Sampling frequency

[Lee Cabell]

Our figure skaters perform quadruple and jumps on the ice with ang. velocity of 5-5.5 rev/s. Their air time is 0.7-0.75 s. Do you know what video sampling frequency will be necessary to collect data on these skaters? What is your suggestion to clean this signal?
Thank you,
Lee Cabell

[Stefan Litzenberger]

Re: Sampling frequency

Our figure skaters perform quadruple and jumps on the ice with ang. velocity of 5-5.5 rev/s. Their air time is 0.7-0.75 s. Do you know what video sampling frequency will be necessary to collect data on these skaters? What is your suggestion to clean this signal?
Thank you,
Lee Cabell

Hi Lee!
This strongly depends on the parameters you are planning to assess and which sensors you will be using.
Usually it is wise to stick to the Nyquist Shannon sampling theorem ($f_{sampling}$ > 2 x f of expected $f_{max}$ content of your signal). So you need to ask yourself a few questions first, or give us some more information on your parameters.
Kind regards
Stefan

[Lee Cabell]

Re: Sampling frequency

Hi Lee!
This strongly depends on the parameters you are planning to assess and which sensors you will be using.
Usually it is wise to stick to the Nyquist Shannon sampling theorem ($f_{sampling}$ > 2 x f of expected $f_{max}$ content of your signal). So you need to ask yourself a few questions first, or give us some more information on your parameters.
Kind regards
Stefan

Hi Stefan,
I understand the Nyquist theorem, but the only thing (besides what I already mentioned) are the cameras with 400 Hz. Are they fast enough to capture the fast movement and still satisfy the theorem?
Thank you,
Lee

[Stefan Litzenberger]

Re: Sampling frequency

Hi Stefan,
I understand the Nyquist theorem, but the only thing (besides what I already mentioned) are the cameras with 400 Hz. Are they fast enough to capture the fast movement and still satisfy the theorem?
Thank you,
Lee

Hi Lee
Ah, ok. You are doing motion analysis at 400 Hz. Your subject rotates with approx 5.5 rev/s which are 1980 deg/s, with 400 Hz the difference (assuming constant $$\omega$$ !!! ) between two frames is approx. 5 deg.
An airtime of 0.75 gives you 300 frames at 400 Hz.
I reckon that $$\omega$$ is not constant but will be increasing and decreasing very rapidly during take-off and landing. Additonally the arms might be moving at an even higher rate when initiating the rotation (correct?) meaning your angular resolution worsens.
Again it depends on the parameters you want to measure. In case it's only rotation angle and/or flight time, I guess 400 Hz might be sufficient, in case you wanted to analyse velocity and acceleration of limbs, a higher frequency might be required to get accurate results.
Kind regards
Stefan

P.S.: Sorry fo the $$, does anyone know whether it's somehow possible to use LaTeX notation in this forum?

[Lee Cabell]

Re: Sampling frequency

Hi Lee!
This strongly depends on the parameters you are planning to assess and which sensors you will be using.
Usually it is wise to stick to the Nyquist Shannon sampling theorem ($f_{sampling}$ > 2 x f of expected $f_{max}$ content of your signal). So you need to ask yourself a few questions first, or give us some more information on your parameters.
Kind regards
Stefan

Stefan,

The only information I have is the sampling frequency of 400 Hz. I don't know the expected frequency of video data. Is it posible to calculate it based on the information I provided? What is the "content of the signal?'
Thank you,
Lee

[Lee Cabell]

Re: Sampling frequency

Hi Stefan,
Correct, it is a motion analysis at 400 Hz. You are right, the body segments will increase (od decrease) their angular movement at take off and landing. I will measure not only body's overall angular velocity and time flight, but the segments kinematics in detail. I am very concerned about the validity and repeatability of the measurement. I guess, I will not know until I collect and analyze the trials. I will need to run the FFT for the frequency content.
Thank you,
Lee