i have a legacy processed audio wav file. Another UN-processed copy of the source is processed by my algorithm. From both processsing blocks i get processed wav files as output. Now i want to do null tests, because i read (will soon quote the reference) null tests are suitable for objective loudness comparision. So merely sutracting is enough?? But merely subtracting is not working. I see high of difference value. So what other parameters should be considered for null-testing!

wav file descriptions:

Source : 16 bit pcm ,48Khz sampled,32 Seconds length

Output : Both 24 bit,48Khz sampled,32 Seconds length

  • $\begingroup$ Which property of the signal do you want to compare? Do you know if your algorithm introduces a delay? $\endgroup$ – pichenettes May 6 '14 at 8:51
  • $\begingroup$ Null tests are absolutely useless for loudness comparison. You can use them to check for audible equality. $\endgroup$ – Jazzmaniac May 6 '14 at 11:01
  • $\begingroup$ @pichenettes i want to compare loudness and spectrum! My part uses filter of order 6. Similarly the legacy code should use either 6th order IIR or FIR filter (i assume). But if the delays introduced are diffrent, then how to proceed? Any way to find sample delay? $\endgroup$ – nmxprime May 6 '14 at 11:50

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.