As mentioned in my
initial posts (I was the first one pointing out the DC issue), DC protection doesn't work at hardware level, and software level fix won't work for hardware issues. Here is the reason to that.
In a HP amp, DC protection is done via measuring the DC level of the output. It is usually very low cost to do so, as it only needs one capacitor, one resistor, and one comparator, so some products do it per channel if they have ample space on the PCB. But the majority of the products sum L and R channel and let it run through such circuit.
The cost cutting way still works reliably in real world. If your music is poorly mastered, or your amp is running into hardware issues and triggers DC, usually it's either L or R channel having the problem, or both channels having the problem but the DC levels don't cancel out when summing L+R. Unless intentionally created, you don't find music that has positive DC on L and negative level DC on R. In addition, when your amp breaks, it is also very rare that your L channel outputs +V and your R channel outputs -V. So this is very reliable in real world.
But topping goes further by summing the 4 balanced outputs (L+/ L-/R+/R-) together in their DC detection circuit. They went too far. There are plenty of situations this will never work. Suppose your music has DC in the L channel, it got cancelled out to 0V by adding L+ with L- phase. After I pointing out the problem, topping released a software solution, which works fine against problematic music tracks at digital level. However, when your amp malfunctions, it cannot guard against it. Suppose your DAC IV's opamp broke and output +V on L channel, or your balanced output then outputs +V and -V, as the negative phase is produced by reversing the positive phase, and after summing the two phase it become 0V. So the hardware will never be able to detect such fault.