Pitch Practical Effects, VFX, or a Hybrid Without Choosing by Habit
Pitch Practical Effects, VFX, or a Hybrid Without Choosing by Habit
A treatment that says “we’ll do it practically, with a little VFX polish” has proposed a method before anyone has said what the method has to accomplish. The sentence sounds decisive, but it commits nothing: it doesn’t tell the producer what to protect on the shoot day, it doesn’t tell the effects team what to build, and it doesn’t tell the director which part of the image they’re still free to change.
The useful version of the decision is smaller and harder. Name the moment the audience has to believe, say what has to exist in front of the camera for that moment to be photographed, and say what can be made afterwards. Then the choice between physical work, post work, and a split between them stops being a matter of taste and becomes a consequence of the shot.
Define the effect before sorting its methods
Three questions do most of the work, and none of them is “practical or digital.”
What must the performer respond to? An actor reacting to something that is really there has a different job than an actor reacting to a timing cue. Neither is automatically better. A real object can give a performance a genuine target; it can also give the actor something unpredictable to chase, take after take. A performer working against a mark can be precise and repeatable, but the precision is now a task they have to perform rather than a gift the set hands them.
What must the camera see? Not “does the shot look real,” but which elements share the frame and which must be in the same photograph at the same instant. Two things joined across an edit are a much easier problem than two things joined inside a single unbroken shot. This one question will reshape the comparison more than any other.
Which changes need control or repetition? Some effects need to vary — a natural, unrepeatable fall reads as real precisely because no two takes match. Others need to land on a music cue, a cut, or a specific frame. Those requirements pull against each other, and a method that serves one may cost you the other.
The vocabulary is worth keeping straight because a muddled pitch invites a muddled bid. Physical effects — sometimes called practical effects — are made and photographed in the scene: a real object, a real fall, a real burst of something. Visual effects is the broader term for work done to the image after capture: cleanup, rig removal, compositing, and computer-generated elements among them. CGI is one kind of visual-effects work, not a synonym for all of it. Some of the most heavily post-produced shots in a commercial contain no computer-generated imagery at all — just a lot of patient erasing. And most shoots labeled practical contain some post work, because something holding, guiding, or hiding an object was standing in frame.
That last point matters for the pitch. If the physical arrangement that makes an effect happen is visible, removing it is visual effects work. Calling the result “practical” doesn’t make it so; it just keeps the removal off the page, which is a good way to be surprised by it later.
Compare what happens at the point of contact
Take the scene this article keeps returning to. It’s invented, and it’s deliberately ordinary: a gift is being handed over. The ribbon comes free of the package, curls through the air toward the recipient’s raised hand, and settles across the palm. Two things have to be true at once. The curl’s path and timing have to be precise enough to land on a specific beat in the track. The landing has to read as contact — something with weight arriving in an open hand, the ribbon settling, the fingers closing a beat later.
As written, the treatment asks for all of it in one unbroken shot, from the moment the ribbon leaves the box to the moment it touches the palm. Hold that assumption for now; it will turn out to be the hinge.
A physical-performance route. A real ribbon travels the whole distance. Something has to make it leave the box and follow a path — an off-camera arrangement, a movement built into the performer’s own action, or some combination. The recipient’s hand is in a frame with a real object, so the actor genuinely tracks it and genuinely closes on it. The path, though, is not fully yours. It will vary take to take, and the take you fall in love with is the take you now have to match if the ribbon appears again in a later setup in the same curled, kinked state. Also, if the guiding arrangement is visible in frame, that visibility has to be removed later. This is not a fully practical route. It is a practical route with a post component, and the treatment should say so.
A digitally represented ribbon. Nothing is in the air on set. The recipient performs to a reference — a mark the hand has to reach, a timing cue, a position the palm must hold at a particular frame. The film’s geometry has to be decided before the shoot, because the animation will be built to land where the hand was. Then post produces the ribbon: its path is exactly controllable and exactly repeatable, adjustable by frames, and free of the physics of an actual ribbon. The demands move rather than disappear. The animation has to match the plate’s lighting, the hand’s micro-movement, the camera’s motion, and whatever grain, blur, and texture the rest of the shot carries. And it has to arrive at the palm at the precise instant the palm is there.
A hybrid. Real ribbon for the contact, digitally developed movement for the intervening travel. The recipient’s hand closes on something with weight and drag, which is the requirement that made this scene worth comparing in the first place. The travel is built separately, where it can be timed exactly to the beat. The cost is a join — the point where the intervening movement becomes the real object — and that join is the whole argument.
Notice what happens to the performance in each. In the physical route the actor has a real target, but the target is unpredictable, which can be wonderful or unusable depending on the take. In the digital route the actor has a cue, not an object, and the performance carries the burden of making an absent thing feel present. In the hybrid the actor has a real target only at the end, so most of the reaction is performed against nothing and the last moment is real. None of these guarantees a better performance. A real ribbon in frame is a fact about the set, not a fact about the acting.
Notice also what doesn’t disappear. Adding something later never removes the need to plan its interaction. The ribbon has to land in the palm; the palm’s position is the actor’s; so either the actor is constrained to a mark the animation will hit, or the animation is built to follow whatever the actor did. Someone decides which, and that decision is usually made weeks before the shoot, in a conversation the treatment can either anticipate or discover later.
Compare control and the burden of verification
Follow one required change — the landing on the beat — through each route and the trade-offs stop looking like a preference.
Timing. In the physical route, the beat is not something you command; it’s something you shoot toward and then choose. Take eleven might land near the beat, and a retime does the rest. An edit would absorb the timing too, but an edit is exactly the staging assumption this version of the shot has set aside. In the digital route, the animation is built to the hand, so the path can be timed to the frame. The cost is the reverse side of that: the beat has to be settled before the shoot rather than found in the edit, and the hand has to reach its mark on it — a frame-accurate obligation on the performance that the physical route never places on it. In the hybrid, the travel is timed exactly and the contact is real, which means the join has to sit somewhere that hides the transition — and whether anywhere hides it is exactly the thing you cannot know from an armchair.
Repeatability. Digital work is endlessly revisable, which is a real advantage and a real trap: revisions need approvals, and each one reopens a shot that was finished. Physical work is done once, for good or ill, on the day. If the director wants to keep adjusting the curl after seeing an edit, that preference argues for post. If they want the thing to be over, that argues for the set.
Continuity. Here is where the hybrid earns its scrutiny. A ribbon that is real at the palm and generated in the air is one object in two states. Its width, its sheen, how tightly it’s curled, whether it looks kinked from being handled — all of it has to match across the join. If the join falls on a cut, that cut is a free repair: the audience has already accepted a new setup, and the mismatch has somewhere to live. If the join falls inside the unbroken shot the treatment wanted, the swap has to happen in front of the audience, concealed by whatever is moving at that instant — the ribbon itself, the hand entering frame, a fast turn. Concealment like that is a question for a test, not a claim for a pitch.
Which brings the hinge back into view. An edit is a free join; an unbroken shot is not. If the release, the travel, and the landing share one continuous take, the hybrid’s difficulty moves from a place you can cut around to a place you have to hide. That single line in the staging changes the relative difficulty of every route in the comparison, and no amount of adjectives about authenticity or flexibility substitutes for noticing it.
The verification burden follows from the same place, and it takes a different form in each route. Before the hybrid can be claimed, someone has to demonstrate the join: a representative shot, the join placed at the frame being proposed, the ribbon’s material and state matched across it, and a look at whether the result survives being watched in motion. The physical route’s own test is the removal — whether whatever moved the ribbon can be taken out of the frame without leaving evidence in the plate. The digital route’s is the integration — whether a generated ribbon can be made to match the plate’s lighting, the hand’s micro-movement, and the camera’s motion well enough to land in a real hand.
Nobody in this article has run any of those tests. No mechanism, rig, animation, specialist estimate, or trial has been produced here, and no published breakdown was consulted. That matters, because a finished breakdown documents the route that was used. It does not establish what the rejected route would have cost or whether it would have worked, and a before-and-after reel is not a comparison of alternatives — it’s an advertisement for the one that happened.
Choose conditionally and retain a real alternative
So the treatment can commit to a preference while stating the question that would overturn it, which is more useful than a guarantee it can’t keep.
Preferred candidate: the hybrid. The reason is tied to the action, not to the technology. The two requirements pull in opposite directions. The curl needs exact, repeatable path and timing — that favors building the movement in post. The landing needs tangible contact in an open hand — that favors a real ribbon at the palm. A split honors both, and the split is honest about its own cost: a join.
Main uncertainty: whether the movement-to-contact join preserves the action inside the shot as written. Not “does it look real” in general — whether the specific join, at the specific frame, in the specific unbroken shot, survives being watched at speed.
Condition for switching. If the join fails when tested, the choice becomes which requirement to bend. Keep the unbroken shot and give up some of the contact’s physicality, and a fully digital ribbon becomes the leading candidate, with the palm’s performance now carrying the landing. Keep the tangible contact and accept an edit at the join, and the hybrid’s problem largely evaporates because the cut absorbs it. Keep both and go fully physical, and the path’s precision becomes the thing under strain, with rig removal waiting at the end of the shoot either way.
That is the shape a treatment can responsibly take. Something like: We see the ribbon come free in one unbroken shot? Then we’re proposing a real ribbon at the palm, movement built to it, and we need to test the join before we lock the staging. If the test fails, we would rather move to a cut at the join than lose the contact.
What that sentence does is separate the creative preference from the feasibility claim. It says what the shot is for, names the dependency it creates, and asks a question a producer or effects specialist can actually answer.
The next step is not a longer list of methods. It’s narrow: get a specialist account that discusses more than one route for a single effect and says why the rejected options were rejected, then design the smallest demonstration that answers the question that would change the choice — a representative shot, the proposed join, watched in motion.
The decision note, in the end, is four lines: preferred candidate, the reason tied to the action, the uncertainty that remains, and the condition that would flip it. Everything else in the pitch is either craft or furniture.
Frequently asked questions
Why is 'we'll do it practically, with a little VFX polish' an incomplete method decision?
It names a method before saying what the method must accomplish. It does not tell the producer what to protect on shoot day, the effects team what to build, or the director what can still change.
What three questions should define an effect before choosing a route?
What must the performer respond to? What must the camera actually see in the same frame or instant? Which changes need control or repetition? The answers reshape the method comparison.
In the ribbon example, what are the trade-offs among physical, digital, and hybrid routes?
A physical ribbon gives real contact but an unpredictable path and possible rig removal. A digital ribbon gives exact path and repeatable timing but the performer works to a cue, and the animation must match lighting, hand movement, camera motion, grain, and the exact instant of arrival. A hybrid uses real ribbon for contact and generated movement for travel, at the cost of a join.
Why does an unbroken shot change the difficulty of a hybrid?
An edit is a free join; an unbroken shot is not. If the join falls on a cut, the audience has already accepted a new setup. If the release, travel, and landing share one take, the swap has to be hidden in front of the audience by whatever is moving at that instant.
What should a conditional recommendation include, and what verification does each route need?
A recommendation can state the preferred candidate, the reason tied to the action, the uncertainty that remains, and the condition that would flip it. Verification differs by route: the hybrid needs a demonstration that the join at the proposed frame preserves the action in motion, the physical route needs a removal test, and the digital route needs an integration test.