THE AIRCRAFT. THE APP. THE ART OF A PLANNED SHOT.
NEMESIS.
THE SKY,
CHOREOGRAPHED.
My DJI Mavic 2 Pro, the Litchi app and twelve ways to turn a carefully planned flight into a film worth watching.
By John “NIX” Nickolls • 35 minute feature • Creative scenarios + practical flight planning
DIRECT THE MOMENT.
A PROPER CAMERA IN A FOLDING AIRCRAFT
Some machines earn their place.
There is a particular pleasure in a piece of equipment that still makes you want to get outside and create something. For me, my Mavic 2 Pro has a name: Nemesis. It has been part of my kit since 13 May 2022, and this is a feature about what I would like to make possible when that aircraft meets a properly considered Litchi plan.
I am interested in the moment a camera move starts to feel deliberate. The van becomes smaller and the landscape opens. A shoreline separates into layers. A building stays in the frame while the countryside moves behind it. The flight becomes part of the storytelling instead of something the audience has to politely sit through.
DJI introduced the Mavic 2 Pro in August 2018, with its Hasselblad partnership and a camera aimed squarely at photographers and filmmakers. That history matters to me: technology can move on without every older tool suddenly losing its purpose. A good composition does not check the release date before it works.[2]
Nemesis has a 20-megapixel, 1-inch-type sensor; a 28mm-equivalent lens with an f/2.8–f/11 aperture; 4K recording up to 30fps; DNG RAW stills; and 10-bit D-Log M or HLG video options. The aircraft weighs 907g. DJI’s quoted 31-minute flight time comes from a controlled, no-wind test at 25km/h. It is a specification, not my mission budget.[1]
What makes Litchi interesting is the possibility of treating that camera as a repeatable filmmaking tool. I can think about a beginning, a middle and an end before the propellers turn. I can decide where the subject belongs in the frame and how quickly the background should reveal itself. Then I can ask whether the real location actually allows the shot.
“Litchi can do the choreography.
I still have to direct the film.”
This is my creative notebook for Nemesis: proposed shots, the thinking behind them and the practical choices that turn an idea into a workable plan. The illustrations bring those ideas to life; the diagrams explain the decisions I would want to understand before taking off.
START WITH THE RIGHT APP
One aircraft. Three names worth separating.
The Litchi family has grown, and an old tutorial can now send you in the wrong direction. The current compatibility list places the Mavic 2 Pro in Litchi for DJI drones, the original iOS and Android flight app. Litchi Pilot is a different application for other supported aircraft; it is not the app I would choose for Nemesis.[3]
Litchi for DJI drones
The supported original app for the Mavic 2 Pro. This is the name to verify when selecting the software that will connect to Nemesis.
Official Litchi site ↗Litchi Hub
The browser workspace for preparing and previewing a flight. I would use the larger canvas of my MacBook to think through the shot.
Open Litchi Hub ↗Litchi Pilot
Its name appears in newer guides, but its aircraft support differs. I would not assume a feature demonstrated on a newer drone transfers to the Mavic 2 Pro.
Check compatibility ↗The current Hub offers browser-based planning, 2D and 3D previews, camera projection and mapping tools. Its selected flight application determines the available functions. I would set up the project for Litchi for DJI drones and Mavic 2 Pro, then judge the features available in that configuration.[5]
Litchi’s own team confirms that the newer Hub supports the Mavic 2 Pro. Some newer waypoint components, including integrated orbit and spiral components intended for newer aircraft, do not carry across. That is distinct from the original app’s separate Orbit mode. The browser Hub also covers much of what older tutorials achieved with the separate, third-party Virtual Litchi Mission program.[6]
For me, the workflow begins at a desk. A proper screen, an unhurried look at the landscape and a cup of tea are useful planning equipment. At the flying site, I want to be confirming a simple plan, not building a complicated one while the light disappears and the battery sits powered up.
I would check the controller, cable, phone, operating system and exact app version together. The App Store listing is useful evidence of current app requirements; it is not proof that every newly released iPhone and iOS combination has been flight-tested. My iPhone 17 Pro Max would earn its place in the setup through a ground connection check and a short, uncomplicated test flight.[9]
A planned route with a considered camera view.
Framing assistance while the pilot controls movement.
A circling shot, when the whole arc is appropriate.
A coordinated sequence of overlapping views.
Position-based following using the mobile device.
Visual tracking of a selected image subject.
High-level original-app modes; availability and details depend on the aircraft and settings. Follow and visual Track use different information. Neither replaces route assessment or supervision.[8][9]
FROM A PIN ON A MAP TO A SHOT IN A FILM
Plan the camera move before drawing the route.
A row of waypoints can be technically tidy and visually dull. I would begin with a sentence describing the result: “Start with Vanilla as the subject, then slowly reveal the countryside around her.” That tells me more than placing eight pins and wondering what they might produce.
Next I would choose the start frame and the end frame. Is the camera looking across the scene or down into it? Does the subject stay in roughly the same place on screen? Will the movement reveal useful depth? Is there a clean beginning and ending that an editor can actually use?
The key distinction is between where the aircraft travels and where the camera looks. A sideways move can keep looking at the same subject. A retreat can widen the context while the camera continues pointing towards the beginning of the shot. That separation is where a simple flight starts to acquire a visual purpose.
A useful way to organise the plan is to give every setting a job. Position answers where. Height answers at what level. Speed answers at what pace. Heading and camera tilt answer what the viewer sees. Capture settings answer what I bring home.
I would use the smallest number of points that expresses the shot cleanly. More pins do not automatically make a better move. They can make the route harder to inspect, the camera behaviour harder to predict and the job of finding an error unnecessarily fiddly.
| Decision | The useful question | What I would write down |
|---|---|---|
| Purpose | What should the viewer notice? | One subject, one change in perspective, one clear finish. |
| Launch reference | Where will the height reference actually begin? | A repeatable, lawful launch position and the terrain around it. |
| Flight corridor | Can I see and assess the complete movement? | Obstacles, people, boundaries and a clear route back. |
| Camera story | What belongs in the first and last frame? | Subject position, direction of light, heading and tilt concept. |
| Time | How much airborne time does the whole job need? | Take-off, positioning, recording, return, landing and contingency. |
| Stop decision | What would make me abandon this shot? | Unexpected people, wind, poor visibility, warnings or an uncertain route. |
The practical MacBook-to-field routine
I would give the project a useful name—something like Vanilla meadow reveal / short / autumn light—and save a separate version whenever I change the geometry substantially. The name should remind me what the mission is for, not merely record that I once clicked “Save”.
In Hub, I would inspect both the overhead route and the camera preview. The 3D view is particularly useful for asking whether a foreground feature will block a subject or whether a camera angle makes visual sense. I would then verify that the intended flight is available in the flight app before travelling.[5]
At the location, the ground gets the casting vote. Maps and previews help me prepare; actual wires, growing trees, livestock, walkers and the day’s weather decide whether the plan still works. I would inspect the launch-to-first-point segment as carefully as the attractive middle of the shot. Getting to the opening frame is still flying.
My first test would be short and uncomplicated. I would check the framing, inspect the resulting clip, confirm that stopping and regaining control behave as expected, and only then consider a more involved version. Finding a problem in a twenty-second experiment is considerably better than discovering it halfway through a cinematic masterpiece with nowhere sensible to go.
THE NUMBER THAT DESERVES YOUR ATTENTION
Forty metres above what?
Litchi waypoint height normally references the launch elevation. A mission showing +40m does not necessarily remain 40m above whatever ground passes underneath it. Terrain-assisted planning uses elevation information; it is not a live scan of the trees, wires and structures along the route.[4]
Here is an intentionally simple example. Imagine that the launch ground is 100m above sea level. Add a nominal 40m mission height and the aircraft’s planned elevation is 140m above sea level. Over ground at 100m, the arithmetic gives 40m clearance. Over ground at 125m, only 15m remains. Ground reaching 145m would intersect that flight plane.
The mistake is easy to picture: I draw a lovely line on a map, type a reassuring number into the altitude box and mentally turn it into a protective bubble. The map has not agreed to that arrangement. A hillside, tall tree or overhead cable has no interest in what I intended.
I would also inspect the space between the waypoints. Checking only the points is rather like checking that both ends of a footbridge are sound and declining to look at the middle. The route, its turns, its approach and its return all need to fit the real three-dimensional space.
There is a separate legal issue when terrain falls away. In UK Open-category flying, the limit is 120m from the nearest point of the earth’s surface, with terrain taken into account. A take-off-relative display alone cannot establish that the entire route complies.[11]
If I cannot confidently explain the route’s height relative to the actual landscape and its obstacles, I have not finished planning it. A handsome 3D preview is a rehearsal aid, not a site survey.
SMOOTH VIDEO AND STOPPED PHOTOGRAPHS NEED DIFFERENT PLANS
Make the route serve the capture.
With the original app, curved turns ignore waypoint actions. Straight-line missions can stop at a point and carry out an action; a curved cinematic pass should not depend on a stop-point photo or record command firing.[4]
My preference for a simple video move would be to start recording deliberately before the useful movement, verify that recording is actually running, and leave a little time at either end for the edit. For a series of still viewpoints, I would design the job around arriving, settling, composing and capturing. Those are different photographic intentions.
Camera tilt deserves the same thought. Focus POI aims the gimbal towards a selected subject and needs the controller link. Interpolate transitions between chosen camera angles; the optional DJI onboard gimbal-interpolation setting supports that mode, not Focus POI.[4]
For a reveal, I would ask whether the camera should remain anchored on the van or slowly hand attention to the horizon. Both can work. A POI-centred version says, “Here is the subject in its setting.” A gradually changing camera angle can say, “Now look at the place around it.” The choice should come from the story.
I would watch the first few seconds especially carefully. A dramatic yaw correction before the aircraft has settled can make a refined route begin with the visual equivalent of someone straightening a wonky picture frame. Starting from a sensible position and leaving an editing handle can make the finished film feel much more composed.
INTERACTIVE SHOT LAB
A short route can make a big picture.
Explore four simple camera-move concepts. This is an editorial animation and a distance ÷ speed calculation, with no real coordinates, terrain data or connection to a drone.
Orange = travel · cyan = camera attention. The map is invented; the timing arithmetic is exact for the stated simplified model.
CAMERA-MOVE STUDY
The reveal
A 60m retreat gives the subject a progressively wider setting. Keep the planned route visible and clear throughout.
A timing illustration, not a recommended flight speed.
Excludes acceleration, approach, turns that change speed, hovering, return and landing. It is not an endurance or safety calculator.
Animation plays at 3× the nominal travel speed.
At 2m/s, a 60m straight movement takes 30 seconds before acceleration and other allowances. That is already plenty of material for a short establishing shot. I would usually prefer one purposeful move with a clear visual payoff to a long tour that asks the viewer to invent the reason for watching.
The lesson is not that every route should be sixty metres long. It is that the amount of travelling and the amount of storytelling are different quantities. A small change in perspective can reveal a lot when the foreground, subject and background have been chosen well.
THE NIX MISSION NOTEBOOK
Twelve scenarios.
One reason to fly: the picture.
These are creative storyboards, not ready-to-upload missions. The figures are illustrative starting points for discussion. Every stated height is relative to the launch point unless explicitly described otherwise. They do not establish ground clearance, legal separation or suitability at a particular location.
For each idea, I would choose an operating category and location that actually permit the flight, assess the entire corridor, retain visual line of sight and plan a clear return. Some scenes would need an appropriate A2 operation or other authorisation; an attractive location is not automatically available to a legacy Mavic.
REVEAL.
Vanilla, then the whole world.
The opening frame belongs to the van. The ending frame explains why she is there.
I would choose suitable isolated open land, park Vanilla where her orange details catch the light, and design a short retreat that gently rises. A 40–60m movement at roughly 2m/s, rising from an illustrative +30m to +45m, gives me something useful to evaluate before adapting it to a real site.
The picture would begin with the van comfortably readable, then allow more landscape into the composition. I would keep her slightly away from dead centre so the reveal has somewhere to develop. The important thing is the changing relationship between subject and place.
The deciding check: the complete backward route, the launch reference and the people around it. A campsite or public parking area is not automatically an A3 location. The van stays parked and I stay focused on flying.
Vanilla gets the opening credit. The landscape gets the standing ovation.
The headland that unfolds.
A quiet sideways move can reveal more than a dramatic charge towards the sea.
I would look for two layers of coastline visible from a lawful flying position. A short, shore-parallel movement could allow the nearer feature to slide across the background and reveal the next bay. As a storyboard, 60–100m at about 2–3m/s is enough to explore the effect; the site decides the actual distances.
I would keep the camera’s attention on the shape of the coast and let the changing foreground do the work. A few seconds before the reveal and a few afterwards give the edit room to breathe.
The deciding check: cliff winds, wildlife, airspace, recreational-area restrictions and an aircraft that remains directly visible. I would not send it around the hidden side of the headland or turn a short coastal shot into an ambitious offshore journey.
The coast opens like a page, one layer at a time.
Old stone. A fresh perspective.
Let a structure hold the frame while its surroundings quietly rearrange themselves.
For a suitable isolated building or ruin, I would consider a wide partial arc—perhaps 60–90 degrees—rather than insisting on a complete orbit. The point is to show how the structure belongs to the landscape. I would plan the stand-off from its nearest part, not mistake a radius measured from its centre for clearance.
A slow camera move can reveal a doorway, the depth of a wall and the countryside beyond without becoming an architectural inspection. I would favour a short section with useful side light over a complete circle that ends in flat shadow.
The deciding check: permission, visitor access, the applicable category, wires and every part of the arc. UK A3 currently requires at least 50m from any building, as well as its other separations. A tourist attraction can also be a recreational area.[14][11]
History supplies the character. The camera supplies the introduction.
The reservoir beyond the obvious shot.
Follow a curve, a reflection or a meeting of water and hillside.
I would begin by looking for a graphic relationship: a shoreline leading towards a hill, a patch of reflected light or a quiet inlet that changes shape as the camera moves. A restrained lateral pass with an oblique downward view can make that relationship clear.
The storyboard might use a short movement at around 2m/s, but I would choose the actual path from a position that keeps the aircraft visible from land. The aim is a calm composition, not a low skim that makes the entire flight depend on the surface below.
The deciding check: water-company land rules, facilities, nature protection, walkers, birds and any airspace restrictions. I would not assume that a reservoir, dam or quiet shoreline is available simply because a map lets me draw over it.
Sometimes the most dramatic thing the camera can do is keep its composure.
One place. Four different worlds.
Let the season change while the composition appears to continue.
This is one of the most appealing reasons to save a carefully designed move. I would choose a short route over suitable open land, record the intended launch reference and return in different seasons. A recognisable hedge, lone tree or field boundary would help anchor the edit.
I would aim for a similar direction of light and a similar composition, then align the footage in editing. Spring green could become late-summer gold; autumn could give way to a winter structure of bare branches. A gentle dissolve would make the change feel deliberate.
The deciding check: the route must be reassessed every visit. Trees grow, wires appear and public access changes. Repeating a saved consumer-drone route is not survey-grade or pixel-perfect positioning; I would expect some correction in the edit.
The camera revisits. The landscape answers differently.
Twenty excellent seconds.
A partial orbit gives depth without requiring the audience to complete the lap.
I would select a simple, isolated subject and study a broad arc with good visibility throughout. At a conceptual 60m radius, a quarter-circle is about 94.2m long. At 2m/s, that is roughly 47 seconds of nominal travel before other allowances—more than enough material for a much shorter edit.
The arithmetic is useful because a circle on a screen can look deceptively compact. I would decide which part has the best background and light, then build the shot around that part. There is no artistic obligation to use a full revolution.
The deciding check: the aircraft’s sideways corridor and the scene behind it. I would never treat obstacle sensors as an assurance that a lateral route is clear. Buildings require their own stand-off assessment; the example radius is not permission.
Round and round is a route. A changing perspective is a shot.
When fields become graphic design.
Tracks, hedges and textures can become a remarkably clean composition.
I would look for an open, lawful area with a useful diagonal or a repeated pattern. An overhead view and a short, slow translation can turn the countryside into lines and colour blocks. An illustrative +40–60m height and 1–2m/s pace would be a planning study, then checked against the actual terrain and visibility.
The composition might be stronger with the diagonal entering from a corner rather than dividing the frame exactly in half. I would also consider a still photograph: a pattern that looks wonderful frozen does not always need movement to justify it.
The deciding check: farm activity, animals, walkers and the complete forward route. A downward-facing camera gives the screen a beautiful view of the ground and very little help with what is ahead. The pilot’s direct observation remains essential.
A landscape can be a place and a pattern at the same time.
The same view, a different chapter.
Make progress obvious by keeping the viewpoint comparable.
For an appropriately permitted project, I would choose two or three clear viewpoints and document their intended framing. The value lies in returning to a similar composition: a landscaping project taking shape, a restoration progressing or a piece of open land changing through the year.
I would keep a simple record of date, weather, launch position, camera choices and anything that forced a change. A before-and-after slider could then reveal the difference much more clearly than a folder of unrelated angles.
The deciding check: an active worksite or plant yard is not an uncomplicated A3 setting. Suitable A2 conditions or another appropriate authorisation, site coordination and separation may be necessary. This is a photographic progress record, not a measured survey, engineering inspection or assessment of structural safety.
Consistency makes the change visible.
The wide view starts with patience.
Choose one excellent viewpoint before collecting many frames.
I would first find a composition that deserves the extra width. A panorama needs more than a large quantity of scenery: a foreground anchor, a clear horizon or a strong relationship between nearby and distant features gives the finished image a reason to exist.
From a suitable stationary hover, I would use the supported panorama workflow and inspect the result for stitching problems. Water, moving branches, clouds and changing light can all make the joined image less tidy than the original idea. A simpler view may sometimes produce a stronger result.
The deciding check: the camera sequence must not distract me from drift, people, battery condition or surrounding aircraft. I would keep the hover in direct sight and leave adequate time to return, even if the view is doing its best to persuade me to stay.
A panorama should widen the story, not merely the file.
Clouds race. Nemesis takes her time.
An interval-photo experiment can turn a gentle move into a hyperlapse-style sequence.
I would begin with a very conservative capture plan and assemble the photographs afterwards. This is a proposed workflow, not a promise of one-button Litchi movie creation. The camera’s actual capture rate, write time and selected settings must all support the interval.
The arithmetic puts the ambition into perspective. One hundred photographs played at 25fps produce four seconds of footage. With the first frame at time zero and five seconds between frames, the last of those 100 frames arrives at 495 seconds: 8 minutes 15 seconds. Positioning and the journey home come on top.
The deciding check: acquisition time and battery margin. A four-second finished clip can demand a surprisingly long period in the air. I would shorten the experiment or change the idea before stretching an ageing battery to reach an arbitrary frame count.
The edit accelerates time. It cannot refund battery.
Torque enters the frame.
Let a cyclist give the landscape scale without turning the flight into a chase.
I would use a separate, competent pilot and an appropriate private open route. The camera could execute a short, predictable movement while the rider enters the planned composition, crosses it and leaves. The landscape remains the subject; the person supplies movement and scale.
This is often a more manageable creative idea than asking the aircraft to follow a rider through an increasingly complicated route. A rehearsal on the ground could establish timing without trying to solve the shot while everyone is already moving.
The deciding check: I cannot be both rider and supervising pilot. No road chase, blind corner or tree corridor is implied. Consent and a separate pilot do not remove the other operating conditions. The CAA specifically warns against piloting while also operating a moving vehicle.[12]
Give the scene a human scale. Keep the flying role unambiguous.
Leave the picture room to breathe.
End by letting the subject become part of a larger place.
For a closing shot, I would consider a brief, gently rising retreat within an already assessed corridor. The start would be clear of known obstacles; the movement would gradually open the composition; and the ending would leave a few seconds of calm material for a title or fade.
I would decide the end position together with the return plan. It should be somewhere useful for flying, not merely the place where the image looks most impressive. The last waypoint is a waypoint, not a promise that the rest of the job has taken care of itself.
The deciding check: visibility, return direction, battery margin and a clear landing area. I would stop the creative part while there is still comfortable room for the practical part. A beautiful ending on the memory card is only half of a satisfactory finish.
The film can finish with a flourish. The flight should finish with a calm landing.
THE HASSELBLAD END OF THE EQUATION
Automation cannot choose beautiful light.
A beautifully repeated move through dull light is still a beautifully repeated move through dull light.
I would put as much thought into when I fly as where the waypoints go. Side light can describe a hillside. Low light can separate the van from the grass. A passing cloud can flatten the scene halfway through a move. The planner cannot make those creative decisions for me.
Before recording, I would check focus, the horizon, exposure and what happens when the camera turns towards the brightest part of the scene. A route that changes direction can also change the demands on the image.
| What I want | What I would try | Why I would choose it |
|---|---|---|
| A straightforward film | 4K/25fps and a normal colour workflow. | A practical starting point for a UK-oriented edit when I want a relatively direct route to a finished SDR video. |
| A considered grade | D-Log M with the appropriate colour conversion and shot-by-shot adjustment. | Use it when I intend to do the post-production work; a flat-looking file is not the finished picture. |
| An HDR project | HLG with a deliberate compatible editing and delivery workflow. | Choose it because the output is HDR, rather than treating it as an automatic upgrade to an ordinary web-video export. |
| Consistent motion | At 25fps, consider around 1/50sec as a cinematic starting point. | This follows a conventional 180-degree shutter choice. The subject and intended effect may justify a different shutter. |
| A steady-looking sequence | Set a considered white balance and exposure before the useful part of the shot. | I would avoid distracting colour or brightness changes when repeatability is the purpose. |
| Editable photographs | DNG RAW, checked focus and a shutter suited to the movement. | Detailed stills have different priorities from deliberately motion-blurred video. |
Those are editorial starting points, not a universal settings recipe. DJI documents the camera’s D-Log M and HLG choices, while its ND guidance explains how reducing incoming light gives more freedom over shutter speed and aperture.[2][18]
I would select an ND filter for the light and the intended exposure, rather than fitting one because the morning happens to look sunny. For a video move, a slower shutter may give the motion the character I want. For crisp interval photographs, I may have a different priority. One filter and one shutter speed cannot answer every photographic question.
I would also separate a stills mission from a video mission when necessary: DJI states that this model cannot take photographs while recording video. And the Pro has no optical zoom, so a change in perspective comes from the camera’s position, framing or the edit—not from borrowing the Mavic 2 Zoom’s lens capabilities.[1]
Edit for the audience, not for the flight log
A planned route can produce a long clip, but the finished film does not owe the route its full duration. I would keep the part where the composition changes meaningfully. If the useful reveal takes eight seconds, that may be all the audience needs.
I would leave a little movement before and after the part I expect to use, avoid needless speed changes in the final cut and let the scene’s natural rhythm guide the music or ambient sound. A few different kinds of shot—a reveal, a sideways pass, an overhead still—usually give an edit more shape than several versions of the same orbit.
For the seasonal scenario, I would align recognisable landmarks first, then adjust the transition until the change feels intentional. For a panorama, I would inspect the joins before celebrating its size. For any clip, I would check the full-resolution file rather than judging the result solely from the small live view.
THE FLIGHT CONTINUES AFTER THE SHOT
Plan the return with the same care.
The Mavic 2 Pro can execute an uploaded waypoint route onboard. Litchi’s team confirms that it may continue a mission after a link loss according to the configured settings. DJI’s waypoint API separately exposes whether the mission should exit when the RC signal is lost.[6][7]
That is a behaviour to understand, not a filming technique. I would not deliberately plan a radio blackout or a route out of sight. Nor would I assume that every camera function continues just because the aircraft can continue along its route.
Mission finish
What should happen when the intended route ends?
Link loss
What happens if the controller connection is unexpectedly interrupted?
Battery response
What protection or landing behaviour could interrupt the creative plan?
I would treat those as three separate questions. A finish action does not explain every failure case. A return-home setting does not inspect the path for me. A remaining-battery percentage is not a guarantee that a headwind, a detour and a climb can all be absorbed.
The official manual distinguishes Smart RTH, Low Battery RTH and Failsafe RTH, and warns that critically low energy can trigger automatic landing. It also warns against relying on obstacle avoidance during return when the relevant sensing is unavailable.[19]
Give every minute a job
I would budget the sortie in parts: take-off and checks; travel to the start; the useful camera movement; travel back; landing; and a margin for change. Repeating the shot belongs in the budget too. If the first version is wrong, the solution is not automatically “one more go”.
I would pay particular attention to the direction of the return. A favourable outward breeze can make the opening minutes feel easy while quietly making the homeward leg the harder part. The plan should be judged against the whole flight and the actual battery’s condition, not the brochure’s best case.
Before committing to a programmed route, I would know the app’s pause and stop controls and practise the documented takeover method. For the Mavic 2 family, Litchi documents switching from P to S to cancel, then returning to P after taking control. Sport mode changes handling and disables vision obstacle sensing, so it is not a mode to leave engaged casually.[4][19]
Sensors are help, not a route survey
DJI’s current documentation limits side sensing to particular modes and explicitly cautions that “omnidirectional” does not mean a complete protective envelope. I would not assume the same sensing behaviour in every Litchi operation.[1]
That matters especially for arcs and lateral shots, where the camera can be looking somewhere other than the direction of travel. My route assessment follows the aircraft, not merely the view on the screen. If the shot requires an unverified clearance to work, it needs redesigning.
A UK AIRCRAFT WITH EUROPEAN AMBITIONS
The rules travel with the plan.
This section is checked against official sources on 29 October 2026. It describes the ordinary, unclassmarked Mavic 2 Pro, not a differently approved or modified aircraft. My current qualifications and any particular site permission are not established by this article.
The useful UK point is that the legacy aircraft is not automatically A3-only in 2026. The current A2 rule retains aircraft below 2kg placed on the market before 1 January 2026, with the required certificate and at least 50m horizontal separation from uninvolved people.[13]
Far from people.
- Flyer ID and Operator ID requirements apply.[21]
- At least 50m horizontally from uninvolved people.
- At least 150m from residential, recreational, commercial or industrial areas.
- At least 50m from any building.
Those are minimum category conditions, not proof that any individual flight is safe.[14]
More location options, with conditions.
- A valid A2 Certificate of Competence, plus the applicable IDs.[21]
- This legacy aircraft needs at least 50m horizontal separation from uninvolved people.
- Do not plan to overfly uninvolved people.
- The reduced class-marked 30m/5m options do not apply to Nemesis.
Appropriate location permissions, airspace checks and every other operating condition still matter.[13]
Visual line of sight means being able to see the aircraft directly well enough to judge its orientation and control it safely. The live video is not a substitute. If using an observer, the CAA requires them to be beside the pilot with continuous communication; a distant chain of observers is not the arrangement described here.[10]
Beaches, parks, campsites and visitor attractions need careful thought because recreational-area restrictions can matter even when the photograph looks peaceful. An app allowing a launch is not a landowner’s permission or an airspace clearance. I would check the current location rather than rely on a saved route’s age or popularity.
The CAA currently makes Remote ID mandatory from that date for legacy camera-equipped aircraft such as this one in the Open category. I would check the compliant solution for Nemesis before the deadline; this article does not claim its existing firmware already provides one.[15]
Taking Nemesis to Europe
For an ordinary unclassmarked Mavic 2 Pro, the EU Open-category legacy route is A3 following the transition that ended on 1 January 2024. I would not carry the UK’s continuing legacy A2 allowance across the Channel and assume it works in France or Spain.[16]
As a UK resident visiting EASA countries, I would arrange the appropriate operator registration and competence requirements through the relevant European system, beginning with the first EU country where I intend to operate. EASA says non-EU qualifications are not automatically recognised. National geographical zones and local requirements still need checking for each destination.[17]
That matters for the kind of touring I enjoy. Planning a beautiful coastal sequence on the MacBook is part of the pleasure; checking that the actual country and location permit it is part of making the plan usable.
THE NIX VIEW
Give an older aircraft a thoughtful new brief.
What appeals to me about Nemesis and Litchi is the chance to be more deliberate. I can separate the business of designing a shot from the pressure of trying to invent one while flying. I can give the camera movement a purpose, preview the idea and arrive with a plan that I am prepared to change.
The first mission I would choose is small: a suitable rural location, a short visible corridor and one clear reveal. I would check the result, understand the behaviour and save what I learned. Then I would build the next idea on that knowledge.
I would be especially interested in the four-season project. There is something satisfying about returning to a place and noticing how much has changed. A repeatable camera move provides a visual thread through those visits. The aircraft supplies the viewpoint; time supplies the story.
And for Vanilla, I can already see the kind of sequence I would like to plan: the van in the opening frame, the surroundings slowly arriving, a few seconds of space at the end. Enough to make the viewer feel the place without turning a campervan stop into a feature-length orbit.
Nemesis does not need to pretend to be the newest aircraft in the catalogue. She needs an appropriate job, a considered camera setup and a pilot who remains involved. That feels like a very good use of a capable piece of kit.
Plan with care.
Fly with attention.
Bring home a story.
A FEW QUICK ANSWERS
Before the propellers turn.
Is this a set of downloadable missions?
No. The routes, figures and animation are explanatory storyboards. They contain no real coordinates and have not been surveyed or validated for any location. A real mission needs its own planning and checks.
Which idea would I start with?
The short rural reveal. One subject, one visible corridor and a clear return give me a useful first experiment. I would check the resulting footage and the aircraft behaviour before adding complexity.
Why not just fly every shot manually?
Manual flight remains valuable. A saved plan is useful when I want to prepare a move carefully or revisit a similar composition. I would choose the method that gives me the best combination of control, attention and creative purpose.
Can I trust a route because somebody else shared it?
I would treat a shared route as an idea to inspect. Its creator’s launch point, conditions, aircraft, permissions and assumptions may differ from mine. A public mission is not an approval to fly it.
When would I abandon a planned shot?
Whenever the real situation stops matching the assumptions that made it sensible: unexpected people, an uncertain clearance, worsening weather, poor visibility or a warning I need to resolve. I can revise a storyboard another day.
RESEARCH NOTES / PRIMARY SOURCES
The reading behind the flight plan.
Technical and regulatory checks: 29 October 2026. Official documentation supports the factual statements; the shot concepts, editorial preferences and labelled calculations are the NIX creative interpretation. App behaviour and regulations should be checked again before a flight.
- DJI — Mavic 2 specifications and supportCamera, weight, stated endurance, capture limitations and sensing cautions.
- DJI — original Mavic 2 announcement, 23 August 2018Hasselblad partnership, the Pro camera and its colour-workflow features.
- Litchi — supported drone modelsOriginal Litchi app versus Litchi Pilot compatibility.
- Litchi — original-app user guideWaypoint heights, curved-turn actions, gimbal modes and takeover controls.
- Litchi — Hub overview and current planning toolsBrowser planning, simulation, camera projection and app-specific integration.
- Litchi staff — older-aircraft support in the new HubMavic 2 Pro support, unavailable newer components and configured mission continuation.
- DJI Mobile SDK — waypoint mission referenceUnderlying waypoint execution and RC-signal-loss setting.
- Litchi — original app feature overviewWaypoint, Focus, Track, Orbit, Follow and Panorama features.
- Litchi for DJI drones — UK App Store listingCurrent iOS requirements, release notes and mode descriptions; not a guarantee for an untested device combination.
- CAA — flying safely and responsiblyDirect sight and the role of an observer beside the pilot.
- CAA — where you can flyTerrain-aware height limits, people, recreational areas and restricted locations.
- CAA — remote-pilot control responsibilitiesAutonomous operations, intervention, moving vehicles and avoiding foreseeable link loss.
- CAA Regulatory Library — current A2 ruleLegacy aircraft below 2kg, competence and the 50m people separation.
- CAA Regulatory Library — current A3 rulePeople, area and building separations for the UK category.
- CAA — Remote ID datesThe current 1 January 2028 requirement for relevant legacy Open-category aircraft.
- EASA — drones without class identification labelsThe EU legacy Open-category position after 1 January 2024.
- EASA — non-EU visitors flying in EuropeRegistration and recognition of pilot qualifications.
- DJI — ND filters and exposure choicesLight reduction, shutter speed and intended motion rendering. No ND stop-count table is reproduced here.
- DJI — Mavic 2 Pro / Zoom user manual, v2.2Flight modes, return-home behaviours, battery warnings and sensing limitations. Refer to the Pro-specific sections where models differ.
- DJI Mobile SDK — Waypoint classNormal and curved path geometry, and waypoint-action behaviour illustrated in the route comparison.
- CAA — introduction to more advanced flyingFlyer ID, Operator ID and the competence needed for A2 and A3.