Cubase mix analysis — find what's wrong and how to fix it

Drop a bounce from Cubase and Unveil tells you why it isn't sounding pro yet — loudness, dynamics, tonal balance, phase, and frequency conflicts — then gives you fixes written for Cubase's own stock plugins. No third-party plugins to buy — just a free account to try it.

Cubase's signal path and the pan law you start with

Cubase processes audio internally at 32-bit float, with the option of 64-bit float processing in the project settings in recent versions. As with other floating-point DAWs, a red clip light on the Stereo Out in your mixer does not mean internal distortion, but the exported file and your hardware are bound by 0 dBFS.

Cubase's Project Setup includes a Stereo Pan Law setting, with Equal Power as the default. This keeps perceived loudness steady as a source is moved across the stereo field, at the cost of a slight dip when centered. If you open a project from a colleague whose pan law is different, centered mono elements will change level slightly, so check this setting before comparing mixes.

Mixer meters can be configured in the preferences to show different behavior, such as peak with hold, or RMS-style readings, and pre- or post-fader. Set post-fader so the meters show what actually reaches the next stage.

SuperVision and the metering you already own

Recent Cubase Pro versions include SuperVision, a modular metering plug-in. You can assemble a window with a spectrum analyzer, level meters, a loudness module and a phase scope in one place. Put it on the Stereo Out as the last insert, post-fader, and play the song from the beginning to read an integrated loudness figure.

Which meters are available depends on your Cubase edition, so open the plug-in and see what modules you have. If your edition lacks a loudness module, use a third-party meter, since BS.1770 integrated LUFS cannot be inferred from the peak meter. The references to compare with are roughly -14 LUFS for Spotify and YouTube and roughly -16 LUFS for Apple Music, with a true peak of about -1 dBTP.

Make a habit of reading three things together: loudness over the whole program, peak or true peak at the loudest moment, and the correlation or phase scope. A mix can look fine on level and still collapse in mono if the bass is wide.

Frequency, Compressor and Maximizer

Frequency is the modern equalizer, with up to eight bands, per-band stereo modes including Mid/Side, and a dynamic range display. Older projects may use StudioEQ, a four-band equalizer, still perfectly good for broad moves. Because Frequency allows each band to work in Mid or Side, it is the right tool for cutting low frequencies from the sides to improve mono compatibility without touching the center.

Compressor in Cubase includes a variety of controls including attack, release, an auto-release option, a Punch control and a built-in sidechain filter. Use it on individual instruments, and on groups with moderate settings, such as 2:1 and 1 to 3 dB of gain reduction.

Maximizer is the stock mastering-style limiter. It has Optimize and Output controls, and Modern and Classic modes. Set Output to -1.0 dB, raise Optimize slowly, and stop when punch fades. Cubase also provides a Brickwall Limiter, a simpler ceiling-focused device useful as a safety stage at the end of a chain.

Export Audio Mixdown, step by step

Choose File, Export, Audio Mixdown. Select the Stereo Out channel as the source, set the locators around the whole song including tails, and pick WAV.

1. Set the sample rate to the project rate unless a delivery spec says otherwise.
2. Set the bit depth to 24 for a mastering handover, or 16 for a final delivery.
3. For 16-bit output, put a dither plug-in last on the Stereo Out inserts (Cubase includes UV22HR from Apogee) and leave it off for 24-bit.
4. Check whether the dialog has any normalization or loudness options enabled, and disable them for analysis.
5. Use Real-Time Export only if hardware inserts or plug-ins require it; otherwise offline is faster and sample-accurate.
6. Import the resulting file onto a new track in a separate project and play the loudest section against the original to catch any differences.

Worked fix: boxy mids near 400 Hz

Imagine analysis puts the low-mid band 3 to 5 dB over target, with the sound described as boxy, like the music is in a small room. Often the source is acoustic guitar, snare body or a cluster of keyboard parts.

1. Open Frequency on the most prominent offender and enable the analyzer.
2. Create a bell filter at 400 Hz, Q about 1.4, and boost 6 dB briefly to locate the box tone. Sweep from 300 to 500 Hz.
3. Once the offending zone is found, reduce to a cut of 3 dB. If the problem sits on the sides only, use the band in Side mode.
4. Apply a smaller, wider cut of 1.5 dB with Q 0.7 on the group bus holding all the mid-range instruments. Two small cuts are less audible than one deep one.
5. Check the vocal: if the cut thins it, give back 1 dB with a gentle shelf at 200 Hz, or reduce the cut.
6. Re-open SuperVision and compare spectrum before and after. A combined 4 to 4.5 dB reduction in the zone should bring the band back near target.

Cubase-specific stumbling blocks

• Forgetting that Channel strip EQs and the inserts are two separate places; hidden EQ on the channel strip can double the correction applied by a plug-in.
• Leaving the Control Room routing active and monitoring through a different path than the export; the mixdown follows the output bus you choose, not the Control Room.
• Using Maximizer on the Stereo Out during mixing and sending the result as a pre-master.
• Ignoring the project Pan Law when matching to references imported from another DAW.
• Exporting from the wrong locator range and losing a reverb tail, which also shifts the measured loudness range.

Closing checklist

Measure the final file again to confirm the corrected balance.

• SuperVision last on Stereo Out, post-fader, read over the whole song.
• True peak about -1 dBTP for a final; pre-master peaks near -6 dBFS.
• Low end narrowed with a Side-mode cut or mono function if the phase scope shows spread.
• Dither only for the final 16-bit render.
• Re-import the mixdown and compare with the live project.

What you get

• A 0–100 mix score with a per-metric breakdown (LUFS, true peak, dynamic range, stereo, tonal balance).
• Time-localized frequency conflicts — the exact moment and band where elements collide.
• Step-by-step fixes using Frequency EQ, Compressor, and Maximizer — e.g. a corrective Frequency EQ band at the flagged frequency.
• Genre-aware targets so a trap bounce isn't judged against an acoustic record.

Questions

Related

• Free online LUFS checker
• Dynamic range analysis: DR, crest factor & LRA
• Hip-Hop mix analysis
• Modern Pop mix analysis
• Frequency spectrum analysis, explained for producers
• Ableton Live mix analysis