@amirm Did the frequency response change in any way ?
Chances still are that low impedance headphones might change their behavior when the current/voltage drive ratio is different.
My suspicion is that the increased output resistance in the XLR out (2ohm) is a sensing resistor, so in series with the driver, and that voltage is used for feedback.
I may not have thought this through enough when a 300 ohm headphone is used that goes to say 600ohm at the resonance frequency the current will of course change at the resonance but across 2ohm a sensing voltage will barely be there.
As the trick here seems to be a mix of current + voltage feedback then it makes sense most of the feedback will come from voltage feedback and a very small amount by current feedback. So for high imp. headphones nothing substantial might actually change.
Since the level difference between 1/4 and XLR4 seems to be small (I expect it to be with just 2 ohm) the output voltage will be voltage determined.
That could be different for lower impedance headphones with a varying impedance (bass and/or treble). In that case the voltage drop across the 2 ohm (still assuming this is the trick) will be relatively higher and more influence could be there.
I don't know if this would be visible in distortion though as the corrections will be small and distortion is mostly SPL related but might be visible in FR (tone).
Is 'trickery' an objective term? The circuit sends feedback through the transducer. Trickery?
What feedback is 'sent through' the transducer and how is this achieved ? Do you know how this is done exactly ?
When you combine current sensing feedback with voltage feedback and no other amps (aside from the Bakoon) do this then I do not call that a trick as long as it is clear what exactly is done.
It may well be tricky to call this a trick but because what exactly is done is kept secret (as in not disclosed) and such is the same as the trick of a magician where he does not reveal what exactly is done. So to me it remains a 'trick' until it has become known what is done.
Amir just lifted a small tip of the entire curtain and when the frequency response did not change either.
Still an unknown but am sure the distortion measurements is sweeped so FR is also present and only needs to be plotted. The fact that Amir did not do this suggests nothing is visible there either.
At least now we know the Halo trick seems to de nothing spectacular for high impedance headphones.
I think for this trick to work for all impedances the device needs to sense the connected impedance and adjust the gain of the current sensing part (or sensing resistor value) accordingly.
But again... this is just how I think it is done. This is based on the very few components in the amp section.
