Add zero allocation helpers for ulaw

[dennwc]

This PR adds *To method variation for ulaw. These methods avoid allocation allowing the caller to reuse the buffer.

Also, add methods that operate on int16 slices for LPCM.

(upstream changes from livekit/sip)

[dennwc]

@zaf A friendly ping :)

[zaf]

Hello and thank you for the suggestion. I was already working on refactoring the code and I m thinking to introduce this feature there. This is the PR: #3
And this is the change that introduces the zero allocation functions: 10be6f3

Bench results:

BenchmarkEncodeUlaw-4     	   13824	     86166 ns/op	 472.39 MB/s	   20480 B/op	       1 allocs/op
BenchmarkEncodeUlawTo-4   	   14311	     82727 ns/op	 492.03 MB/s	       0 B/op	       0 allocs/op
BenchmarkDecodeUlaw-4     	   35204	     33762 ns/op	 602.82 MB/s	   40960 B/op	       1 allocs/op
BenchmarkDecodeUlawTo-4   	   45838	     25614 ns/op	 794.58 MB/s	       0 B/op	       0 allocs/op
BenchmarkUlaw2Alaw-4      	   78462	     14701 ns/op	1384.36 MB/s	   20480 B/op	       1 allocs/op
BenchmarkUlaw2AlawTo-4    	  101139	     11891 ns/op	1711.47 MB/s	       0 B/op	       0 allocs/op

Would you mind if we close this one and continue this work on the aforementioned PR?

[dennwc]

Sure! Will check that PR to see if anything is missing there. Thank you!