This really depends on your context. How are you representing your sounds? What have you tried so far?
I’d recommend taking a look at this guide as well: