How to check a mix before sending to mastering is mostly a matter of catching preventable problems while keeping the mix’s intent intact. You do not need to make the track louder, imitate a mastering chain, or remove every imperfection. You need to confirm that the balance works, the export is correct, and the mastering engineer receives a clean, clearly described file with enough room to work.
A useful pre-master check is repeatable: listen from the beginning, inspect a few technical measurements, compare against a reference at a matched level, and audition the final export outside the session. Each pass answers a different question. A meter cannot tell you whether the vocal feels right, and a good listening impression cannot reveal a truncated render or an accidental mute.
Start with the whole song at a comfortable monitoring level. Follow the arrangement rather than jumping straight to the loudest chorus. Listen for changes in vocal level, transitions, edits, noise, clicks, unwanted tails, and sections where an instrument unexpectedly masks another. Confirm that the intro, breakdowns, fills, and ending behave as intended. If the song has alternate versions or edits, make sure you are evaluating the correct one.
Then listen to the most exposed elements. Check the lead vocal or principal melody, kick and bass relationship, and any part that carries the hook. Ask whether the important material stays intelligible when the arrangement gets dense. Avoid making a broad tonal correction because one source is difficult to hear in one section; first identify whether the problem is level automation, masking, arrangement, or the source itself.
Check the mix quietly, then return to your normal working level. Low-level listening can make balance problems easier to notice, but it does not replace listening at a sensible reference level. Also audition in mono briefly. This can reveal elements that become weaker or disappear when the stereo channels are summed, often because of phase relationships or stereo processing. Mono is a diagnostic check, not a demand that the mix sound identical in mono and stereo.
If a decision still feels uncertain, take a break and return with a specific question. “Is the vocal clear through the second chorus?” is more useful than repeatedly asking whether the entire mix is good. Change one thing at a time, level-match the before-and-after, and keep a version you can restore.
Choose a reference that is relevant to the production, arrangement, and intended sound. Compare similar musical sections, such as chorus against chorus, and turn down the louder playback until the perceived level is reasonably close. A louder signal can seem fuller or more exciting; an unmatched comparison can lead you to over-EQ or over-compress your own mix.
Use the same monitoring path for both files. Check that your player or DAW is not applying different loudness normalization, EQ, or other processing to one of them. Do not assume the reference’s master-bus processing is a template to copy. The purpose is to identify differences worth investigating—such as vocal placement, low-end balance, brightness, or stereo width—not to erase the character of your production.
If the reference sounds more forward, decide whether that is a real mix issue or simply a difference in arrangement, performance, or mastering. A sparse verse will not carry the same density as a full chorus. If you use meters to support the comparison, compare equivalent sections and interpret the readings as descriptions of the audio, not targets that guarantee a particular sound.
Check the rendered mix’s peak level and listen for clipping or distortion. Sample-peak meters report the highest digital sample values; true-peak meters estimate peaks that can occur between samples during reconstruction. They answer related but different questions. If delivery instructions specify a peak ceiling or format, follow those instructions. Otherwise, do not force the mix to an arbitrary number solely because a meter displays it.
Integrated LUFS can help describe the average perceived loudness of a full track, while short-term and momentary readings describe shorter windows. None of these values tells you by itself whether the mix is balanced or whether it should be louder. Review the MixDiagnose LUFS checker to inspect loudness and peak readings, and listen to the song in context before deciding whether a number indicates a problem.
A spectrum display can help you investigate persistent low-frequency buildup, an unexpectedly thin range, or a narrow resonance, but it cannot determine whether a sound is musically appropriate. The MixDiagnose spectral analyzer can show where energy is concentrated. Use it to guide a focused listening check: solo or mute candidate sources, compare sections, and decide whether the energy is intentional. Do not cut every visible peak or try to make your spectrum match a reference curve.
Listen for audible consequences first: bass that obscures the kick, harshness that distracts from the vocal, or low-mid energy that clouds a dense passage. Then use meters and spectrum analysis to narrow the search. A measurement is most useful when it helps answer a question you can also verify by listening.
Review the master bus for unintended processing, clipping, or gain changes. If you have added a limiter, maximizer, or heavy bus compression only to make the mix feel finished, compare with it bypassed at a matched output level. The louder version will often seem preferable until you compensate for its gain. Keep processing that is part of the sound you want; remove processing that is there only because you assume mastering requires a pre-limited mix.
Do not normalize or raise the entire mix just to reach a preferred peak value unless the delivery brief calls for it. Leave the mix’s dynamics and headroom in a state that allows the mastering engineer to work, and communicate any intentional clipping, saturation, or bus processing that should remain. There is no universal amount of headroom that suits every session and workflow. The receiving engineer’s technical requirements take priority.
Check stereo effects and low-frequency content in mono, but do not remove width automatically. A wide chorus may be intentional; the important question is whether essential parts remain stable enough for the intended use. Listen for sudden shifts, distracting phase effects, or a bass element that loses too much weight when summed. If those issues are part of the sound by design, make that clear in your notes.
Export the exact version you reviewed. Confirm the file begins and ends where intended, includes the full reverb and delay tails, and has the requested channel count, sample rate, bit depth, and file format. These specifications vary by project and mastering workflow, so use the engineer’s instructions rather than assuming a universal delivery setting. If no specifications were provided, ask before converting or changing the file.
Re-import the rendered file into a fresh session or open it in a separate player. Listen from start to finish and confirm it is not silent, clipped unexpectedly, missing a channel, or cut off early. Check that the filename identifies the song and version clearly. If you are sending multiple mixes, label instrumental, clean, or alternate versions unambiguously.
Include concise notes: the reference tracks and what you value about them, any intentional mix-bus processing, known edits or alternate versions, and anything the engineer should preserve. Do not ask the mastering engineer to solve a mix balance problem without mentioning it. If the vocal is intentionally understated or the bass distortion is a deliberate production choice, say so. Clear context helps distinguish an intentional decision from a technical oversight.
How to check a mix before sending to mastering comes down to listening in context, making fair reference comparisons, using meters to investigate rather than dictate, and verifying the actual export. Fix audible mix problems in the mix, follow the recipient’s delivery requirements, and send notes that explain the intent. A careful final check can prevent avoidable revisions without turning the mix into a premaster imitation.
Before you send the track, run it through MixDiagnose for another view of its technical balance and loudness. Use that information to guide your final listening pass, then deliver the version you have actually checked.