Nobody remembers your track selection. They remember how the night felt — the bit where it lifted, the bit where it broke, the record that arrived exactly when it should have. That feeling isn't luck. It's a shape you decided on before you left the house, filled with records you know inside out. This lesson is how you build that.
By the end of this lesson you'll be able to…
- Draw the energy arc of a set and name its four phases.
- Play a warm-up properly — and know exactly what not to touch.
- Pick tracks that belong together on tempo, key, energy and character.
- Dig with intent and prep a set — record bag or crate — that plays itself.
01A set is a story, not a playlist
Here's the difference. A playlist is your twenty favourite records in a row. Every one is a banger, so it must be a great set, right? It isn't — it's exhausting. Twenty peaks in a row is the same as no peaks at all, because peak only means anything next to something quieter. Contrast is the whole engine. If nothing is ever restrained, nothing can ever release.
A set has a shape. It starts somewhere, goes somewhere, and lands. Think of it exactly like a film: you don't open on the explosion. You open on a room, and someone walking into it. The explosion means something an hour later because of that room.
The four phases
- Warm-up — establish the tempo, the mood, the fact that music is happening. Low energy, deep, patient. You're building the floor everything else stands on.
- Build — the long climb. Energy creeps up: drums get busier, basslines get bolder, vocals appear. Nobody should be able to point at the moment it changed.
- Peak — the pay-off. The biggest records, the moment you've been buying credit for all night. Shorter than people think — this is a stretch, not a plateau.
- Come-down / close — the landing. You bring it back down and let the room out gently, or you drop the one that everyone leaves humming. Both are legitimate; picking neither is not.
The arc below is the shape. It isn't a rule — it's the default that works, and you should understand it thoroughly before you go bending it.
Figure 1 · The energy arc of a set
Read the slope, not the height. The line rarely goes straight up — it climbs, eases off, climbs further. Those small dips in the build are how you buy the next lift: you can only raise energy from somewhere lower. Notice how short the peak is, and that the line comes back down. A set that ends at maximum is a set that just stopped.
Setting up / letting go
The climb — most of your set
Peak — spend the credit
Energy is a budget, not a tap. You start the night with a fixed amount of "lift" available, and every big record spends some of it. Play your three biggest tracks in the first half hour and you've spent the lot — the next two hours can only go sideways or down, and the room will feel it even if nobody can explain why. Peaks are affordable because valleys paid for them. So when you're tempted to drop the massive one early: don't. That's not restraint for its own sake, it's arithmetic.
02Warm-up is a skill, not a punishment
The most misunderstood job in DJing. New DJs treat the warm-up slot as an insult — an hour of playing to nobody while they wait for permission to do the real thing. So they don't warm up. They play their peak-time set at 9pm to eleven people and a bar staff, blow their best records on an empty room, and then wonder why they never get booked again.
Here's what's actually happening: you are setting the table. The headliner is going to walk into whatever room you hand them. Hand them a room that's warm, settled, at the right tempo, in the right mood, with somewhere left to go — and you've done a technically difficult job that only other DJs will notice. Hand them a room you already peaked and exhausted at 10pm, and you've made their job impossible. Promoters absolutely know the difference. It's the single fastest way to get re-booked.
A warm-up isn't "boring music". It's music that leaves room. Deep, groovy, hypnotic, atmospheric — often the most beautiful records in your bag. You're just refusing to resolve them.
Play this
- Below the night's tempo. If peak-time lands at 128, start around 120–124. Leave the headliner somewhere to accelerate to.
- Groove over drama. Deep, rolling, hypnotic. Records that are good without being events.
- Long blends. Let things sit. Sixteen bars is not too long. You're creating continuity, not showing off.
- Instrumentals and dubs. A big vocal hook is a landmark — landmarks belong later.
- Music that fits the next DJ. Listen to their sets. Aim the room at them.
- Atmosphere. Space, pads, texture. An empty-ish room sounds enormous — use it.
Not this
- Your best three records. They're not yours to spend. Wrong room, wrong hour, wasted.
- Peak-time tempo. Starting at 130 means the headliner either goes faster than the night wants, or drops the energy to fix your mess.
- Big drops and huge builds. You cannot pay off a drop to a room that hasn't arrived yet.
- Anthems. Every anthem you play is one the headliner now can't.
- Trick mixing. Nobody at 9:15pm needs six-second cuts and an echo-out every track.
- Sulking. Playing badly on purpose because the room is empty is the loudest thing you can do.
And the honest reward: warm-up is where you're allowed to be interesting. Peak-time is conservative by nature — a full room narrows your options. At 9pm you can play the weird nine-minute one. That's not the booby prize; a lot of DJs quietly prefer it.
03Choosing tracks that fit together
Two records can both be brilliant and still have no business next to each other. Whether they fit comes down to four filters — run every candidate through all four.
1 · Tempo
Work in a tempo range, not a fixed BPM. Most sets live inside a 6–8 BPM window that drifts upward across the night — say 120 at the start, 128 at the peak. Two tracks more than about ±6% apart start to sound stretched when you pitch them together: the drums get flabby or brittle, and vocals go obviously wrong. Modern key-lock helps, but it doesn't make a 118 record and a 132 record friends. Sort your crate by BPM and you'll see your set's spine immediately.
2 · Key
Two tracks in clashing keys will fight, and no amount of EQ fully hides it — you'll hear it as a sour, seasick wobble under the mix. Harmonic mixing solves it: stay on the same Camelot code, move ±1, or swap the letter. If none of that means anything yet, read Music Theory for DJs and then go play with the interactive Camelot wheel — ten minutes and it clicks. When you're prepping, filtering your crate by compatible keys turns "what next?" into a two-second glance.
3 · Energy
The one nobody tags and everybody needs. Energy is not tempo — a 124 BPM record can be a bulldozer and a 128 can be a hammock. Give every track in your crate an energy score of 1 to 5 yourself (most software has a star rating; use it for exactly this, not for "how much I like it"). Now your arc has coordinates: your warm-up is 1s and 2s, your build walks 2→4, your peak is your 5s, your close comes back to 2. Suddenly the set almost picks itself.
4 · Character — "does it sound like the same night?"
The subtle one, and the one that separates good DJs from correct ones. Two tracks can match on tempo, key and energy and still feel like a scene change — because one is warm, analogue and human and the other is cold, digital and hard. Ask it plainly: do these belong to the same night? Listen for the drum sound, the reverb, the era, the production polish, whether it's built for a dark basement or a festival field. When it works, the room stops noticing the joins and just stays. That's the whole aim.
Change one thing at a time. That's the working rule behind all four filters. Big lift in energy? Then hold the key and keep the tempo where it is. Jumping tempo? Then keep the character identical so the room only hears the speed. Moving into darker territory? Do it at the same energy so it reads as a mood change, not a crash. When two or three of those move together, the mix stops feeling like a transition and starts feeling like a different DJ took over. Change one, hold the rest — that's how a set stays continuous while going somewhere.
04Crate digging & finding your sound
Digging isn't browsing. Browsing is opening a store's new-releases page and scrolling until something goes "ooh". You'll do that forever and end up with 400 unrelated tracks and no set.
Digging with intent means digging for a hole. You come off a gig knowing exactly what you lacked: "I had nothing to get from the deep bit into the peak," or "my closers are weak." That's the brief. You go looking for a 124 BPM, energy-3, warm-and-rolling record in 8A to bridge that gap. Now a search has an answer, and every record you buy earns its place.
Where the good stuff hides
- Follow labels, not charts. A label is a filter someone else already ran for you. Find three you trust and go through their back catalogue properly — that's where a sound comes from.
- Mine other DJs' sets. Track ID from a set you loved, then look at what else that artist and label did. One good set is a reading list.
- Go backwards. Everyone digs this month's releases. Almost nobody digs 2004. The back catalogue is where you find records the whole room hasn't heard.
- Follow the credits. Loved a remix? Chase the remixer, not the original artist. That's usually the trail.
- Second-hand vinyl. The £1 bins are full of records nobody has played out in twenty years. Genuinely nobody else has them.
Why a small crate beats a huge one
Ten thousand tracks you've never heard is not a library — it's a search problem you brought to a gig. The room's waiting, and you're scrolling. Meanwhile the DJ with 300 records they know cold hears the current track's breakdown coming and their hand is already on the right one, because they know what's in there. Depth beats breadth every single time.
So set a bar and defend it. A track earns its way into your gig crate by proving it does a job — this is my energy-4 rolling bridge, this is my closer, this is the one for when it's not landing. If you can't say what a track is for, it isn't ready to be there. A well-known 300 will outplay an unheard 10,000 in any room on earth.
And that's where your sound comes from, incidentally. Not from deciding on one — from digging with intent for two years and then noticing the pattern. Your sound is the residue of what you kept saying no to.
05Prepping the set
Prep is where sets are actually won. The gig is just the bit where you find out whether you did it. The aim is simple: make every decision you can at home, so the only decision left in the booth is "what next?"
The digital prep checklist
- Analyse everything — and check the grid. Your software's beatgrid is a guess. On anything live, old, or hand-played it's often wrong by a hair, and a wrong grid means sync will drift and your cues won't line up. Fix the downbeat manually. This is the single most skipped step and the single most common cause of a mix falling apart.
- Key it — get Camelot codes on every track so harmonic filtering works. Trust the analysis about 95% and believe your ears over the tag when they disagree.
- Set cue points that mean something. Not one at 0:00. Cue 1 = the mix-in point (first proper beat you'd bring in on). Cue 2 = the breakdown. Cue 3 = the drop. Cue 4 = the mix-out. Use the same convention on every track and your hands stop thinking.
- Rate for energy, 1–5. As above — stars mean energy, never preference.
- Tag by job. Comment/tag fields: warmup, rolling, peak, closer, emergency. Genre tags tell you what a record is; job tags tell you what it's for, which is the only question you'll ask at 1am.
- Build crates by role, not genre. One crate per gig, sub-crates by phase. "Friday_Basement / 1_warm / 2_build / 3_peak / 4_close". Your set structure is now literally your folder structure.
- Trim it. A 3-hour gig needs maybe 60–80 tracks, not 500. Being able to see your whole crate is the point.
- Back it up. Two USB sticks, both tested in the actual player, plus the library on your laptop. Sticks fail on the night. This is not paranoia, it's Tuesday.
Prepping a vinyl set — a different discipline
Vinyl changes the rules, because on vinyl you cannot skip. There's no search field. If it isn't in the bag it does not exist tonight, and a record you can't find fast enough is a record you don't own. Everything below exists to solve that.
- The bag is the crate — and it's brutal. Forty to sixty records for a night, and every one has to justify weight. You'll leave good records at home. That constraint is exactly what makes vinyl sets feel decided.
- Order the bag by phase. Front-to-back: warm-up, build, peak, closers. You should be able to reach in blind and land in roughly the right decade of the night. Some DJs flip the peak records upside-down so a thumb finds them instantly in the dark.
- Sleeve notes are your metadata. Write on the sleeve — "124 · 8A · rolling · long intro · B2". That's your BPM, key, job, mix-in and which side. This is what the digital DJ gets for free and what you must build by hand.
- Know the side and the position. "B2" is not a detail, it's the whole thing. Fumbling a flip while a record runs out is how vinyl sets die.
- Check the condition before you bag it. A jump you've never noticed at home will find you at peak-time. Play the ones you're unsure of, all the way through, at home.
- Know your intros by length. No waveform means no warning. You need to know that this one has 32 bars of drums up front and that one drops at 16 — because you're mixing on memory and your ears. This is why learning to DJ on turntables makes everything else feel easy afterwards.
- Have a spine. Not a fixed order, but 3 or 4 records you know are going in and roughly when. On vinyl a plan does more work, because your improvisation is limited to what's actually in the bag.
The trade is real, and worth being honest about. Digital gives you infinite choice and instant recall; vinyl gives you commitment. Vinyl DJs plan harder because they have to — which is precisely why prepping a vinyl set will teach you more about set-building than a year of scrolling ever will. Even if you never play a record out, prep one bag properly once. It rewires how you think about the whole thing.
Figure 2 · The prep bench — digital vs vinyl
The formats differ in method, not in principle. Digital prep front-loads metadata so search does the finding. Vinyl prep front-loads decisions so your hands do the finding. Either way you're doing the same job at home: removing every choice from the gig except the one that matters.
Keep an "emergency" crate of five records. Five tracks that have never once failed you, tagged so you can find them with your eyes shut. Not your five favourites — your five most reliable. Every DJ has a night where the plan is simply wrong: the room is older than you were told, the previous DJ ended 10 BPM off, the promoter's brief was fiction. That's when you stop steering and reach for a known-good record, get the floor back, and rebuild from there. The emergency crate isn't an admission you'll fail — it's the thing that lets you take risks the rest of the night, because you always know the way home.
06Plan or improvise? The honest answer
This argument never ends, and both sides are wrong in the same way.
Fully pre-planned — every track in order, rehearsed at home. It'll be technically clean, and it will be dead. You are now performing a recording to people who are physically present, and you'll ignore the actual room to protect your track 7. Everyone can tell. If the set can't change, you're not DJing, you're pressing play with extra steps.
Fully improvised — "I just vibe it, man." Usually means an unprepped library and a set that wanders: no arc, no build, three peaks and a random ending. Freedom without a plan isn't spontaneity, it's drift.
What actually works
Plan the shape. Know the crate. Choose the record live.
- Fixed in advance: the arc. Where you start, roughly where the peak lands, how you're getting out. Your tempo range. The job you're doing in this slot.
- Fixed in advance: the crate — prepped, trimmed, tagged, and known cold. Plus a spine of 3–5 records you're fairly sure are going in, and one or two you know how you're closing with.
- Decided in the moment: literally every individual track choice, and the exact timing of every phase.
The plan is a route, not a script — you know you're heading north and roughly when you want to arrive; which roads you take is up to the traffic. And you must be able to abandon it. If the room tells you something different from what you assumed at home, the room wins. Every time. It's the only feedback that's actually real, and reading it is the pillar you can only build live — which is exactly why it sits where it does in the full curriculum.
The paradox worth sitting with: preparation is what makes improvisation possible. The DJ who can genuinely react to anything is the one who prepped hardest, because they're choosing from a crate they know instead of searching in one they don't. Freedom in the booth is something you build at your kitchen table.
07A worked example — two real slots
Same DJ, same crate, two completely different jobs. This is what "play the slot" means in practice.
Example A · The 2-hour warm-up
9pm–11pm · opening for a 128 BPM headliner · room fills at ~10:15
| Time | BPM | Energy | The job |
|---|
| 9:00–9:30 | 118–121 | 1 | Music exists. Deep, atmospheric, long blends. Nobody's dancing and that is correct — you're establishing tempo and mood for an empty room. |
| 9:30–10:15 | 121–123 | 2 | Drums firm up, first proper grooves. People drift from the bar. Still no vocals, still no drops. You're making the room feel safe to stand in. |
| 10:15–11:00 | 123–125 | 2 → 3 | Room's filling. Nudge — bolder basslines, a little more melody. Hand over at 125 and climbing, energy 3, with everything above it still unspent. |
You just did a hard job well. The headliner walks into a full, warm room at the right tempo with the entire top of the arc available to them. You played zero anthems and no big drops. Nobody in the crowd will mention you — and the promoter will book you again, because they watched the room fill on your watch.
Example B · The 1-hour peak-time set
1am–2am · room is full and already at energy 3 · you close the night
| Time | BPM | Energy | The job |
|---|
| 1:00–1:10 | 126–127 | 3 | Take the handover, don't reset it. Match where the last DJ left the room, sound like a continuation for two tracks, then start steering. Never open by dropping the energy to "make space for yourself". |
| 1:10–1:35 | 127–128 | 3 → 4 | The push. Tighter mixes, shorter blends, bigger records — but still climbing, not arrived. One deliberate dip around 1:25 to buy the last lift. |
| 1:35–1:50 | 128 | 5 | Peak. The 5s. Fifteen minutes, that's all — this is what the whole night has been paying for. Any longer and it flattens into noise. |
| 1:50–2:00 | 126–128 | 4 → 2 | The landing. Either the euphoric closer they leave singing, or bring it down and let them out gently. Decide which before you get there. |
Same arc, compressed and shifted up. An hour has no time for a warm-up — you inherit one. So the shape is the same (climb, dip, peak, land), it just runs faster and starts at energy 3 instead of 1. That's the real lesson: the arc doesn't change, the slot changes where you enter it.
Two hours, one hour, six hours in a basement — it's the same job. Work out where the room is, work out where it should be when you hand it over, and build the road between them out of records you know. That's set building. Everything else is detail.
Check your understanding
Three quick questions
Pick an answer — you'll get instant feedback. No sign-up, nothing saved.
Q1. You're opening at 9pm for a headliner who peaks at 128 BPM. What should your first half hour look like?
This is the classic warm-up mistake. Your biggest records spent on eleven people and the bar staff — and now the headliner inherits an exhausted room with nowhere left to climb. Those records aren't yours to spend in this slot.
Correct. Start below the night's tempo and low on energy. You're setting the table: establishing mood, filling the room, and leaving the entire top of the arc unspent for the DJ after you. Nobody in the crowd notices — every promoter does.
Consistent tempo isn't the goal — an arc is. Open at 128 and the headliner must either go faster than the night wants or drop the energy to repair it. Start 6–8 BPM lower and leave them somewhere to accelerate to.
Q2. Two tracks match on tempo, key and energy — but the mix still feels wrong. What's most likely missing?
The numbers are only three of the four filters. Tempo, key and energy get you a technically clean mix — they can't tell you whether the two records belong to the same world.
Not it — you have roughly ±6% of pitch to play with, and identical BPMs guarantee nothing anyway. Two records at exactly 126 can still be from completely different planets.
Correct. That's the fourth filter: character. Drum sound, era, warmth, production polish, dark-basement vs festival-field. One warm and analogue into one cold and digital reads as a scene change even when every number lines up. Ask it plainly: same night?
Q3. What's the most useful way to think about planning vs improvising?
Correct. Fix the shape — start point, roughly where the peak lands, how you get out — prep the crate until you know it cold, then pick every actual track in the moment. The plan is a route, not a script. And here's the paradox: preparation is exactly what makes real improvisation possible.
That's a recording performed at people who are physically present. It'll be clean and it'll be dead — because when the room says something different, a script can't listen. If the set can't change, you're pressing play with extra steps.
"I just vibe it" almost always means an unprepped library and a set that drifts — no arc, three peaks and a random ending. Freedom without a plan isn't spontaneity. The DJs who react best are the ones who prepped hardest.
You scored 0 / 3 —