What Is the Lowest (best) Possible Bitrate for a Speech Audio Track?

MP3 format was a hit, when it first introduced (still is). It compressed the audio CD tracks so well, that even dudes like me with slower internet connections could easily download a lot of em within a shorter period of time :). Although it's just not the only reason for its "such" popularity, because other than 14-16 track/song per a CD in an Audio CD... with MP3, you could insert hundreds of tracks in CD with the same capacity (geez, we didn't know that Gayan :) ). 

But MP3 is certainly not the best of its kind. Just because it's popular, it doesn't mean it's the best. In fact, did you know that the GNU licensed Vorbis codes is more efficient than the MP3 for a given a bitrate!. However, in my experience, Vorbis can take slightly longer periods to convert a track than the MP3. In that sense, they're both head to head. 

And on the other hand, MP3 is supported by almost all the players while Vorbis (or commonly known as Ogg which is not the real name, Ogg is the container format) isn't that "popular". Anyhow, let me come back to the topic.

Let's rock this speech :)...

Both MP3 and Vorbis are lossy codecs. Meaning that, once the encoding is done, it irreversible!. So, after encoding, the first thing that you did was deleting the original file and then realized that the encoded track sucks so much that it makes you puke... then my friend, no one can help you :). So, don't delete your original track before you listen to the encoded track and happy with its quality.

What is the lowest possible value/bit-rate that can be achieved for a speech audio track using MP3?

The higher the quality of your original vocal/speech track the lower the bit-rate that you can achieve. And remember, I'm talking about an audio track that contains a only a vocal track (a speech made by someone). And if you apply these below values to a song/music-track... it'll suck, big time!.

*. Bit-rate = 32

I can't give you how to change your bitrate since from app to app the GUI differs ... so you'll have to figure that out.

*. Also make sure to change the channels from 2 Stereo to 1 Mono.

This is again important because with such low bitrates sometimes the "stereo" effect might hurt rather than it helps. So by cutting it out and going for the Mono will come in handy while removing some of that background noises in my experience. However, I've seen in very few occasions... by having Stereo "on" helped... but most of the time you'll be better off without it. 

*. Change the audio sample rate from 44khz (44100Hz) to 22khz (22000Hz).

Sample rate is : the number of audio samples per second that are used to "create" the output sound.

So the lower it is the lower the quality it will be. However, 44khz is usually the standard for 2 channel stereo, therefore, when you change from Stereo to Mono (as said above), then by also reducing the sample rate to 22khz can also help to minimize the file size since 44khz is pretty much unnecessary since we don't have 2 channel stereo anymore.

If your source is not that high in quality, then try going for slightly higher bit-rates such as 48-64, etc, experiment.

So by doing those above mentioned changes, again if you have a high quality source file, not only you can get a much smaller file afterward, but it'd still be able to give you a nice output with minimal distractions as well. Good luck ;).