why separation audio channel always have result 16 bits per sample?

Hi everyone

I wanna ask you about audio separation

Why separation audio channel always have result 16 bits per sample?

This happens when I do a channel split from 2 channels to 1 channel the first audio has 24 bit-sample and 2 channel

The second audio has 8 bit-sample and 2 channel

when converted to 1 channel, the audio is changed to 16 bits per sample

Therefore, how do we maintain the sample bits?

For example, the original audio is 24 bits per sample with 2 channels, the audio is separated into 24 bits per sample and 1 channel

I use this code:

thank you

[y,Fs]=audioread('water.flac');
size(y)
% single_channel_from_y = y(:,1);
x=y(:,1)
size(x)
plot(y)
plot(x)
audiowrite('onechannel(24).wav',x,Fs);
% sound(x)

[Mathieu NOE]

Please check this , audiowrite by default will generate 16 bits audio if you export in wav format