Is there any way to automate an FX flush?

I'm trying to get FabFilter Timeless 2 (a delay) to repeat the last word of these phrases in a vocal passage. I can easily have the effect bypassed and then unbypass it on each word that I want repeated. Then I can bypass it again when I want to shut the repeated word off. The problem is, when I unbypass it again to catch the next word, the repeats from the last word are still going. It's like when you bypass an effect via automation, it saves the state of the effect just before it was bypassed. Is there any way to clear its state so I can repeat the next word without the repeats of the old word present? If not, is there any way to do this without rendering to audio? (im using Melodyne and I'd like to be able to go back and edit)

If you can automate the feedback of the delay then you can lower it to zero during the silence, which should flush the delay and back it up again before the next word comes in. Maybe you can also live with a lower feedback amount in the first place.

Here is how I would deal with it. Instead of automating "bypasss", instead automate the "feedback" or "regen" or whatever determines the delay fade. That will likely for flush the old stuff out. EDIT: Gofer beat me by a long shot WTF can't believe it took me 4 mins to type that lol.

Believe it or not I did try this. The problem with my specific case though is the following: I can't live with a lower feedback because I want the word to repeat at nearly the same volume and then abruptly stop when the next line comes in. And I HAVE to use the bypass to make it cut out abruptly and not catch any of the words of the next line. The problem with this is that when the FX is bypassed, automating the delay to 0 doesn't do anything because the FX is bypassed. And there isn't enough time before I bypass the FX to set the delay to 0 to clear it because it takes time to clear it, and as stated before I want an abrupt cut out.

Usually the send is being automated for this, the delay stays active all the time. You automate the e.g send volume or mute and send only the word to be repeated to the delay.

This going to sound crazy, but after seeing what you are actually trying to achieve, I have a suggestion from watching an engineer track an old band of mine. Instead of trying to figure out the best way to automate and what to automate, just use small loops of the word you are trying to "echo" and copy/paste them wherever you need them. This is digital audio, things that seem crazy or difficult often are awesome solutions. I use this trick sometimes, and now that I get the process, it's actually much faster than setting and automating a plugin. Another cool thing about this method is you can eq (or use any fx) on each item, if you want, for progressive tonal changes.

Lots of moved threads lately... This ends up in recording techniques while "Direct Monitoring Revisited" stays on the main page? O.o

Richie has a good strategy ... I offer a different idea. Can you 'slice' the track up, placing the various 'slices' on different tracks ... then automate the mutes to start/end for each segment. just throwing ideas out. [and this SHOULD be in the Rec Technique section i think] keep us posted :)