Unless I'm wrongly interpreting Schiit or Amir's measurements, it seems that for 33 Ohms load the two measurements seems identical, but for 300 Ohms load some differences arise.
There is
1dB difference between the 300 ohm measurements at the onset of increased distortion between Schiit and Amir's measurements at 300ohm.
It would appear that the one Amir tested could reach 9V in 300ohm where the one Schiit tested could reach 10V in 300 ohm
Let's analyze.
The max output voltage of the used opamps as well as the voltage rails determine the max output voltage.
There is one opamp for gain (OPA1662) and some opamps in unity gain as power buffers (OPA1688)
Data sheet shows that at the knee where distortion starts to climb the output top voltage is Vs -0.7V for the OPA1688
Data sheet shows that at the knee where distortion starts to climb the output top voltage is Vs -3.7V for the OPA1662 (THD+N measurement)
Using SMPTE / DIN: Two −Tone 4:1, 60 Hz and 7 KHz test = Vs-1.6V
Vs = power supply rail voltage.
This means the output devices won't contribute to the max output voltage swing so the OPA1662's voltage rail determines the output voltage.
To reach 9V you will need 16.4V rail voltage.
To reach 10V you will need 17.8V rail voltage.
The rail voltage is determined by 2 resistors of the regulators so can easily be changed by selecting different resistors.
So.. my best guess would be the Schiit tested device was probably a prototype where the rail voltage was closer +/-18V.
On the PCB it states power supply voltage = 17V so perhaps the production models have a power supply rail voltage closer to +/- 17V which then would explain the 1V difference in output voltage.
The power in the 30ohm load is not determined by the rail voltage in this amp but output current limited (OPA1688 + series resistor voltage drop) so will be the same despite the voltage rails being slightly different.
As Amir got a production run model I would trust Amir's measurements over the ones taken by Schiit. These could be obtained during prototyping.
On Schiit's website, it is said that at 300 ohms the amp's power output is 410mW RMS per channel ... but this review says max power at 300 ohms is 274mW ??? Can someone explain why this is ?
Max power rating on the schiit website:
Maximum Power, 300 ohms: 410mW RMS per channel
The trick here is that distortion isn't mentioned. When looking at Amir's measurements we can see that the 410mW is reached but at 0.02% distortion. So there is no disagreement about output power at all. Both measure 410mW in 300 ohm but Amir shows the value at the onset of clipping.
Of course, we cannot hear 0.02% distortion when our headphones, at those levels are doing much worse at those voltage levels.
So there is no discrepancy there.