Discover MakerZone

MATLAB and Simulink resources for Arduino, LEGO, and Raspberry Pi

Learn more

Discover what MATLAB® can do for your career.

Opportunities for recent engineering grads.

Apply Today

Thread Subject:
is it necessary to taper the begin and the end of a time series before fft?

Subject: is it necessary to taper the begin and the end of a time series before fft?

From: Ha

Date: 15 May, 2014 07:41:08

Message: 1 of 4

Someone told me that if the value of the beginning is quite different from the value of the end of the time series, some unexpected results may rise.

Is it necessary to taper the begin and end to zero when using the built-in fft function of matlab?

Subject: is it necessary to taper the begin and the end of a time series

From: TideMan

Date: 15 May, 2014 10:14:15

Message: 2 of 4

On Thursday, May 15, 2014 7:41:08 PM UTC+12, Ha wrote:
> Someone told me that if the value of the beginning is quite different from the value of the end of the time series, some unexpected results may rise.
>
>
>
> Is it necessary to taper the begin and end to zero when using the built-in fft function of matlab?

Yes.
Google this: Gibbs Phenomenon

Subject: is it necessary to taper the begin and the end of a time series

From: Paul Mennen

Date: 15 May, 2014 20:52:12

Message: 3 of 4

> Yes.

To answer this complicated question with a "Yes" is certainly an oversimplification. This "tapering" you refer to is usually called "windowing". Windowing is not necessary and FFTs are often used without them. More often than not however, windowing is quite useful and allows us to see things in the results that would otherwise be obscured. To use windowing successfully it is usually important to understand a bit more about this deep subject. To learn more about it, you may want to get "winplt" (one of my file exchange submissions). Contained in there are two papers about windowing. The first is paper by Harris that is almost certainly the most cited reference on FFT windows. It's not that great for the mathematically timid audience however which is why I wrote the other included paper which will give you the basic ideas needed without getting into the deep mathematical questions.

~Paul

Subject: is it necessary to taper the begin and the end of a time series

From: Ha

Date: 16 May, 2014 01:04:13

Message: 4 of 4

"Paul Mennen" <nospam@mennen.org> wrote in message <ll39ds$6f8$1@newscl01ah.mathworks.com>...
> > Yes.
>
> To answer this complicated question with a "Yes" is certainly an oversimplification. This "tapering" you refer to is usually called "windowing". Windowing is not necessary and FFTs are often used without them. More often than not however, windowing is quite useful and allows us to see things in the results that would otherwise be obscured. To use windowing successfully it is usually important to understand a bit more about this deep subject. To learn more about it, you may want to get "winplt" (one of my file exchange submissions). Contained in there are two papers about windowing. The first is paper by Harris that is almost certainly the most cited reference on FFT windows. It's not that great for the mathematically timid audience however which is why I wrote the other included paper which will give you the basic ideas needed without getting into the deep mathematical questions.
>
> ~Paul


Oh, mathematician...... they are terrible aliens.

Thanks for the suggestions. I'll read the 2nd one.

Tags for this Thread

What are tags?

A tag is like a keyword or category label associated with each thread. Tags make it easier for you to find threads of interest.

Anyone can tag a thread. Tags are public and visible to everyone.

Contact us