Following the path through the I mixer, users will get to node 3. This path is really the same as if they were simply using a traditional down-converting mixer, and it is apparent that the RF and the UI signals have both been mixed down to the IF frequency and are sitting on top of each other. This is where the degradation in performance would result, as the noise and spurious components from the UI would add to their IF signal. There are actually two terms to the mixing process; at the sum and difference frequencies. The sum term will be eliminated in the signal chain with low pass filters, so they can disregard it in the math. The only term left, at the IF frequency, is the frequency difference term.