The 4-Cable Method (4CM) is just a way to physically hook up pedals / processors / modelers to an analog amp that allows you to control which pedals or effects are placed before the pre-amp and which pedals or effects are placed after the pre-amp in the amp's effects loop. The main reason to do this is flexibility & the full utilization of fx in "proper" (usual) places (for clarity). The analog amp MUST have an effects loop to do this. Of course you can put whatever fx wherever too... & some, like chorus, have no standard place whatsoever. Here's the basic setup with pedals: Guitar plugs into pedals that "normally" go before the pre-amp. EG: compressors, dirt, wah's. Then into main pre-amp in. Then the delays, verbs go in the fx loop to affect the sound after the input section (pre-amp) & then out to cab (or separate power amp /cab). It's a bit more complicated with modeler but is much more powerful: Guitar plugs into modeling unit. The modeler's FX Send output goes to amp's main input (pre-amp in) The amp's FX Send goes to the modeler's FX Return. Lastly, the modeler's main unbalanced output goes to the Amp's FX Return. From there it goes out to actual power section and to cab/speaker. I think modelers are best if using the 4CM because of how easy it is to substitute a large selection of modeled PRE-amps (vs. "amps") in place of the real one. You just need to have the modeled pre-amp DISABLED if enabling the FX Loop so you're not going through both, and vice versa if want to bypass the analog pre-amp & just use the modeled one. (Remember, usually you don't want an "amp" model to replace your analog pre-amp, you want a modeled PRE-amp only) But in some units (like Helix), you can actually have the Sends and Returns (of same OR different loop #'s) be separate blocks away from each other & you can be even more creative / flexible sandwiching (or entirely skipping) custom segments of fx, and a simple snapshot enables / disables both the fx loop blocks (therefore enables / disables the modeled one in place of real one). So to recap: If you want to use modeled amp, you DISABLE the fx loop block going into real pre-amp & ENABLE the model pre-amp block so that the signal never gets to the analog pre-amp input & only the modeled one is used & signal flow continues on. If using separate send / return blocks then obviously you have 2 blocks to make sure are disabled/enabled, but can sandwich the modeled pre-amp that never needs to be disabled itself if you want since it's only heard if the fx loop that is sandwiching it is disabled.