Originally Posted by Jambo
The capacitance determines the frequency response. The cap forms a high pass filter with your headphone speaker. The -3dB frequency of this filter is f = 1/(2 PI x Z x C) where Z is the impedance of your headphones. As you can see, a smaller capacitor will pass less bass - how much less depends on the impedance of your headphones.
It's not just the capacitor and headphones though. You also have to take into account the input impedance of the amplifier you're using. I think a typical amp has about 10Kohm to 50Kohm of impedance, so that in conjunction with the headphones determines the bass response. Assuming you're using Grados at 32ohm impedance and an amp with 50K input impedance and 47uF BGs,
f = 1/(2*pi*(50,000+32)*47uf)
f = 1/(2*pi*50,032**0.000047)
f = 0.0677Hz
Like dumbears, I went with a slightly smaller capacity 22uF BG NX Hi-Q, but even that has a corner frequency of 0.145Hz with this theoretical setup. I think typical CMoy setups have around 20K input impedance. I'd say worry about the amp's impedance rather than the caps.
dumbears:
As for your iPod photo, if you could take an aerial shot of the WM8975 perhaps I can help you a bit better. From that angle I can't see the necessary traces. I don't think shorting the caps will make that much of a difference, as the goal is to completely avoid the internal circuitry. I haven't looked into it extensively, but I imagine you could solder the wires from the traces to the iPod's dock. I hadn't even thought of doing it that way. I'll look into it tonight. It would make things a lot cleaner. *wanders off in thought*