audio convolution problem

[dflarsp_]

hi guys,
i'm relatively new to dsp but im trying to learn. I wrote a convolution function for audiofiles in matlab. Just out of interest i convoluted white noise with a discrete dirac, and compared the frequency response of the convoluted white noise (outputsignal) with the orignial white noise (input signal). for frequencies around 20 - 200 Hz there are very small differences. with higher frequency the responses become identic, like it should be. the length of the convoluted signal is signal + dirac. i noticed when the length of the output signal is equal to the length of the input signal, i get a completely identic frequency response. I would greatly appreciate if anybody could explain this to me!

regards

 

[albert22]

I think that if you are convolving with the discrete dirac (unit impulse) the frequency response should be the same. Because the FT of the unit impulse is 1 and the convolution in the time domain maps to a product. Then you are multiplying your signal by 1 in the frequency domain.
Are you saying that the response differs for low frequency when you take the full output and it is identical if you truncate your output?