[mocaptoolbox] Re: Questions on using odd mocap data
- From: Burger Birgitta <birgitta.burger@xxxxxx>
- To: "mocaptoolbox@xxxxxxxxxxxxx" <mocaptoolbox@xxxxxxxxxxxxx>
- Date: Thu, 6 Oct 2011 10:40:25 +0000
Hi,
On 6.10.2011, at 12:47, Alexander Refsum Jensenius wrote:
> Hi,
>
> Thanks for rapid feedback, everyone.
>
>> "norm data" represents the
>> distance from the marker (in space) to the origin (without
>> direction), thus one column per marker.
>
> Ah, I see. Btw: there is probably a typo in the manual, since it says "type:
> structure type ('MoCap data')" under norm on page 57?
True! Thanks.
>
>> If you data type is "MoCap data", the functions you call assume 3
>> columns per marker in the .data field. Therefore, it would be a good
>> idea to change your type field to "norm data", if your data has only
>> one column, so that the functions will use the parts for "norm data"
>> (if available). However, the problem with mccumdist is that the
>> current version of the function doesn't support norm data - but this
>> will be fixed for the next release.
>
> Great, thanks. Another function that doesn't seem to work with norm
> structures is mcsmoothen.
Hm, I just tried it and it worked for my norm data. Could there be something
else wrong?
>
> Another question: since my data is coming from an audio file, I guess it
> could make sense to use the .analogdata field. However, I cannot find much
> info about how analog data data is treated or analysed in the toolbox. Did
> you implement anything for analog data? And, considering the great
> MIRToolbox, are there any plans for connections between the two toolboxes?
So far, there nothing really implemented for analog data - apart from some
rudimentary EMG data analysis. However, we have usually used a separate norm
structure for EMG data and not saved it on the analog data field of the
corresponding mocap data. Since the toolbox functions don't use the analog data
field, you would have to change the functions accordingly if you want to use
it...
Best,
Birgitta
>
> Best,
> Alexander
>
Other related posts:
