I've the impression that on 3,3V the MSGEQ7 readings are more unstable than on

Archive FastLED Archive
,
1

I’ve the impression that on 3,3V the MSGEQ7 readings are more unstable than on 5V. So I’m having a hard time filtering the data without loosing the snappy reactions to the beat.
In the video FastLEDs noise that is moved along X.
http://www.youtube.com/watch?v=W3s8V_b-w3M

2

The sparkfun page concurs with this, saying " The chip supply can be between 2.7 and 5.5 volts with 5 volts providing the best performance."

3

Are you feeding the audio signal directly via wire or are you using a mic ? Mics are not very linear …

If you are using it wire based check out this post:
https://forum.pjrc.com/threads/25878-Analog-signal-issues-with-migration-from-Arduino-UNO-to-Teesny-3-1-Help-requested

4

Thanks, @Juergen_Bruegl . I always feed the signal directly in. Beside the even more noise with a mic it takes arround 3 ms/m for the sonic wave to travel from the loudspeaker to the mic. I prefer to have this time already for computing the data = better responsivity of the animation.