Search options

A Cadence tool could be useful Joan. I coach dragon boat racing and sometimes relate cadence data to boat speed or paddler action. Usually, I extract the data from object tracking info. But entering data from the keyboard is useful especially if the video is not suitable for kinematic analysis.
1. General interaction
Keyboard interaction is fine
2. Cadence display
- count, cadence, half cadence, double cadence.
- The raw count can be interesting for example in 400 m Hurdles where we want to know how many strides are taken between each hurdle. Or simply as a repetition counter for any sort of exercises.
- half cadence and double cadence: for example, in swimming it's easier to mark the beat when a specific arm goes down into the water but we may want to know the stroke rate instead. In running it's easier to tap a beat on each step but we might be interested in a tempo based on the full cycle. cumulative water time vs air time would be very useful, so summing half cadence time (paddle in, paddle out) will address this
- Precision: integer cycles or fractional cycles. (70 rpm vs 70.512 rpm). Fractional raw count may be interesting as well. For human subjects, integer cycle count is fine, and cumulative water/air time in ms is good
Question
- Do you think the simple counter and the cadence tool should be two different tools? It seems a bit convoluted to use a cadence tool if all you want is to quickly count reps. maybe a 2 step tool, first count, then analyse
3. Partial cycles in a given period
I am not sure what you are trying to say here.
I think cadence count is always full cycles over a selected time span
4. Compute period
- The basic idea is thus to compute cadence by averaging over the time section (full cycles divided by their span).
- Another idea would be to compute the cadence based on the last X seconds or the last X cycles. How useful is this?
cadence will vary at different stages (e.g. start, steady, burst, finish). So being able to record and save data for each selected piece is important
5. Distance calibration
- If we know the distance covered during the time period we can compute extra metrics. that data may come from other sensors. It would be nice if users can import such data and display on the video
- Stride or stroke length.
- Pace (time per distance unit).
- So it should be possible to go and specify the distance coordinate of time section end points.
- However, this calibration is going to be over the whole time section, so now we definitely need to to deal with the partial cycles at the start and end of the section.
extra metrics need to be for each time interval. It would be nice if they can be shown in the kinematic analysis grahs
6. Units
Hertz is fine
7. Cadence deviation
I don't think that is needed, coaches and athletes know what the reference is and will judge whether the observed cadence is significantly different from what's desired

Hope this helps

The camera I used is a Kayteon 120fps global shutter USB camera.
I also made 2 recordings under the same conditions using Kinovea 0.8.15 and Amcap 3.0.9 (Amcap demo version doesn't render the image properly for me).
Then I played both using Avidemux 2.7.6 and to make sure, advanced 40 frames and checked the time difference.
The Kinovea recording gave 727ms for 40 frames = 55fps
The Amcap recording gave 333ms for 40 frames = 120fps