2c:null 2d:["$","title",null,{"children":"Investigating mountain bike rear suspension | johnwelford | Observable"}] 32:T1144,whatWillWork = html`

If it's your first time on this then click here to check what will work in this tool.

What will work?

The picture we are using for analysis should be a plain side view of all of the bike, similar to the following:

✔

In order to get accurate results the following should be avoided:

Pictures that are not square to the camera, or where wheels are turned

✘

✘

✘
Pictures that don't include all of the bike

✘
Frames without wheels fitted

✘

✘
Mixed or strange wheel sizes
Pictures where the suspension is compressed – it should be at zero sag
Pictures that are distorted, including squashing or stretching
Bikes that are facing to the left (drive side away from view)

The algorithms are designed for single pivot and four-bar suspensions, including virtual-pivot type systems; they are not currently built to support the following less common designs:

Designs that rely on the flexure of one or more parts, rather than pivots. e.g. Cannondale Scalpel
Designs that utilise a floating bottom-bracket. e.g. GT iDrive bikes
Five-bar and more complex linkage systems. e.g. Tantrum Missing-Link or Felt Equilink bikes
Mechanisms that use a sliding part rather than a pivot. e.g. Yeti Switch infinity bikes (although it is possible to fudge these with a sensibly placed imaginary pivot)
Systems with multiple shocks, or extra actuators. e.g. Kona Magic-Link bikes, Canyon Shapeshifter bikes
Unified rear triangle designs (is anyone still making these?)
Complex linkage driven shocks (although the axle-path analysis should still work)
Idler pulley driven designs (although the axle path and shock linkage analysis should still work)

Most others should work. If you're unsure then give it a go and see what the output looks like.