A/B Testing: Simple signal addition make sense?

[Blur]

I was thinking about how I used to take music files of different resolution and invert one and add them together to see what was left. Often shockingly nothing or just some artifacts in the compressed versions, but nothing really audible.

Now I got to thinking about how to easily do the same with sound systems such as 2 different complete setups for headphones. I'm talking 2 DAC's and 2 amps going into the same signal path into the same headphone. My idea is to simply invert the polarity of one of the setups and combine the two setups while listening to the headphones. Turning up the volume on both and then matching the volume where the sound disappears. Does this seem sensical? My guess is the hard part is keeping one setup from loading the other.

I expect there may be some timing errors as well, but if all of a sudden the signals overlap while inverted there would be some interesting results.

 

[solderdude]

That is basically nulling with 2 variables. The problem here could be a different phase response which, in the null, will result in an amplitude difference which could be more 'audible' than the reality and would only work with amps.
Small timing differences in DAC's will prevent nulling this way.
DAC's may be better compared with Paul's software.

Fine for cables and comparing op-amp circuits though.