I've been trying to do some research on audio watermarking and what i wanted is to insert string into audio file in the transform domain (DWT), i've read some literature of DWT especially HAAR transform and try to implement it. So far i can embedd and extract watermark from the audio which is wav file beacuse i following this literature but apperently it is not lossless watermarking as i try to implement it. What i want to know is from your logical view is that literature really give a lossless method because it said that they got the analisis which mention it supposed to be lossless (BER equal to zero or near zero) when attacked. Or maybe i'm the one who implement it wrong, my analisis give me result that the proposed method is not lossless(sometime the result is wrong). If there is any embedding method that support haar and which is lossless, because i hardly find the literature that exactly pointing out audio watermarking (that insert some text to audio).If anyone know any lossless embedding method in transform domain please help me?

  • $\begingroup$ What do you mean by "BER equal to zero or near zero when attacked"? $\endgroup$ – Jim Clay May 14 '13 at 1:16
  • $\begingroup$ That mean even when attacked e.g mp3 compression, noise, etc. BER zero mean the watermark embedded doesn't damaged at all. $\endgroup$ – Wawan Ma-chun May 28 '13 at 14:44

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.