opus - Mar 2011 - Please reinstate CELT_SET_SIGNALLING_REQUEST

Hi Jean-Marc,

I'd like to lobby to put CELT_SET_SIGNALLING_REQUEST back into the 
public area. In our case the frame sizes are constant and we're not 
trying to be inter-operable with anyone else, so the in-band signaling 
byte is just wasted bandwidth. I know I could just modify the default 
value for "signalling" in the source code, but I'm trying to avoid
doing
that as much as possible. Thanks,

John Ridges

On Tue, Mar 22, 2011 at 3:30 PM, John Ridges <jridges at masque.com>
wrote:
> I'd like to lobby to put CELT_SET_SIGNALLING_REQUEST back into the
> public area. In our case the frame sizes are constant and we're not
> trying to be inter-operable with anyone else, so the in-band signaling
> byte is just wasted bandwidth. I know I could just modify the default
> value for "signalling" in the source code, but I'm trying to
avoid doing
> that as much as possible. Thanks,
You're really concerned with overhead in the range of 3200 bit/sec
(for 120 sample frames) to 400 bit/sec (960 sample frames)?

The ability to change sizes will probably be used by future encoders
to achieve greater quality,  and the ability to send mono frames to a
stereo RX will be used.

In any case, if this is that much of an issue you'd probably be best
off just skipping the byte and adding it back in on the remote side.
This doesn't require any changes to the library.