Mixing

Mix Checklist Scorer

Most mix feedback collapses into one vague sentence: it sounds muddy. That sentence is useless in the session, because a muddy mix and a bright mix fail in opposite directions and need opposite fixes. This scorer works as a checklist instead. Fourteen checks, each on a one to five scale, each carrying a weight, grouped into four regions: low-end, dynamics, space and translation. The kick and bass relationship carries more weight than the ambience behind the vocal, because getting it wrong costs you more than getting it wrong the other way. Every check comes back as a number with a specific action attached, and the regions come back as four scores so you can see which one is actually costing you the mix. The output also names the weakest region and tells you what to fix first, and it prints as plain text so you can drop it into a session note. This is a rubric, not a measurement: it does not read your audio file, and it cannot tell you what is actually in the low end. It tells you what to listen for and in which order.

Balance · Kick and bass share the low end cleanly

1 is a mess, 5 is a locked low end

Balance · Low mids stay clear under a dense arrangement

Listen for guitars and pads covering the vocal

Low end · Sub sits below the bass fundamental

Does the sub carry on a phone speaker

Low end · Low end stays mono below 200 Hz

Check the sub image in mono

Dynamics · Transient detail survives compression

Snare and kick should still have an edge

Dynamics · Crest factor keeps the peaks punchy

Compare against the crest factor of your references

Dynamics · Sidechain control feels intentional

Pumping should not chop the arrangement

Space · Stereo width reads on headphones

Check the image in mono as well

Space · Depth comes from space rather than from level

Pull the sends down and the dry up

Space · Ambience and room sit behind the vocal

The words should sit in front of the room

Translation · Levels match the reference tracks

Level match, do not eyeball it

Translation · The reference reveals masking you could not hear

A reference exposes what the session hides

Translation · The mix survives a phone speaker

If the vocal vanishes, the mid scoop is too deep

Translation · Headroom is left for mastering

Aim for a peak near -6 dBFS on the mix bus

Monitoring level is calibrated

An uncalibrated session makes every score on this page unreliable

Result

How to use it

STEP 1

How to use this scorer: rate the mix as it is right now, not as you want it to be. Every check is a listening test with a specific thing to listen for, and a score of 4 means you can hear it working. If you are not sure, mark 3 and move on, because a wrong 5 costs more than an honest 3 once you start fixing things in the wrong order.

STEP 2

Do the listening pass on the mix bus, not on the individual tracks, and check it in the order the regions are listed. Low-end first, then dynamics, then space, then translation. A mix that fails low-end and translation at the same time usually needs two separate fixes, and the second one is invisible until the first one is done.

STEP 3

Read the weakest region line before anything else. It names the area that is costing you the most, and the summary paragraph under the scores gives you the first move for that region. Fixing a 42 out of 100 low end buys more than moving a 75 out of 100 dynamics up to 90, because the second change is harder to hear once the first one is fixed.

STEP 4

Take the printable block into the session. It lists all 14 checks with their weights and marks the ones at 4 or above, so it works as a punch list for the next hour and as a record of what the mix looked like before you touched it.

Three common mistakes

Common mistake: scoring the session instead of the mix. The mix bus sounds better than the print, and the print sounds better than the release. Score what you would actually send, and if the only version that exists is the rough bounce, score that and write down the gap. Scoring a mix you have never played back on a phone is scoring a session.
Watch out for giving every check a 4. When the sliders all sit in the same place the region scores stop meaning anything, and a scorer that always returns 75 is decoration rather than a tool. If you cannot tell whether a check is a 4 or a 3, it is a 3. The value of the page comes from the spread, not from the average.
Key point that gets missed: the low end is where translation dies first. A sub that fights the kick and a low mid scoop that hides a guitar both pass on studio monitors and both vanish on a phone speaker, which is where most people actually hear the track. Fixing the low end before the space work is the order that saves sessions, not a style preference.
Easy to get wrong: treating the headroom check as a formality. A mix bus peaking at 0 dBFS with no headroom gives the mastering engineer nothing to work with, and the limiter on the master then sets the loudness for you. Aim for a peak near -6 dBFS, which is not a rule but a reasonable place to leave room for glue compression and the final limiter.

What the four regions cover, and what they do not

The fourteen checks are grouped into four regions, and each region answers a different question about the mix. Low-end covers four checks: the kick and bass relationship, low mid clarity, where the sub sits relative to the bass fundamental, and whether the low end is mono. Dynamics covers three: transient survival, crest factor and whether the sidechain control is intentional. Space covers three: stereo width, whether depth comes from sends rather than from level, and whether the ambience sits behind the vocal. Translation covers four: level matching against references, whether a reference exposes masking, phone speaker survival and headroom.

Low-end is the region that decides whether the track works on a phone, and it is the region people habitually rate highest and get wrong most.
Dynamics is where the transient lives, and losing it is the most common cause of a mix that sounds expensive but dead.
Space is the cheapest region to improve, because sends and levels are free to change, but it is also the one most often confused with loudness.
Translation is the only region that depends on something outside the session, which is why it needs a calibrated monitoring level.

None of this is a measurement. The page does not read the file, it cannot measure LUFS, true peak or crest factor, and it cannot hear masking. It is a structured list of things that are usually wrong, each one phrased so you can check it against your own ears in about a minute.

How the weights and the score work

Each check is rated from 1 to 5 and carries a weight from 1 to 3, and the score is a weighted average rather than a plain one. That is the part that matters: a 2 on the kick and bass relationship costs you three times as much as a 2 on whether the ambience sits behind the vocal, and the totals are arranged to reflect that.

Every check is converted to a 0 to 100 scale, so 1 is worth 0, 3 is worth 50 and 5 is worth 100.
The check score is multiplied by its weight, and the region score is the weighted average of the checks inside that region.
The total is the weighted average across all 14 checks, with a total weight of 36 across the four regions.
Weights of 3 go on the failure modes that cost a mix its translation: the kick and bass, the sub, transient survival, the crest factor, depth from sends, level matching, phone survival and headroom.
Weights of 1 and 2 go on the checks that are cheaper to fix and cheaper to get wrong.

The default set of ratings in the form describes a typical home-studio mix: strong on space and dynamics, acceptable on translation, and weak in the low end, which is the most common shape of a mix that a friend says sounds muddy. It scores 57.6 out of 100, and the low end is named as the region to fix first.

Turning the weakest region into a session plan

The value of a score is what you do with the worst number on it, so the page turns each weak check into a specific action rather than a description of the problem. The actions are written to be done in order, because they interact: a mix bus that is too loud cannot be fixed by adding a compressor, and a low end that is masking itself will read as a translation problem even after you fix the translation.

Low-end first: sidechain the bass to the kick or high-pass it at 120 Hz, pull the sub down 2 to 3 dB, then fold the low end to mono and listen for the bass thinning out.
Dynamics second: drop the ratio to 2:1 or 3:1 on the mix bus, or use parallel compression on a separate send, and check that the snare transient is still there.
Space third: pull the reverb and delay returns down 2 dB, bring the dry up, and keep the ambience behind the vocal rather than around it.
Translation last: level match the references at around -9 LUFS integrated, check the phone speaker, and leave the peak near -6 dBFS so the mastering engineer has headroom.

Each of those steps has a measurable result, which is the other reason to work through them in this order. You can hear whether the fix worked, which is more than most mix advice offers.

FAQ

Can this scorer tell me what is actually wrong with my mix?

No, and it is built the other way round. The page gives you a list of the failure modes that cost mixes the most, in an order that reflects how often they are the real problem, and asks you to check each one by ear. That has a real advantage over a measurement tool: it catches things a meter cannot, such as a sub that masks a bass note at one frequency, or a reverb return that is drowning a vocal while the numbers look fine. What it cannot do is replace checking LUFS, true peak and crest factor in a real meter.

Why do the region scores not add up to the total score?

Because the total is a weighted average across all 14 checks, while each region is a weighted average of the checks inside it. The regions carry different total weights, so a strong region with many checks moves the total more than a weak region with two. With the default ratings, space scores 67.9 and low-end scores 42.5, and the total lands at 57.6 because the low-end checks carry a heavier weight each. That is intentional: the point is to make the heavy items count for more.

What does calibrated monitoring actually mean?

It means your monitor level is set to a known reference, usually 85 dB SPL for a calibrated pair, and you do not change it while you work. Without that, every loudness judgement in the session is a memory rather than a measurement, and memory drifts in the direction of the last thing you played. A cheap sound level meter and one reference track that you level match at the start of every session will make this page useful; without it, the translation region is a guess.

Should I score the rough bounce or the finished master?

Score the finished mix, before mastering. Mastering changes loudness, adds a limiter and can flatten dynamics, so a score taken after mastering measures the mastering engineer rather than your mix. If the only file you have is the master, score it anyway and note the difference, because the headroom and translation checks will read lower than they should and you will know not to chase them.

How long should a mix session with this list take?

The rating pass takes about ten minutes on a mix you know well, and less if you are honest about the checks you have never thought about before. The fixing takes as long as it takes, but working the weakest region first usually means the second session is shorter, because the checks that scored low often have a single cause. A useful habit is to run it at the start of the session and again before you bounce, and to print the block so the second run has something to compare against.

Limits of this tool

Limits: this is a rubric, not a measurement. It does not read your audio file and it cannot measure LUFS, true peak, crest factor or masking, so every score is an approximation of what you were asked to listen for.
Not for: mastering, for mix revisions on stems you cannot hear in context, and for tracks where the arrangement has not been sorted yet. Scoring a mix that is not finished produces confident nonsense.
When this stops: with stems printed per section, with a mix that changes on every pass, or with tracks in a different key and tempo per bar, the region scores stop being comparable and the totals stop meaning anything.
Scope: the checklist covers balance, low end, dynamics, space and translation. It says nothing about arrangement, writing, sound design, selection or mix bus routing, which is where a large share of real mixing time goes.