How to Measure Your Cadence From a Phone Video

How do you count cadence from a running video?

Film yourself running from the side for at least 30 seconds, then switch the clip to slow-motion playback (most phones support 120 or 240 fps slow-mo natively). Pick one foot — say, your right — and count every time it strikes the ground for 30 seconds. Multiply that number by 4 to get steps per minute (spm) for that foot, then double it to get total cadence, since both feet contribute roughly equal steps. If your right foot strikes 42 times in 30 seconds, that's 168 spm from that foot alone — meaning your total cadence is close to 168 spm once you account for both feet's steps within the same 60-second window (in practice, most people just count total foot-strikes, both feet, over 30 seconds and multiply by 2, which is simpler and gives the same number).

A tripod or a fence post at hip height gives you a stable, repeatable angle. Side-on framing is what most software (including StrideIQ) expects, because it's the angle that makes foot-strike timing visible frame by frame.

Why is counting from video more accurate than a running watch?

Watch-based cadence estimates come from an accelerometer inferring stride rhythm from wrist motion, not from direct measurement of foot contacts. That's an indirect proxy, and it can misestimate cadence versus a direct video count — arm swing asymmetry, watch position, or even carrying your phone can throw off the accelerometer's pattern recognition. A video count is a direct observation: you're watching the foot hit the ground and counting that event, frame by frame, so there's no inference step to introduce error.

This matters most at the extremes — very slow warm-up jogs, sprint intervals, or hilly terrain — where accelerometer-based algorithms are more likely to drift from what your feet are actually doing. If you've ever glanced at your watch mid-run and thought the cadence number looked oddly low or high, a quick video count is the fastest way to sanity-check it.

What's a good cadence target?

A commonly cited range for recreational distance runners is 170–180 spm, though this isn't a hard rule — cadence scales with leg length, pace, and running experience, and elite runners are frequently observed well above 180 spm at race pace. The oft-repeated "180 is optimal" figure originated from observations of elite runners decades ago, not a controlled trial proving that hitting exactly 180 prevents injury or improves everyone's economy. For a deeper look at where that number came from and whether it holds up, see is 180 spm cadence a myth.

What the research does support more consistently: modest increases in cadence (around 5–10% above your current baseline) can reduce loading at the hip and knee for runners who overstride — meaning they land with the foot too far ahead of the hip, increasing braking forces with each step. That's a tendency worth checking, not a universal fix, and it's most relevant if you're managing a current injury; see cadence for returning after injury and cadence for shin splints for context-specific guidance.

Method What it measures Typical accuracy for cadence
Manual video count (slow-mo, side view) Direct foot-strike count High — you're counting the actual event
GPS/running watch (wrist accelerometer) Inferred stride pattern Moderate — can misestimate vs. direct count
Footpod or foot-mounted sensor Direct contact detection High, but requires extra hardware
Phone-video app (e.g., StrideIQ) Automated frame-by-frame foot-strike detection High for cadence; lower confidence for contact time or joint angles
In-person gait lab Multi-camera + force plate Highest, but not accessible for routine training checks

What does StrideIQ measure automatically?

StrideIQ analyzes a single side-view phone video and detects foot-strike frames automatically, converting that into a cadence figure without you having to count manually. Cadence is the metric a single phone video can measure with the most confidence, because it only requires identifying discrete foot-strike events in time — something a normal video frame rate (even 30 or 60 fps) can capture reliably.

Be realistic about what it can't do with the same confidence. Ground-contact time (how long the foot stays on the ground per step) and vertical oscillation (how much your center of mass bounces) are measured in milliseconds and millimeters — figures that genuinely need high-speed or multi-camera capture to nail down precisely. A phone video can estimate these, but with meaningfully lower confidence than a lab setup, and true joint angles (knee flexion, hip extension) are also rough estimates from a single 2D angle, not goniometer-grade measurements. If you want a fast cadence and general form check from your phone, that's a reasonable use case; if you need clinical-grade joint mechanics, that calls for an in-person gait lab or physio assessment.

How do you interpret your cadence results?

Once you have a number — whether from manual counting or an app — context matters more than the raw figure. Cadence naturally varies with pace: you'll run at a lower cadence on an easy recovery jog than during a tempo effort, and that's expected, not a flaw. Compare your cadence across similar-effort runs over time rather than chasing one static target.

If you're a new runner, don't panic about hitting 180 on day one — pace and comfort matter more early on; see how fast should a beginner run for a more useful starting framework. If your cadence sits notably below 160 spm at an easy pace and you're dealing with recurring knee or shin pain, a modest, gradual cadence increase (5% at a time, over several weeks, not overnight) is one variable worth experimenting with alongside — not instead of — a broader look at your training load, footwear, and recovery.

What can video cadence analysis NOT tell you?

A cadence number, however accurately measured, doesn't diagnose an injury or tell you why your knee hurts. Foot-strike pattern (heel, midfoot, forefoot) and "correct" running form remain genuinely debated in the sports-medicine literature — no single form is proven best for everyone, and changing form carries its own adaptation risk if done too quickly. Cadence itself is only loosely linked to injury outcomes in research; it's a modifiable training variable, not a guaranteed fix.

Video analysis, from any app, also can't measure internal forces on your joints, tendons, or bones — that requires force plates and imaging, not a camera. If you're running with pain that doesn't resolve with rest, or you're returning from an injury and want a structured plan, talk to a physiotherapist or sports medicine physician who can assess you in person. Treat cadence data as one useful input into your training, not a diagnosis or a substitute for professional care.

Frequently Asked Questions

How do I calculate cadence from a running video?

Film yourself from the side for at least 30 seconds, play it back in slow motion, count one foot's ground contacts during that 30 seconds, then multiply by 4 to get that foot's steps per minute — or count both feet's contacts and multiply by 2 for the same total cadence figure.

Is a running watch accurate enough to measure cadence?

Running watches estimate cadence from wrist accelerometer data, which is an indirect proxy and can misestimate cadence compared with a direct video count, especially at very slow or very fast paces. A quick video count is a useful way to double-check a watch reading you're unsure about.

What is a good running cadence?

A commonly cited range for recreational runners is 170–180 steps per minute, but this varies with leg length, pace, and experience — it's a general reference point, not a strict target everyone should hit.

Can a phone video measure ground contact time accurately?

Not with high confidence. Ground contact time is measured in milliseconds and typically needs high-speed or multi-camera capture; a standard phone video can only estimate it, with lower reliability than it offers for cadence.

Should I change my cadence if I have knee or shin pain?

A gradual cadence adjustment is one variable some runners experiment with, but it isn't a guaranteed fix and shouldn't replace evaluation by a physiotherapist or sports medicine professional if pain persists.

Sources