The 1/8 rule isn't a rule, it's a recommendation. The frequency response will change with any output impedance, even 0.1 ohm. 1/8 of the headphone impedance just happens to be a good number to aim for that achieves inaudibility almost all of the time. Most full-sized headphones will cope with much more. Some IEMs might need even lower.
Also AFAIK planar magnetic drivers have flat impedance curves. There was some debate in another thread about whether or not damping factor mattered at all for them.
Yup, and if anybody has looked at impedance curve of dynamic drivers of lower impedance for portable over ears or iems, you will see flat impedance. Even look at the Sennheiser Momentum, it's flat. When you get to high impedance designs like the Beyers or Sennheisers with 300ohm nominal the impedance curve shows significant resonance hump. That hump is resonance where the peak point is where the inducatance and capacitance meet and cancel each other out to be resistive. That means the left side where it rises is inductive and the right is capacitive. The Q factor of the resonance changes with the added resistance which shape the resonance. Within the resonance area is where the added resistance affects the FR. And a measurement I've seen of the HD800 shows that it happens only when the resistance is pretty significant like a value that is equal to it. 180ohms shows only slight boost. And, there was minimal FR changes in the region that is close to flat in the mids to high frequencies even with 600ohm output impedance. Also I read that if you short out the speaker terminal, the cone resists movement, which means that with a resistance in between, it moves more freely. That means the output impedance would cause less damping and cause less control of the driver. Close to 0 output impedance would be like a short at the terminals creating maximum damping. I believe stv posted some distortion measurments of the Beyers showing much higher distortions at the resonance frequencies with added output impedance.
I think damping factor is for full sized headphones or speakers that has the characteristics like the Sennheisers or Beyers with the resonance hump and only matters because of the resonance. I think low impedance iems with non-flat response is different. Not sure if "damping" really applies to BA iems with skewy impedance response. Since they are uneven some parts are going to affect the power output relative to others that would change the FR compared to output impedance close to 0. I would think the reactive areas would be affected differently from resistive areas, but don't know how.
Edited by SilverEars - 10/5/14 at 9:37pm