what is wrong in my understanding ???????????????

Archived from groups: rec.audio.pro (More info?)

Thanks for your knowledge which you all shared with me.
Now I have some answers why I am taking the above approach.
and some doubts in that appraoch also.

1. As it has been suggested that use

Mixsed_samples = a/n + b/n +........so on
what I think is that If we go in this way then we will lose the
amplitude of the Resultant Mixed signal. Hence for mixing the 10 files you will
have the amplitude of the resultant mixed signal will be low as compared to mix
the five audio files.

2. In my case I am doing the
let supose I have the 8bit signed data.

so I am taking the consideration of the fact.

if (a < 0 and the b < 0)

then misxed sample resultant will be equal to
mixed sample = a + b - ( a * b / - 128);

else i will mix in below fashion

mixed sample = a + b - a * b / 127;

In this there is no data loss also and the amplitude of the
mixed signal is also manitain (No dounbt if we are goint to add
more numbers of audio files then I will get the Hiogher Amplitude of
of the mixed signal. So there may be trde off that uptil this
number of signal You can mix with out geting the noise. As what I
think that after the certain number of audio signal (number of files)
u can have the noise in it. Which may be disturbing.

3. I am multiplying the a*b and dividing keeping in the consideration of the
Normalisation concept.


Now I get on more doubt in this.

supose one sample is -ve and other is +ve the also I am doing the
else equation (above) Now does this will give the problem ??????

Yaa I am writing my code i dont need any software for this.