For algorithm, I think it is almost the best. But for implementation, there might be another optimization. Remember, the choice of algorithm is much dependent upon the hardware architecture.

Regards,

Li xiaochao.Jeff ----- Original Message ----- From: mad-dev-request@lists.mars.org To: mad-dev@lists.mars.org Sent: Friday, March 30, 2007 3:00 AM Subject: mad-dev Digest, Vol 20, Issue 2

Send mad-dev mailing list submissions to mad-dev@lists.mars.org

To subscribe or unsubscribe via the World Wide Web, visit http://www.mars.org/bin/mailman/listinfo/mad-dev or, via email, send a message with subject or body 'help' to mad-dev-request@lists.mars.org

You can reach the person managing the list at mad-dev-owner@lists.mars.org

When replying, please edit your Subject line so it is more specific than "Re: Contents of mad-dev digest..."

Today's Topics:

1. Synthesis filter bank in libmad (Karthika)

---

Message: 1 Date: Thu, 29 Mar 2007 07:45:58 +0100 (BST) From: Karthika karthika_lakshmi@yahoo.co.in Subject: [mad-dev] Synthesis filter bank in libmad To: MAD mad-dev@lists.mars.org Message-ID: 333191.28381.qm@web7906.mail.in.yahoo.com Content-Type: text/plain; charset=iso-8859-1

Helloo, I am trying to understand the synthesis filter bank in the LIBMAD. I could not understand in what order the D coefficients are used to obtain the PCM samples. Also,half of these coefficients have some relation to the other half. Then, why all the coefficients are being used in D.dat file? What is the significance of even and odd pointers? How is the windowing operation between U vector and D vector taking place?

Also, I would like to know if we can further optimize the LIBMAD source code. For e.g, by using Cooley-Tukey algorithm for the calculation of fastdct() and dct32() . Could anyone suggest any further optimization techniques for porting LIBMAD on an ARM processor?

Any help will be greatly appreciated.

Thanks and Regards, Karthika Lakshmi.

---

Yahoo! India Answers: Share what you know. Learn something new http://in.answers.yahoo.com/

End of mad-dev Digest, Vol 20, Issue 2

---