Hello!

I am working on a music visualiser project using an Arduino and RGB strip. On my first prototype, I used an LM393-type sound sensor which worked but was very limited in its abilities: it pretty much only measures 'volume' which made the light visualisation look like a total mess!

I started to dive into the vast world of microphones and ended up narrowing my picks to:
- SparkFun's sound detector as it has been used in similar projects
- Adafruit's MAX 4466 and MAX 9814 as they seem to still perform well at high decibels and has automatic adjustable gain

From what I gathered, I am thinking of using the envelope output on these sensors to create my visual effects. Passive band filtering also seems interesting, perhaps to separate bass/mids/highs.

Are these ideas relatively feasible? Did I miss out on crucial details or am overcomplicating things?

Thank you in advance!
Have a great holiday... A

 

Many ways to do this.

Digital signal processing from a fast ADC input would allow the most flexibility, but the math requires a LOT of CPU power, probably more that an Arduino can muster.

Down a notch would be to use a few analog bandpass filters, followed by envelope detectors. Feed these signals into the ADC, this would require FAR less CPU power.

 

I don't see an "envelope output" from either device.
If all you need is a beat-detector circuit, that is fairly easy to achieve either in the analogue or the digital domain.
Have you considered MEMS microphones?

 

I have an obsolete LM3915 IC in a "volume indicator" circuit that has 10 steps with 3dB between each step. Its LEDs show 10 very low levels then automatically switches when levels are loud to show 10 loud levels. My circuit uses a peak detector that shows all sounds, even short duration sounds.

Some people have 10 of these circuits in one project with narrow-band active filters (passive filters are too wide).