If you split the supply with zener diodes, you'll need a capacitor. Zeners are noisy.

And now I'm thinking your microphone is broken. I can get 4 volts p-p out of mine and you can barely get yours to register screaming on a "Line in" connection? What's wrong with this picture?

 

I tried garageband, If I scream, with the software amplification all the way up, it registers slightly which you can hear if you turn your speakers all the way up. Its not good enough for my application :/. Just out of interest, are there any plugins etc that can amplify further?

You've made me curious, so I had to break out my old gear and experiment. My old Shure SM58 mic (labelled unidirectional dynamic, 50 and 100Ω dual impedance) gave results similar to what you describe, although I could get a pretty good response to normal voice if I eat the mic. But not really what one wants.

I tried Guitar Rig and it does give more gain, just as I remembered it. But I wouldn't say the result is any more useful. It was still very sensitive to having your mouth right on the wind guard. Farther away there was almost no measurable response.

So my mic definitely puts out less signal than my Stratocaster, which works OK but just barely when used directly on my Mac's line-in port. The mic signal is detected but not really useful.

 

I tried garageband, If I scream, with the software amplification all the way up, it registers slightly

This is where I get curious about the mic being faulty because I can get 4V p-p out of my mic under the same conditions, and that's generous compared to the specs for "Line in".

Just repeating myself in case what I was referring to got lost in translation.

I still agree that normal conversational talking will need amplification because normal talking is about 40 db lower than singing in public. That's down to about 1 mv p-p coming out of the mic.

Just for something to compare to,
My 'Paul can do 6V p-p if I hammer a full, 6 string chord, but single note picking never produces more than a volt p-p. It's usually a lot less, like a few tenths of a volt p-p on single notes.

 

This is where I get curious about the mic being faulty because I can get 4V p-p out of my mic under the same conditions, and that's generous compared to the specs for "Line in".

Just repeating myself in case what I was referring to got lost in translation.

I still agree that normal conversational talking will need amplification because normal talking is about 40 db lower than singing in public. That's down to about 1 mv p-p coming out of the mic.

Just for something to compare to,
My 'Paul can do 6V p-p if I hammer a full, 6 string chord, but single note picking never produces more than a volt p-p. It's usually a lot less, like a few tenths of a volt p-p on single notes.

I haven't got the parts yet to build the whole circuit but I'm curious. Is there a given specification for the "line in" to recording devices? If so could you please link me a copy or tell me where to find because I can't find a standard specification anywhere.

I don't think the mic isn't broken since it works with a desktop computer (or atleast did) which I no longer use. But using a mic in port rather than line in.

I also looked in some mac forums and managed to find this for an earlier macbook pro:

"During input of a 1 kHz, full-scale 1 VRMS sine wave (44.1 kHz input sample rate, 24-bit sample depth, 0.0 dB input gain, no weighting) the audio line input has the following nominal specifications:

Jack type: 3.5 mm stereo
Maximum input voltage: 1 VRMS (+2.22 dBu)
Minimum voltage input for full scale output: 31 mVRMS (-27.95 dBu) at input gain = +30 dB
Input impedance: > 20 kΩ
Frequency response: 20 Hz to 20 kHz, +0.5 dB/-3 dB
Signal-to-noise ratio (SNR): > 90 dB
Total harmonic distortion + noise (THD+N): < -75 dB (0.02%)
Channel separation: > 80 dB"

 

http://en.wikipedia.org/wiki/Line_level