Guide · Director craft

How to Design a PMV Intensity Arc

You usually find the problem one drop too late. The first chorus already holds everything you were excited to use. The close-ups. The rare clip. The fastest cuts. Then the second half has nowhere to go.

Cutting faster does not fix it. You spent the reveal, the motion, the close framing, and the rare footage all at once. The guy watching is already at the peak you meant for the last minute. After that he is just waiting, or he closes the tab.

An intensity arc is the plan that stops this. It is not one line labeled intensity. It is a set of phase contracts. Each phase gets one job and its own working ranges. You write those contracts before the timeline fills.

This guide is the board you use tonight.

Mark the song as sections, not beats

Start from a verified beat grid. Then zoom out until the song is sections, not beats.

A grid tells you where the pulse lives. That is how you sync cuts to the music. An arc needs a bigger map. Mark every audible turn that changes what you have to direct. Drums enter. The vocal strips back. The hook returns. The bass drops out. The final run begins.

Say what changed, not just the section name. "Drums entering" tells you what you have to do. "Chorus" only names the part. Final Cut Pro's own docs draw this same line: song parts versus bars and beats. The planning idea is the same in any editor.

First board is four columns: start, end, audible boundary, working section name. No intensity yet. Prove the boundaries belong to the music first. If you cannot hear why a line is there, it is not a phase yet.

Markers keep that map visible while you cut. Premiere lets you add markers to clips. Final Cut treats song parts as a different layer from bars and beats. Use those marks as section marks, not as a cut list. You are still zoomed out.

Give each phase one job

Once the song is sections, give each phase one job. A verb. What must be different by the end of the phase.

Working verbs, not official types:

• Orient. Establish the visual world and the first pattern. He should know what kind of video this is.
• Invite. Deepen without spending the rarest material. Closer, hotter, still holding something back.
• Accelerate. Raise the session without closing the argument.
• Withhold. Stay charged while refusing the expected escalation. He wants the next step. You do not give it yet.
• Reset. Open contrast for another run. Change the rules without emptying the screen.
• Return. Bring a pattern back changed. Same motif, new pressure.
• Pay off. Spend what the earlier phases protected.

Do not let the song label choose the job. A chorus can withhold. A breakdown can keep explicit footage on screen. A final verse can pay off a promise the first drop left open.

Verse equals low and chorus equals high produces a diagram of the song. It does not produce a direction for the session. The guy stroking is not tracking verse and chorus. He is feeling whether you are still in charge of him.

Finish this sentence for every phase: by the end of this phase, the edit has ______. "Gotten more intense" is unfinished. "Introduced the close-framing pattern without using the final reveal" is something you can test.

If two neighboring phases have the same job, merge them or change one. A board full of accelerate is how you peak early with extra steps.

Four controls, not one intensity slider

This is the heart of the board. Four independent controls per phase.

Cut cadence. The phase's normal shot-length range in beats. Mostly four-beat holds. A two-beat run at the exit. Cadence is how often the picture changes, not how hard the picture hits. Long holds can stay filthy if the footage is already working. Fast cuts of mild clips can feel busy and cold.

Visual motion. What is happening inside the footage. Low, working, high. These are comparisons inside this edit, not a global scale. A still face can be low here. A thrusting close-up can be high. Another video's scale is not your scale. Motion is the body in the frame: a stroke, a ride, a tongue, a hip that finally commits.

Framing and reveal. Proximity, detail, explicitness, and new information. Track them together only when they serve one decision. Explicitness is one control, not the arc itself. If every escalation means showing more cock, more hole, more cum, the edit runs out of direction at the most explicit clip. Then you have nowhere to take him except the same money shot again.

Rarity. How much of your strongest, least repeated material the phase may spend. No A-tier. One anchor. Spend one, reserve two. Final material available. A-tier is the clip you would miss if it showed up twice. Treat it like that.

A tagged clip library feeds this. Energy and framing tags tell you what can serve the contract. If you are still gathering footage, find clips with those tags in mind, not just "this is hot." Hot does not help you later. Tags do, when the phase says working motion, no A-tier, medium frame.

Four controls expose the choice hidden inside "make it build." Raise one. Hold another. Pull a third back. Save the fourth.

That is how a withhold stays sexy. Explicitness holds. Framing simplifies. Cadence lengthens. Rarity spends nothing. He is still looking at sex. You just stopped feeding him the next gear.

Place the anchors first

Do not fill the timeline left to right. Place the anchors first.

The first unmistakable statement of the visual idea. The first meaningful reveal. The first false peak. The deepest reset. The return of a pattern with a change. The final payoff. The final readable image.

An anchor is a promise about what the viewer perceives at that point. It is not automatically a cut. If the first peak holds all three of your best anchors, move at least one. The ending needs something the opening did not spend.

This is the part that feels like teasing because it is. You decide when he gets the thing he already wants. You can hint. You can let a motif show up in a wider frame. You do not dump the whole bag in the first chorus.

He finishes to the video that still has a last card. He closes the one that already played it.

A worked board for a 3:04 track

Here is a complete board for a 3:04 track at 120 BPM in 4/4. One beat is half a second. One bar is two seconds.

Read what the board is doing, not just the times.

The first peak raises motion, tightens framing, spends one rare clip, and permits a short run. It refuses to max every lane. That is why the second half still has a job.

The withhold does not go non-erotic. Explicit material stays. Cuts lengthen. Framing simplifies. The return has something visible to restore: tighter framing on a motif he already knows. He feels the pattern come back, and it lands harder because you took the tight frame away for sixteen seconds.

The 48-second payoff varies inside its range and saves the fastest run for the final arrival. Reserved reveals arrive with space between them, so each one still registers. If you stack both leftover A-tier clips on the same four bars, you spent them as one event.

The last eight seconds do not introduce anything. One readable image. Let him finish on something his eye can hold, not on another cut.

And the board never says which beat gets a cut. Macro direction is set. Micro selection stays open. That is the work when you place the actual edits: obey the phase contract, then pick the frame.

Test the rough cut three ways

Once you have a rough cut, test it three ways before you polish.

The silhouette test. Draw the four lanes as rough steps. Cadence, motion, framing, rarity. Every line in the same band means flat treatment. All four climbing together means the first half is eating the second half's options. You want a messy little drawing, not a pretty diagonal.

The mute test. Watch without sound. The treatment change should read at the boundaries. If you cannot see that accelerate started, the picture is not doing the job. The music is covering for you.

The audio-only test. Listen while reading the phase jobs. Does accelerate begin where the arrangement gathers? If a job changes mid-section, move the boundary or write a better reason. Sometimes the song fooled you. Believe the arrangement.

These tests are cheap. They catch the usual lie: the song feels like it builds, so you assume the edit does too.

Repair the board, not the symptom

When the rough cut fails, do not reach for random faster cuts. Repair the contract.

The edit is flat. Find the lane that never changes. If every shot is four beats, change cadence in one phase. If every clip is the same distance from the body, change framing. Faster everywhere is still flat.

The first drop beats the ending. Check rarity and reveal before cadence. Move one A-tier anchor later. The first peak should make a statement without closing the argument. If he already saw the rarest body, the tightest frame, and the fastest run, the last chorus is a rerun.

The quiet section dies. You reduced every lane at once. Restore one source of charge. Explicitness. Eye contact. Internal motion. Close framing. A repeated motif. A reset changes the rules without emptying the screen. He can still be hard during a hold. He just should not be bored.

The graph rises perfectly and the video feels mechanical. Break the diagonal. Hold a phase. Let one control fall while another rises. Let framing open while motion stays high. Let cadence slow while the footage gets more explicit.

If you are assembling a first video from scratch, pair this board with how to make a PMV. The how-to gets you a cut. The arc keeps that cut from spending itself by the first chorus. For the psychology underneath the session, see why PMVs work.

This planning is more coordination than it should be. You are holding a grid, a clip library, four lanes, and a set of promises in your head at once. That friction is part of why we are building Kira. It cannot lay this board down for you today, though. The method still works now, in the editor you have open tonight. Write the phases. Place the anchors. Spend the last rare clip last.