#208 Optimize and correct linear2ulaw conversion

[Stefan Brüns]

The attached patch corrects a small error in the conversion function. According to the standard, the quantization invervals are symmetric to zero. At the moment, negative values have a negative bias.

Second, the new version makes use of faster compiler intrinsics, depending on some defines. By default, no intrinsics are used.

[Stefan Brüns]

[Craig Southeren]

Logged In: YES
user_id=774109
Originator: NO

Can you provide information on how to set the defines to enable these optimizations?

[Stefan Brüns]

autoconf test macro

[Stefan Brüns]

Logged In: YES
user_id=691769
Originator: YES

File Added: bsr_ac_test.txt

[Stefan Brüns]

Logged In: YES
user_id=691769
Originator: YES

Put the attached autoconf test macro into opal/configure.ac and add an "#include <opal/buildopts.h>" to g711.c

[Craig Southeren]

Logged In: YES
user_id=774109
Originator: NO

Applied