When Master Tempo makes the sound worse.

When Master Tempo makes the sound worse

What is documented

Before considering heuristics, establish the three facts that support everything else. They are the only hard data available:

1. rekordbox warns about reprocessing. The manual’s statement that sound quality may change because the audio is digitally reprocessed appears twice, beside the Master Tempo descriptions in the function panel and platter sections. It is not a rumour; it is a manufacturer warning built into the feature’s documentation.
2. Traktor documents artefacts between tempo zones. Its manual explains that variable-tempo tracks can be tempo masters without straightening their dynamic structure, and that Key Lock should be enabled on synchronised decks to avoid audible pitch-shifting artefacts when moving between tempo zones. Notice the paradox: here, Key Lock is the solution rather than the problem.
3. Traktor sets an operating limit. In Scratch mode, Key Lock switches off below −30 % or above +50 % speed to achieve more natural scratching with timecode or jog wheels.

A fourth fact joins these three, unrelated to audio: the Traktor manual mentions Key Lock on all decks as an example of a processor-intensive task when explaining multicore support.

Reprocessing is the method rather than a defect

Understanding the warning is worthwhile. Maintaining pitch while changing duration requires stretching or compressing the material without moving its frequencies. That cannot happen without processing: the signal must be reconstructed.

The honest statement is therefore that the sound changes, rather than that it sounds worse. The decision cannot be settled with a fixed rule, because it depends on the material and how far you stretch it.

The design consequence is clear: Master Tempo is a tool rather than an installation setting. Enable it when key matters more than the purity of the reprocessing, and disable it when reprocessing interferes more than the pitch change.

What time stretching actually is

Preserving pitch while changing audio duration is called time stretching. Its sister operation, changing pitch without changing duration, is the pitch shifting mentioned above. Master Tempo is the booth name for the former.

The most detailed public reference for these processes is not a DJ document. It is the Ableton Live manual, which describes the same algorithm families you have to work with:

• Transient-based modes. Material is split into segments and each attack is treated separately. The manual is precise: Beats-mode granulation is optimised to preserve audio transients.
• Spectral or texture modes. Complex and Complex Pro use, according to the manual itself, an entirely different technology, intended for files containing several layers at once.

On one side, AlphaTheta answers the question “Does MASTER TEMPO distort?” with a statement worth remembering: sound quality changes because of the processing that keeps pitch constant, and that change is part of the equipment’s specifications. On the other, Ableton warns that Complex modes are never neutral, even at the original tempo. There is no public manual for the DJ algorithm, but each manufacturer’s description concerns the same problem: stretching without changing pitch has a cost, and that cost is a system characteristic rather than an isolated fault.

Time-stretching artefacts

Artefacts are what you hear when the process becomes audible: a metallic vocal timbre, a granular hi-hat, or a kick arriving with a tail or duplicated attack. They were absent from the original recording; the adjustment produced them.

What is documented:

• They appear even in the most careful mode. Ableton warns that preserving transients in Beats mode can produce artefacts, eliminated by preserving sixteenths instead of transients or switching modes. In booth terms, no mode is cost-free; each has a different cost.
• The more comprehensive modes are not clean either. Complex and Complex Pro, although they sound better with mixed material, are never neutral: the processing leaves a trace even without moving the tempo.
• The DJ equipment manufacturer calls it a change in sound. AlphaTheta speaks of quality changing and being part of the device’s technical specification, rather than artefacts.
• Between tempo zones, Traktor does use the word. It documents audible pitch-shifting artefacts when crossing tempo zones, and recommends enabling Key Lock on synchronised decks to avoid them. Here, reprocessing is the solution.

A practical heuristic: if a passage sounds odd, repeat it in another tempo range before blaming the program. A large jump overloads the algorithm; sometimes the same move over a smaller range makes the problem disappear. The full procedure is in this guide’s comparison section.

Transients and algorithms: where it becomes noticeable first

A transient is the brief energy peak at the start of a hit: kick, snare, hi-hat or an instrument’s attack. Ableton defines transients as amplitude peaks indicating where notes or beats begin. A transient-based algorithm must locate precisely those points to cut and reconstruct the audio.

When that detection fails, the error is heard in the hit rather than the pad: the attack is cut too early, repeated or arrives with a tail that was not there. That is the expected consequence of a method that depends on guessing where each note begins, and fits the documented point that preserving transients can introduce artefacts.

In practice, percussive material reveals stretching before sustained material does. This is a practical heuristic rather than doctrine: pads and long bass notes forgive a poorly cut segment; a duplicated kick does not. The useful check is therefore the rhythm section, with and without Master Tempo, in the actual tempo range you will use. The procedure appears in How to compare without fooling yourself below.

To be clear: nobody has published any brand’s DJ algorithm, so this section explains the mechanism rather than certifying a particular product’s result.

Signs that stretching is becoming audible

Practical heuristics, rather than official doctrine. No verified source lists these cases. They are recurring booth observations that fit what is documented. Treat them as checking criteria rather than laws:

• Large tempo jumps. The further you move speed, the harder the algorithm must work. A couple of BPM and a jump of ten do not behave the same way.
• Exposed vocals or high frequencies. Passages with a solo vocal or lots of information at the top of the spectrum tend to reveal processing first.
• Continuous tempo changes. Variable-tempo tracks combine several stretching regimes within one track; Traktor specifically recommends Key Lock on synchronised sections to avoid pitch-shifting artefacts.
• A system at its limit. If CPU load is already high and you also run Key Lock on every deck, you add more work. The manual identifies that case as demanding.
• Prepared keys. Ironically, Master Tempo helps here: if you planned the mix around compatible keys, pitch following tempo is what breaks it.

How to compare without fooling yourself

This procedure is practical methodology rather than a manual quotation, but prevents the common mistake of deciding from a three-second impression.

1. Prepare two versions of the same moment. Use the same passage and tempo jump, once with Master Tempo and once without it.
2. Listen in both places. Headphones reveal detail; the booth reveals what matters to the dancefloor. The verdict should come from the latter.
3. Raise the level of the fragile passage. Not to force a failure, but so your ear does not have to strain: the solo vocal, hi-hat or pad.
4. Use the actual range you will play. Comparing at ±2 % and then playing at ±8 % invalidates the test.
5. If unsure, decide by key. Where uncertainty is equal, harmonic mixing is a more objective criterion than an impression of timbre.

One detail is supported by the documentation: do not turn off Master Tempo without checking what happens to pitch. The rekordbox manual describes the key display changing as you move tempo with master tempo off. That pitch change is audible and can be a bigger problem in a long transition than the reprocessing you were trying to avoid.

If the problem is the tempo range

Before blaming the algorithm, rule out other causes. “Worse sound” during a synchronised mix can come from several places at once:

• The master’s grid, if the offset forces you to adjust tempo strangely: review What happens when the Master Deck’s beatgrid is wrong.
• Gain, if what sounds dirty is actually saturation: check trim and master levels before attributing anything to stretching.
• The range itself, if you made an excessive tempo jump to align two tracks that should not go together.

Only after ruling these out does it make sense to attribute the loss to Master Tempo, and only then is switching it off worthwhile.

What has not been verified

This deserves explicit treatment, because it is sensitive content and the temptation to fill gaps is strong:

• There is no official algorithm comparison between rekordbox and Traktor. You cannot claim which sounds better or stretches better. Nobody has published a verifiable measurement in this cluster’s sources.
• There is no official percentage threshold at which stretching becomes audible.
• There is no official list of problematic material, whether genres, instrument types or tempos.
• There are no verified Serato or CDJ data in this cluster. Any statement about them would be invented.
• The −30 %/+50 % limit belongs to Traktor and its Scratch mode. It must not be extrapolated to rekordbox or any other product.

Common mistakes

“Master Tempo ruins the sound, so I always turn it off.” The manual says quality may change, not that it always gets worse. Switching it off also introduces a pitch change with every tempo move, which is worse in many mixes.

“Since the manual warns about it, it audibly sounds bad.” The warning concerns digital reprocessing without quantification. Audibility depends on the material and range, and no table fixes it.

“All programs stretch the same way.” That has not been verified. There is no official comparison, and algorithm equivalence cannot be asserted.

“rekordbox disables it beyond a certain percentage.” That is not documented in manual 7.2.18. The figure belongs to Traktor.

“If it sounds bad, I turn off Master Tempo and I am done.” The problem might have been gain, the master’s grid or an absurd tempo jump. Disabling the feature without diagnosis trades one symptom for another.

“Key Lock only costs audio quality.” It costs CPU too, and Traktor documents Key Lock on all decks as a demanding task. This is a system cost, noticeable in long sessions.

Once you understand what Master Tempo does and when to pay its price, learn to align tracks without relying on Sync: How to beatmatch manually, step by step.