The grid is not the truth

A recorded drummer does not play a grid. They play behind, ahead and around it, and the size of those deviations is much of what tells a listener the part was performed rather than assembled. Two recordings of the same part can be measured as equally accurate and differ enormously in how alive they sound.

Quantisation is the process of moving those notes onto the grid. It is a genuinely useful tool for fixing a performance that is inconsistent, and it is destructive when applied to a performance that is already good. Nothing distinguishes the two cases except listening.

Microtiming cannot be recovered after quantisation flattens it. Whatever strength you apply is the strongest the performance will ever be again.

What the strength dial actually does

Most quantisers expose a strength control between zero and one hundred percent. At zero the timing is untouched. As you raise it, notes move progressively closer to the grid, and at full strength every note lands exactly on it. The intermediate values do not snap, they scale the correction, which means a performance with a few notes badly out of place gets pulled in evenly rather than having only its worst moments corrected.

StrengthEffect on timingWhat you are choosing
0%UntouchedKeep the performance as played
25-50%Deviations halvedTidy up without flattening
50-75%Most notes near the gridAcceptable for simple grooves
100%Exactly on the gridAccuracy only, no feel
Abstract scattered timing artwork representing microtiming and the humanisation of performed parts
Perfectly identical timing is one of the least human things in a recording. Small deliberate variations are what make a part sound played rather than assembled.

The swing control is separate and easier to misuse. Swing moves every off-beat note later by a fixed amount, expressed as a percentage of the beat, which changes feel substantially and content not at all. A swing value above about sixty percent is a distinctive rhythmic choice rather than a correction, and it should be chosen deliberately.

Select elements, not the pattern

The most common mistake is applying quantisation globally to every element of a drum pattern. Kick and snare usually carry the intended groove, and quantising them alongside a hi-hat that genuinely needed tightening removes the thing you were trying to preserve.

  • Identify which element is actually inconsistent, and correct only that one.
  • Leave the elements that define the groove alone unless they are audibly out of place with the arrangement.
  • For layered and ghosted material, quantise the layer that fights the rest and leave the one that adds movement.
  • Small deviations on the snare backbeat are usually the most audible loss when strength is raised.

If the material was recorded loosely enough that tempo mapping is involved, timing and pitch become coupled because stretching changes both. That coupling is the reason a badly warped loop can sound worse than a quantised one, and it is worth handling the warp before touching quantisation strength.

Putting some of it back

When a part has been quantised hard and now sounds mechanical, the fix is to reintroduce small deviations deliberately rather than to undo the quantisation. Most sequencers offer a humanise function that adds random timing and velocity variation, and a little of it goes a long way.

The useful range is smaller than people expect. Deviations of a few milliseconds are audible in aggregate without any individual hit sounding wrong, and keeping them that small preserves the tightness that makes a programmed part work. Randomising heavily produces a different kind of wrong, one that sounds like sloppy rather than live.

Abstract swing artwork representing swing ratio and groove depth in a quantised grid
Swing moves every off-beat note later by the same amount. It changes feel more than it changes content, and it cannot be recovered after quantisation flattens it.

The bottom line

Quantisation is a trade between accuracy and feel, not a quality setting. Full strength snaps everything to the grid and removes the small deviations that make a performance sound played. The right setting is the lowest strength that fixes the actual problem, because the deviations you delete cannot be recovered once they are gone.

Judge timing against a reference

A retune moves every channel by the same ratio, so the timing relationship between your parts is preserved exactly while you compare them.

Explore 432Hz MASTER

Frequently asked questions

Should I fully quantise my drums?

Rarely. Full strength snaps every hit to the grid, and the small deviations that make a performance sound human are removed with them. Most producers end up somewhere in the middle, with low strength applied to specific elements rather than the whole pattern.

What is microtiming?

The deliberate placement of a note slightly ahead of or behind the grid, usually by a few milliseconds. It is one of the main tools for controlling feel, and it is the first thing lost when quantisation strength is raised.

Does quantisation change the tempo?

No. Tempo is a rhythmic property and quantisation only moves notes in time, so the tempo value itself is untouched. Any process that changes tempo as a side effect of pitch is a retune, which is a different operation.

Should I quantise at all?

Use it to fix a problem rather than to achieve accuracy. If a hi-hat is genuinely dragging out of time with the rest of the arrangement, quantise that element lightly. If everything is already close, quantising usually makes the track worse.

More from the atixUniverse blog