I see no assurance that any of the devices mentioned will do what you want because none of them mention the potential problem (EDIT: unless your problem is unique to that hardware). Although demos sound fine but maybe because they're kind of avoiding the problem or context. Or there is compression. The usual purpose of active mixing is to isolate the controls of one channel from another. In a sense, what you want here is almost the opposite, for the controls to interact but only when mixing but taken out when not mixing or only using 2 sources. EDIT: Essentially, you're looking for there to be a volume drop if you turn on an effect that has no volume, that would make up for if that effect did have volume. But there are still several ways to do this, hard to automatically compensate for everything. The basic idea is simple though, maybe even someone has done it. [original musings continue...] One way to do this is more of a blend control, not individual channels but 1 vs. 2, 1+2 vs. 3. An internal passive circuit surrounded by buffers and make up gain is the obvious circuit. The switches disable sources by floating them so there is a volume increase when fewer sources are selected. Then that needs to be compensated for. Not only that, but 50% of one sound plus 50% of another tends to not sound as loud, so ideally, the compensation is also adjustable. And there can still be need of seperate level trims. There is a way to do this with seperate levels for each channel, but again there are complexities. Another way would be to have several seperate mixes and switch between them. Full control but fewer switchable choices at one time unless many knobs. Always more knobs... Maybe the simplest would be a seperate circuit that decreases volume based on how many sources are used.