search

alddiaz / HBWT_AnaSynth

Python code for the Harmonic Band Wavelet Transform (HBWT).
https://sites.google.com/site/aldodiazsalazar/
GNU General Public License v2.0
3 stars 0 forks source link

请问如何使用做小波变换呢？ #2

Closed Ruohan1 closed 4 years ago

Ruohan1 commented 5 years ago

不太清楚文件的含义，请问如果用它做bandlet变换应该如何使用呢？

alddiaz commented 5 years ago

Dear Ruohan,

Could you please reply me back in English please?

Aldo

On Wed, Aug 14, 2019 at 5:16 AM Ruohan notifications@github.com wrote:

不太清楚文件的含义，请问如果用它做bandlet变换应该如何使用呢？

— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/alddiaz/HBWT_AnaSynth/issues/2?email_source=notifications&email_token=ACEGZUK4KRW2JTFWYVY37JLQEO5PDA5CNFSM4ILSH5PKYY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HFEQIGA, or mute the thread https://github.com/notifications/unsubscribe-auth/ACEGZUKCSXRYGYG6LP2GNNDQEO5PDANCNFSM4ILSH5PA .

Ruohan1 commented 5 years ago

sorry,,I don't know the meaning of the file. How can I use it to do the bandlet transformation by python?

alddiaz commented 5 years ago

Dear Ruohan,

First of all, sorry for the late reply and thank you for your questions. I am curious to know which project are you working on? I would be happy if you can write some more detail about it.

The HBWT_AnaSynth library is not known for being the best one documented online. However, I would try to explain you a bit further how to use it. I suggest you to read my publication on the subject for additional technical details: [1] A. Díaz and R. Mendes, “Analysis/Synthesis Of The Andean Quena Via Harmonic Band Wavelet Transform,” in Proceedings of the 18th International Conference on Digital Audio Effects (DAFx-15), 2015, pp. 1–4. https://www.ntnu.edu/documents/1001201110/1266017954/DAFx-15_submission_74_v3.pdf

Basically, you will work with two functions called "hbwt" and "ihbwt" as follows:

  1. From libhbwt.py file (the direct HBWT transform, as in Section 2.2 in [1]) you will find the function hbwt(x, h, g, P, N), where: x: input signal x(n) h: low pass wavelet filter coefficients (typically, a Daubechies wavelet filter of order 11). g: high pass wavelet filter coefficients (typically, a Daubechies wavelet filter of order 11). P: the number of MDCT channels or "sidebands" (typically adjusted to match the period of input signal x(n) ) N: the number of wavelet scales or "subbands".

  2. From libhibwt.py file (the inverse HBWT transform, as in Section 2.3 in [1]) you will find the function ihbwt(a, b, h, g), where: a: wavelet filter coefficients called "scale residue" at scale N per channel p (see Eq. (7) in [1]) b: wavelet filter coefficients called "expansion coefficients" at scales n per channel p (see Eq. (7) in [1]) h: low pass wavelet filter coefficients (the same used in Step 1). g: high pass wavelet filter coefficients (the same used in Step 1).

I hope it is a bit more clear now, otherwise, let me know in case of further questions.

Regards,

Aldo

On Thu, Aug 15, 2019 at 2:28 AM Ruohan notifications@github.com wrote:

sorry,,I don't know the meaning of the file. How can I use it to do the bandlet transformation by python?

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/alddiaz/HBWT_AnaSynth/issues/2?email_source=notifications&email_token=ACEGZUNU5S4IXRFQBYY2OWLQETSP7A5CNFSM4ILSH5PKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOD4K36RI#issuecomment-521518917, or mute the thread https://github.com/notifications/unsubscribe-auth/ACEGZUO55JIA56XVN4M4A4LQETSP7ANCNFSM4ILSH5PA .

alddiaz commented 4 years ago

Hej Ruohan1, I've just uploaded new code and example (including graphics). Take a look at it and let me know what applications are you interested into ;)