depends on the genre, the specific generation, and what "better" means for your use case. Many generated tracks may benefit from loudness and peak analysis before distribution, regardless of the generation platform used. The most consistent predictor of a clean mastering result is the starting loudness level and the internal consistency of the mix.

The Mastering Tool Landscape Has Changed

A few years ago, professional-quality mastering required either hiring a mastering engineer or investing in expensive desktop software. Many browser-based mastering tools now provide loudness normalization, peak limiting, and basic EQ correction without requiring traditional desktop software.

What Free Browser-Based Tools Do Well

Free browser-based mastering tools are well-suited to a specific workflow: a finished mix that needs loudness correction and peak limiting before streaming distribution.

• —No installation — works in any browser on any device
• —No subscription — no recurring cost
• —Fast turnaround — upload, process, download in under a minute for most tracks
• —Consistent processing — generally follows a consistent processing approach across tracks

Limitations include less control over processing parameters, no ability to adjust the EQ curve manually, and cloud-based processing may introduce variability in processing time depending on workload and infrastructure.

What Paid Desktop Tools Add

• —Manual EQ — adjust specific frequency bands based on what you hear, rather than relying on automatic analysis.
• —Multiband processing — compression and limiting applied differently to different frequency ranges.
• —Preference learning — some tools learn from feedback over multiple sessions, adjusting processing toward a sound the user has indicated they prefer.
• —Offline processing — files are processed locally, which avoids upload time for large files and keeps audio data on your machine.
• —Stem processing — some tools accept separate stems rather than a stereo mix, allowing more targeted adjustments.

WaveStudio vs PRISM Master Desktop: When Each Makes Sense

WaveStudio (free, browser-based) and PRISM Master Desktop (one-time purchase) share the same core loudness normalization approach. The difference is in control and adaptability.

WaveStudio makes sense when:

• —You want a quick loudness correction pass before uploading to streaming
• —You do not need to adjust processing parameters manually
• —You are working from any device and do not want to install software
• —You are processing occasional tracks rather than working through a catalog

PRISM Master Desktop may add value when:

• —You want to adjust EQ and dynamics parameters for specific tracks
• —You are processing a large number of tracks and want consistent stylistic output
• —You prefer offline processing and local file handling
• —You want the preference learning system to adapt to your taste over time

A Note on AI-Assisted Mastering Claims

Several tools in this space market themselves with strong claims about AI improving audio quality. Mastering is a well-defined technical process: loudness normalization, peak control, and EQ correction. Whether the decisions behind those steps are made by a human, a rules-based algorithm, or a neural network, the underlying operations are the same. The term "AI mastering" is used loosely across the industry — it is worth understanding what a tool actually does rather than how it is marketed.