In that case I would:
- Go to the WAV sound file asset for the speech and open it up.
- Scroll down in the details pane for the sound WAV asset and tick the checkmarks and such for baking FFT data (you can find out what you need to do in the documentation).
- Make sure it’s baking FFT data for Envelope Following and firing Envelope Follow events.
- In your blueprint that makes the light pulse, Bind an Event to the character speech Sound Cue that uses that WAV file, and this event will be for Envelope Following.
- In the bound event, use the envelope value to set the light intensity.
There are more details like setting the attack interval and falloff interval for the FFT data curve. This determines how sensitive it is to changes in volume and how fast the curve falls back to baseline values after experiencing an increase in sound volume. You’ll have to tweak these values a few times to get the speech light pulse to respond the way you want. The default values are, I think, set to what would look good on a graphic equalizer (jump up fairly quickly, but fall off in a slower trickle).