The Schiit Midgard is Schiit’s new affordable, balanced solid state headphone amplifier, replacing the Magnius. The topology is “Fully discrete, DC coupled, highly linear open-loop with low loop feedback, Halo™ topology mixed-mode feedback for Neutrik 4-pin output”. It outputs 4.8W into 32ohms and 750mW into 300ohms.
Unique to this model, Schiit’s Halo “is a mixed-mode feedback topology that incorporates your headphone driver into the feedback loop, offering the potential for better control at driver resonance and improved acoustic results”.
At the time of writing, Schiit is literally giving this amplifier to anyone with a measurement rig who is willing to publish their results with Halo. Full disclosure, that is how I got my Midgard. These are the results I have seen so far measuring multiple headphones using a MiniDSP EARS system.
Halo requires a balanced headphone connection, so comparing the Midgard with/without Halo is just a matter of plugging the headphone into either the single ended output or the balanced output. The output level of the amplifier remains the same between the two.
The EARS system is not the most accurate measurement rig. However I did confirm with Schiit that it was sufficient to meet their requirements before asking for a Midgard. I think it should be sufficient to compare two measurements of the same headphone.
I measured 4 different headphones, 3 dynamic drivers and one planar. All of them measured the same FR between the SE and BAL outputs. One headphone, the Sennheiser HD800S, also measured the same distortion levels. But the other 3 headphones showed a small, but consistent, lowering of distortion when connected through the BAL output that engages Halo. In order to see the advantage consistently, I needed to measure at a pretty high level 100db or more. I don’t know if this is due to the Halo effect needing high output to make a measurable difference or it was a limitation of my equipment and/or environment.
With all that said, here are a couple of distortion plots that illustrate what I am seeing. I don’t want to mention the specific make/model of headphones because of the limitations of my rig and that information isn’t relevant anyway.
This measurement is representative of the 2 dynamic driver headphones. Both have < 100ohm impedance and high sensitivity, unlike the HD800S which didn’t show any difference.
This measurement is of the planar headphone. Its has < 100ohm impedance and low sensitivity.
So does Halo work? I’d like to see what others are seeing before making any concrete conclusion, but it looks like it does give a small, ~0.3db, advantage in THD for some headphones. Would that be audible? I doubt it. But there is the possibility that on a more accurate measurement rig, the difference is shown to be more significant. And who knows? Maybe Midgard 2 will have a more effective Halo.