This comes up constantly from people filming themselves during training. The mount feels solid. The phone is tight in the cradle. You walk, the footage is fine. You start running and the image turns to mush, with a specific low buzzing wobble that autofocus keeps hunting through.
Almost every guide answers this with "tighten the clamp" or "buy a better mount." Both miss what is actually happening.
Walking and running are two different vibration sources
A treadmill under load produces vibration from three places at once: the motor, the belt and roller assembly, and your own foot strike transmitted through the deck. They do not scale together.
At walking pace the dominant input is the motor, which is relatively high frequency and low amplitude. Your mount barely notices it, and most cradles have enough internal friction to absorb it.
At running pace the dominant input becomes deck deflection. Each foot strike loads the deck, the deck flexes and rebounds, and that rebound travels up the frame into whatever is clamped to it. This is a slower, larger movement, and it is the one that ruins footage.
That is the whole answer to why the same mount behaves differently at two speeds. You have not changed the mount. You have changed which vibration source is loudest.
Why a motorcycle dampener does not fix it
The obvious move is to buy a vibration dampener, since motorcycles solved this years ago. Quad Lock and RAM both sell silicone grommet suspension modules, they are inexpensive, and they genuinely work on a bike.
They are built for a different problem. Quad Lock's own support documentation states that its grommets are specifically tuned to dampen higher frequency vibration, and that users may notice more movement at lower frequencies. That is the correct engineering decision for a motorcycle, where the destructive input is high-frequency engine vibration that can damage camera sensors.
Treadmill deck oscillation sits at the other end. Lower in frequency, lower in amplitude, and closer to the band those modules explicitly are not targeting. Fitting one to a treadmill mount addresses the part of the spectrum that was never the problem.
Grommet durometer, and why it decides everything
A grommet-suspension dampener is a mass on a spring. The silicone is the spring, and its stiffness is set by durometer, measured on the Shore A scale. Durometer selects which frequencies get absorbed and which pass straight through.
| Grommet | Behaviour | Suits |
|---|---|---|
| Harder, higher durometer | Stiffer spring, higher resonant frequency, absorbs fast vibration | Engines, power tools, high-frequency sources |
| Softer, lower durometer | Softer spring, lower resonant frequency, absorbs slow oscillation | Deck flex, structural sway, low-frequency sources |
Motorcycle modules use the harder end because that is what an engine demands. A treadmill needs the softer end, which is a straightforward change to make and, as far as we can find, one nobody has made in a product aimed at this use.
The trap: you can make it worse
A damper does not remove vibration, it moves the resonant frequency. Pick a durometer whose resonant frequency lands near the treadmill's dominant oscillation and you amplify it rather than absorbing it. This is why a random soft grommet is not automatically an improvement, and why the durometer has to be selected against a measured input rather than guessed.
The part nobody mentions: arm resonance
The mount arm is a cantilever, and every cantilever has its own resonant frequency set by length, stiffness and the mass on the end. A long arm holding a heavy phone resonates slowly. A short stubby one resonates fast.
This produces a counterintuitive result. On a motorcycle you want the arm short and stiff, so its resonance stays above the engine band. On a treadmill the opposite reasoning applies, because the destructive input is already slow, so pushing arm resonance in the same direction is exactly what you do not want.
Practically, that means a treadmill mount should not simply be a scaled-up motorcycle mount. The arm needs to be stiff for its length, and the damping stage should sit at the base of the arm rather than at the phone, so the arm's own resonance is isolated too.
What to do about it today
- Move off the console. The display panel is the most flexible large surface on the machine and it is directly above the deck. The side rail is stiffer and closer to the frame. This alone helps more than any accessory.
- Add mass at the phone end. Counterintuitive, but a heavier assembly resonates lower and more slowly, which is easier to correct in post than a fast buzz. A cheap metal cradle instead of a plastic one is a real improvement.
- Do not rely on a motorcycle dampener alone. It is not doing nothing, but it is not addressing your dominant input.
- Try a floor tripod first. Standing on the floor rather than the machine removes the problem entirely. If your room allows it, that is the honest answer, and it costs less than any mount.
- Lock portrait mechanically. Rotating heads with a single friction knob work loose under repeated low-frequency loading, which is how mounts drift from portrait to a slight tilt over a session.
Where this comes from
The frequency-band argument here is reasoned from published manufacturer documentation, principally Quad Lock's own description of how its grommets are tuned, combined with standard cantilever and grommet-isolator behaviour. We have not yet put an accelerometer on a treadmill rail to measure the actual spectrum, which is the obvious next step and the thing that would turn this from a well-supported argument into a measurement. If you have done it, we would like to see the data.