Just curious, how are you calculating that? I would not think the output resistors on the bus drivers are parallel. They each connect to a separate output stage and the bus.

First, Science is good. Thats why we have a vaccine. Using science to say that one tone is better is not quite there yet....

The calculated bus impedance is the sum of the bus resistors, provided they are connected between an output driver and/or a switch that grounds them. The source of a bus would be either an opamp (virtual ground) or a send pot or pan pot, which should be 1/10 the value of the bus resistor. After that, it's simple math. All opamps have a point where the 1/n noise increase in gain from the bus value ( the op-amp gain is the sum, each channel gain is only it's resistor) starts to go up more than the 1/n noise calculation, meaning that say with a 2520, if you go below 1200 ohms, the noise will increase more than just the gain calculation, where a 990 is around 600 ohms, so it will appear slightly quieter on the same bus if it's below 1200 ohms, but will sound different.

On the larger APIs I designed, I actually lifted the bus resistor from the summing amp (with some trickery) so the bus amp was always the sum of

**only** the channels connected. That kept the noise down a lot. In the new FIX designs, I sum locally in the 8 channel bucket, then sum each bucket balanced into the master section, so the noise gain is never more than the equivalent of 16 channels, even in a 64 input console, plus I am hitting the bus amps at +4dBu instead of -10.