Short answer: a 5-7 ft tripod with a sandbag hung from the centre column, placed 6-10 ft behind one baseline. That's the default. The rest of this post is what to do when the default doesn't fit your court — magnetic mounts on steel poles, monopods stuck into ground spikes, and a brutally honest list of the setups we've watched fail in the wild.

If you're filming for AceSense (or any AI tennis app), the rules of the camera don't change because there's no fence. The pipeline still needs the phone above net height, behind the baseline, framed on the whole court. The only thing that changes is what's holding the phone up. And on public-park courts, school courts, and a surprising number of clay clubs across Europe, there is no fence to clip against.

TL;DR

Default: 5-7 ft aluminium tripod + phone clip + sandbag on the centre column. €30-€45 all-in.
If there's a steel pole nearby: a magnetic phone mount (MagSafe-compatible) on the pole at net height or above.
If you have a car: a window mount or a roof-rack mount, parked behind the baseline. Surprisingly common at amateur tournaments.
If you have nothing: a folding step-ladder from a hardware store + a phone clip = €40 and 6 ft of height in a kit-bag.
Don't: lean the phone against a water bottle on the court. The angle is below the net cord and the report will be useless.

Why "no fence" is a real problem (not a niche one)

The fence-clip mount is the hidden assumption behind most tennis filming guides — including ours. It's the easy answer: a €15 clip from Amazon, snap it to the chain-link, you're filming. But a non-trivial chunk of European tennis happens on courts that don't have a fence behind the baseline: red-clay courts at French and Italian clubs, padel-style indoor halls with glass walls, public-park hard courts where the perimeter is a row of bushes, and most school sports halls.

There's a Quora thread where the question is literally "a good camera I could just hook to a fence, cheaper the better" — that's the standard mental model. The first three answers all assume a fence. None of them help you if you don't have one.

Here's what does.

Option 1: the default — tripod plus ballast

For 80% of fence-less courts, this is the answer.

What you need:

A 5-7 ft aluminium tripod (€20-€35). Look for one rated for at least 1 kg payload, with a screw mount in the standard 1/4"-20 thread.
A phone clip with a 1/4"-20 mount on the bottom (€8-€15). Most adjust to 5.5" - 7" phones.
A weight you can hang from the tripod's centre column hook. Options: a sandbag (€10), a reusable shopping bag with two 1.5L water bottles, your tennis bag.

How to set it up:

Position the tripod 6-10 ft (1.8-3 m) behind the baseline, centred on the court's lengthwise axis.
Extend the legs fully and lock them.
Extend the centre column to put the phone clip at 6-7 ft (above net height).
Hang the ballast on the centre column hook. This is non-optional in any breeze.
Mount the phone in landscape, frame the whole court, lock orientation, press record.

Why this works: the phone ends up at 6-7 ft, looking slightly down at the court — exactly where ball tracking, court detection, and pose detection all want it. We cover the why in detail in the filming guide, but the headline is: every foot above net height meaningfully improves shot detection accuracy.

Where this fails: wind above 20-25 km/h, soft surfaces (sand, mud near a clay court), and any court where you can't put a tripod 6-10 ft behind the baseline because there's a wall there. For those, see options 2 and 3.

Option 2: magnetic mounts on steel poles

A lot of fence-less courts have something metal upright near the baseline — a flag pole, a light pole, a netpost extension, a perimeter post for crowd ropes, the upright on a permanent scoreboard. If that pole is roughly 6-10 ft behind one baseline and at least 5-6 ft tall, a magnetic phone mount is the fastest setup in tennis.

What you need:

A MagSafe-rated magnetic mount with a flat steel base (€15-€30). The cheap "fridge magnet" ones do not work; look for rare-earth magnets rated for at least 1.5 kg pull.
A phone case that's MagSafe-compatible or a metal disc adhered to the phone's back.

How to set it up: wipe the pole with a sleeve to remove dust, stick the magnet to the pole at net height or above, attach the phone in landscape. Whole thing is 30 seconds.

Why this works: zero footprint, zero ballast needed, no risk of tipping, and you can adjust the height by sliding the magnet up the pole. The pole is doing the engineering for you.

Where this fails: painted steel where the paint is thick (the magnet decouples), aluminium poles (not magnetic), and any pole thinner than ~3 cm (the magnet contacts only on a tiny area). Test before the match.

Option 3: the step-ladder (yes, really)

This is the underrated solution most people don't think of. A small folding step-ladder — the two-step or three-step kind that lives in most apartments — gets you to 5-7 ft instantly, doesn't tip, and packs flat in a car boot.

A €30 aluminium two-step gives you 4-5 ft. A €50 three-step gives you 6-7 ft. Stick a phone clip on the top rung, weight the bottom step with your tennis bag, you're done.

Why it's good: dead-stable. No tripod-tip risk in wind. Zero balance work. You can also stand on it to set up framing, which is harder than it sounds on a tripod where the phone is already at head height.

Why most people skip it: it looks ridiculous. We've watched players walk past a perfectly good step-ladder solution because they didn't want to carry one to the court. Suit yourself.

Option 4: car window or roof-rack mount

Underrated for public-park courts where you've parked just behind the baseline. A standard suction-cup phone mount on the rear-side window or a magnetic mount on the roof rack puts the phone at 4-5 ft for a sedan, 6-7 ft for an SUV. The car itself is the ballast; nothing tips.

This is the standard amateur-tournament setup we see at events that don't have permanent video. The car parks behind the back fence (or the back row of bushes), the phone goes on the roof, the player tournament-films their own match.

What doesn't work (a brutally honest list)

We have seen all of these submitted as videos. None of them produce a useful report:

Phone leaning against a water bottle on the court. Angle is below the net cord. Court detector loses the corners. Don't.
Phone in the cup-holder of a pop-up chair. Height is 30-40 cm. Same problem.
Phone on top of a tennis bag. Tippy, low, the bag sags. Same problem.
Phone on the net post. Centred on the net, looking across the court — it's a side-on angle. Looks great in a tennis brochure photo. Useless for AI.
Phone tied to a tree branch with a hair tie. This actually happened. The branch swayed in the wind enough that the court drifted across the frame and court detection re-anchored every five seconds. Not recommended.

The common thread: any setup that ends with the phone below net height, or moving relative to the court, will break the pipeline. Doesn't matter how clever the mount is.

A note on clay and indoor

If you're playing on clay with no fence (very common at European clay clubs — the back wall is often a low wooden barrier), a tripod with sandbag is fine, but place it on a concrete or paved area behind the court if there's one — soft clay under a tripod leg slowly sinks during the match and the camera drifts. There's a whole post on filming on clay courts — read that next if it's your surface.

If you're playing indoor with no back wall (rare but it happens — some converted-warehouse halls), a tripod is fine, but mind the lighting; ceiling lights cause flicker that 60fps cameras handle badly. The indoor courts post covers this.

What about hardware tracking systems?

You may have seen courts with permanent overhead cameras — Baseline Vision sells one for around €1,800, and PlaySight has SmartCourt facility installations. If your club has one, use it. For everyone else (the 99% of amateur tennis), a phone on a tripod gets you 80% of the same value at 1% of the cost. We compare phone-based tools to facility hardware in AceSense vs SwingVision: the honest 2026 comparison.

The mount is the unsung hero of AI tennis

Most of the AI tennis pipeline — TrackNet ball detection, court keypoint detection, MediaPipe pose, the bounce classifier — assumes a stable, above-net, on-axis camera view. That assumption is doing a lot of work. When the mount is wrong, the model isn't broken; it's being asked the wrong question.

Get the mount right and AceSense's shot detection feels obviously good. Get it wrong and even the best AI tennis tool produces a noisy report. That's true for us, it's true for SwingVision, it's true for everything in the category.

If you're a club player working through this for the first time, the 5-minute starter guide walks you through the first session end-to-end. If you're a coach setting up players for between-lesson homework, the junior-coach workflow has the player-facing version of this checklist.

Get the phone above the net, behind the baseline, on something that doesn't move. Everything else takes care of itself.