How to segment a voice signal into frames of 256 samples?

How to segment a voice signal into frames of 256 samples?

Answer 1

[Honglei Chen]

Do you mean something like this?

x = rand(1024,1);
y = buffer(x,256);

[Anurag Pujari]

windowsize = Fs/10;
trailingsamples = mod(length(YourSignal), windowsize);
sampleframes = reshape( YourSignal(1:end-trailingsamples), windowsize, []);

Is it the way to do so?

[Honglei Chen]

You can do that too. Your command there (I edited it to show the format) requires only MATLAB. But you can use buffer command if you have Signal Processing Toolbox and you can achieve the same by simply say

sampleframes = buffer(YourSignal(:),windowsize)

Then if you really don't want the tail, you can remove the last column.

See the reference of buffer at

http://www.mathworks.com/help/signal/ref/buffer.html

[wafa derouaz]

Please, can you tell me how to recover the original signal after using 'buffer'?

[Walter Roberson]

Saad Rehman comments to Honglei Chen:

for matlab 2016a version FRAMING is done via BUFFER function. Thumbs up!

[Walter Roberson]

To recover the original signal after using buffer, use the (:) operator

buffered_signal = buffer(inputsignal, length_of_buffer);
reconstructed_signal = buffered_signal(:);

This assumes that no padding was needed and that there was no overlap used. If padding was needed then to reconstruct the original you would need to know the length of the original.

Answer 2

[guada]

i think i also have a similar problem, could you also please help me with this: given a with 7.238 second duration and fs=16000, i need 25 msec per frame and overlaps 10 msec every succesive frame.

Answer 3

[sufyan masood]

how to decompose audio signal in transfered domain ?