Is Your Library Actually Big Enough for a 3-Hour Set?

You've booked a 3-hour slot — maybe an opening set, maybe a full headline night, maybe a b2b that's really just you playing for 90 minutes twice. You go looking for the maths: how many tracks does a set that long actually need? You find the answer, plug in your genre, and get a number. Forty-eight tracks. Fifty-two. Something in that range. You feel briefly reassured, then you open your crate and the question changes shape entirely: do you actually have forty-eight tracks that work?

That's a different question, and it's the one this piece is about. If you haven't planned the actual shape of a long set before — where the peak sits, how you land the ending, how energy should move across three hours rather than one — our set-planning guide is the place to start first. This piece assumes you know roughly what a long set is supposed to feel like, and asks a narrower, more practical question: is your library built to deliver it.

The formula everyone quotes, and the question it can't answer

The standard track-count maths is genuinely useful, and it's worth restating because it's the baseline everything else in this piece builds on. As a general rule for standard mixing, you're looking at 18–22 tracks per hour — roughly 2.5 to 3.5 minutes of true “play time” per track once you account for overlapping blends. Tech house, with its longer intros and extended arrangements, tends to run a little lower at around 20 tracks per hour. Drum and bass, where tracks are shorter and transitions come faster, pushes past 25 an hour. Our set length calculator does this maths properly for your genre and mixing style, and it's the right first stop if you haven't run the numbers yet.

But that number describes a hypothetical library, not yours. It tells you how many tracks a 3-hour tech house set needs in the abstract — somewhere around 60, give or take your buffer. It says nothing about whether the 60 tracks sitting in your folders can actually do the job. Sixty tracks that are all mid-tempo, mid-energy, four-to-the-floor fillers at 124 BPM isn't a 3-hour set. It's one very long segment of a set, repeated until the clock runs out. The calculator answers “how many.” It has no way of answering “which ones, and are they the right shape.”

What “big enough” actually means

A library can hit exactly the right total track count and still fail a long set, in three specific ways.

Not enough depth in your core genre. Sixty tracks that satisfy a genre filter isn't the same as sixty tracks that satisfy a chemistry check against each other. If forty of them cluster around the same three artists, the same label sound, the same handful of chord progressions, you'll run out of genuinely different-feeling transitions long before you run out of tracks — the back half of the set starts repeating itself, even though nothing on paper is a duplicate.

No coverage at the ends of the energy arc. Most DJs stockpile the middle — peak-time tracks are exciting to buy and easy to find, so everyone's crate has plenty. True openers, the kind that hold a low-energy room without demanding attention, and true closers, the kind that land a set rather than just stopping it, are the two spots almost nobody's crate is stocked for. That's also the part of a set with the least room for error. Our energy flow guide covers why the ends of the arc carry more structural weight than the middle, if you want the fuller theory before you start auditing your own crate.

No bridge tracks between zones. A 3-hour slot rarely stays in one genre pocket the whole way through. Most DJs cross at least two adjacent zones over the course of a night — house into tech house, tech house into peak-time techno, melodic into driving. If every track you own sits solidly in the middle of one zone and none of them sit on the edge of two, you don't have a bridge — you have a wall, and the join between zones is exactly where sets fall apart live.

Run the audit on your own crate

This is the part that tells you something, and it only works if you've already imported your Rekordbox collection into Mixgraph. Open your Library and work through it in three passes.

Start with the top-line numbers. Your Library shows a matched count and a total count — matched being tracks Mixgraph has resolved against its catalogue, total being everything you imported. Note both; you'll need the split again in a moment.

Then narrow to the window your set actually lives in. If your Rekordbox crates are already split by genre or intent, use the playlist filter to pull up the right one. Either way, work through the list and use the BPM and key Mixgraph shows on every row to count what actually sits in your range — for a 3-hour tech house slot running roughly 122 to 128 BPM, that's the number you're after. This is your real, usable pool for the night — not your whole collection, which is doing you no favours if half of it is 140 BPM techno you'll never touch in this set.

Next, work through that same pool by ear and sort it into four rough bands: opener, build, peak, closer. This is the one pass Mixgraph can't do for you — don't overthink the boundaries, you're looking for shape, not precision. A healthy 3-hour pool typically wants something like 8–10 tracks that could open, 20–25 that sit in the build, 10–12 genuine peak-time tracks, and 6–8 that could close. Count what you actually have in each band and write the numbers down.

The gaps show up immediately once the bands are in front of you. Most DJs auditing their crate for the first time find a healthy peak band, a workable build, and almost nothing at the two ends — three or four openers when they need eight, one or two real closers when they need six. That's not a track-count problem. The calculator would tell you your sixty tracks are plenty. The band count tells you where those sixty tracks actually sit, which is the thing that matters.

The variable the track-count formula ignores: matched versus unmatched

Here's the bit most DJs don't think to check, and it changes how much buffer a library-built set needs. When you import your library, every track gets matched against Mixgraph's catalogue where possible. A matched track gets full chemistry — harmonic, rhythmic, energy, groove, mood, and vocal, the whole picture, same as any track already in the catalogue. An unmatched track — a rare edit, a promo, something Mixgraph hasn't seen before — still gets scored, but only on BPM and key. That's a real and useful signal, but it's a narrower one, and it carries a lower confidence for exactly that reason.

In the safe middle of a set, that difference barely matters. If an unmatched track's BPM-and-key chemistry with its neighbours looks solid, it'll almost certainly play fine — you've got tracks either side absorbing any surprise, and a slightly-off groove or mood in the middle of a build is forgivable. At the structural pinch points, it matters a lot more. Your opening transition, the swap into your peak sequence, and your final transition into the closer are the three moments in a set with the least margin for error — get one wrong there and there's no next track to quietly correct course with. An unmatched track sitting in one of those exact slots is a bigger gamble than the same unmatched track sitting in the middle of your build, purely because of where it is, not what it is.

The real buffer math for a library-heavy set

The standard advice is to prepare 25–30% more tracks than you think you need — if your set needs 50, have 63–65 ready. That's sound advice and it still applies. But if a meaningful share of your pool is unmatched, the buffer needs a second layer: the spares shouldn't be spread evenly across the night, they need to sit specifically at the three or four structural moments where a wrong call costs you.

Take a concrete example. You're playing a 3-hour peak-time slot, tech house, running at roughly 20 tracks an hour — that's 60 tracks for the set itself. Add the standard 25–30% buffer and you're prepping around 75 tracks total. Say a third of that prepped pool — about 25 tracks — turns out to be unmatched imports once you check the split. Sounds like plenty of margin either way. But if you haven't checked where those 25 sit, you might find they're concentrated right where you need them least: your opener, your first peak swap, and your closer are all unmatched tracks with no matched alternative sitting next to them. The fix isn't more tracks overall — it's 2–3 confirmed, matched alternatives specifically at those three moments, so an unmatched track never has to carry a pinch point on its own.

The gaps DJs hit most often

A few patterns show up again and again once people run this audit for real.

• The everything-at-125 hoard. A crate stacked with tracks clustered within 2–3 BPM of each other, almost all at peak energy, with no true opener anywhere near the bottom of the range. It plays like a great 90 minutes, twice, with nowhere to start and nowhere to land.
• The energy floor that doesn't exist. No tracks below a certain energy level at all — everything in the collection was bought to be played loud, to a full room, at full attention. Fine for the middle two hours, a real problem for the first twenty minutes.
• The single deep pocket. One genre or sub-genre with real depth — forty or fifty strong tracks — and nothing that sits on the edge of an adjacent style to bridge into for the back half of the night. Great first 90 minutes, then a hard stop where the set needs to evolve and can't.

Closing the gaps before the gig

Once the audit has named a specific gap, closing it is a targeted job, not a general “go buy more music” instruction. If your audit turned up a shape — say your build and peak are strong but your opener and closer bands are thin — the Energy Arc Planner will show you what a properly proportioned version of that shape looks like across genre and energy, so you know exactly what you're filling before you go looking.

From there: search the catalogue directly for tracks that sit in the specific band you're short on — don't browse broadly, filter to the exact energy and BPM window the gap sits in. Shortlist candidates into Favourites as you find them, so you're not re-covering the same ground twice. And if a load-bearing slot — your opener, your closer, a peak swap — only has an unmatched track to fill it, it's usually worth requesting the track properly rather than trusting the confidence-discounted version in a spot where you can't afford it to misbehave. A thin substitute in the safe middle of a set is a minor risk. The same substitute in your closing slot is the whole night's last impression.

From audit to actual set order

Once the gaps are filled and you've got real coverage across opener, build, peak, and closer, the audit stops being a spreadsheet of counts and becomes an actual running order. In Flow Builder, start from the Marathon Flow formation — it's built as an extended-journey starting template, a useful skeleton for a long set rather than a slot-length-specific fit. From there, switch into Freeform mode, which adds slots as you add tracks rather than locking you to a fixed count, and grow the flow out to the real number your audit called for. The bands you labelled — opener, build, peak, closer — become the actual sequence, track by track, instead of a set of numbers on a page. That's the difference between knowing your library is big enough and knowing your set is actually built.