#38 Float32 encode functions

[Nicholas Humfrey]

Attached is a patch to add two new encode functions to LAME:

lame_encode_buffer_float32()
lame_encode_buffer_float32_interleaved()

Unlike lame_encode_buffer_float(), these two new encode functions use audio samples in the range +/-1.

This streamlines code that use float32s internally, as many pro-audio APIs do (Jack, Mac OS X's coreaudio, LADSPA etc).

[Nicholas Humfrey]

[Gabriel Bouvigne]

Logged In: YES
user_id=1056
Originator: NO

I am afraid that having 2 encode_buffer_float with different input ranges would be error-prone.

Perhaps we should have a look at implementing a generic encode_buffer, with some enums/flags to set input data type, interleaving, range.

[Nicholas Humfrey]

Logged In: YES
user_id=711618
Originator: YES

I don't see that it would be that confusing / error-prone if it was properly documented?

Having a generic encode_buffer is more like to be error-prone as it won't be possible to do type-checking/number of parameters at compile time...

It is just a shame that encode_buffer_float didn't use the same range of values as most other audio applications seem to now :(