When you are in MIXING you can activate the Dual Insertion Effect that a Voice has from VOICE mode. By that I mean, you do not edit the active Insert effect from the MIX - it is edited and stored with the Voice. On the screen where you are editing the effects of your MIX (MIXING/EDIT/F6 Effect/SF1 Connect). Notice in the upper left corner you can activate it on a PART. What Insert Effect is actually recalled depends on what Voice is in that part. For example, you have the Insert effect set to PART 01. If you recall a Guitar with heavy duty Amp Simulation as the Insert Effect will recall its original effect setting from Voice mode when you select it as the PART 1 Voice. If you select an Organ that has a Rotary Speaker effect for PART 1 - it will automatically recall that effect. The EFFECT is actual stored with the VOICE. A good analogy is: In a MIX the STUDIO has 3 system effects - each part with an individual aux send. One of the musicians can bring his (dual insertion) effect from home.
If you are working in Pattern mode. It is an easy thing to make audio tracks of key parts (those that need their insertion effect active) Record the part as MIDI with the Insert Effect assigned to that part - then resample that part as an audio track. Then you can reuse the insertion effect on other parts.

When you are in Voice Mode and are looking for the Insert Effect (VOICE/EDIT/F6 Effect/SF1 Connect) simply move the cursor over the diagram's INS1 (for the small Insert) and INS2 (for the large insert). Use the data wheel to select the effect algorithm. The MultiBand Compressors are in Insert 2. The small 'c' in front is for category. c:INS1 and c:INS2

If the Voice is a Preset and you want to change the effect settings, you will need to store it to a USER location and replace the PART in your MIX with your new variation.