My younger brother has explained to me some of the intricacies of perceptual sound and yada-yada-yada
you’ll typically find that if the band for 18khz is showing any appreciable
level, that level will be reflected louder, lower down. Sounds with an
18khz component will almost certainly have matching (in terms of volume
envelope) components across a wide range (in linear terms) of frequencies in
the top end. For example – a hi-hat stills sounds like a hi-hat if you
low-pass filter it at 10k, the higher components aren’t really doing
anything different for our purposes.
…
So limiting your graph to say 60 > 12k or so would still be representing all
the zingy top and hefty bottom visually, but would do it more accurately (by
doing it less accurately, if you know what i mean!?!?).
level, that level will be reflected louder, lower down. Sounds with an
18khz component will almost certainly have matching (in terms of volume
envelope) components across a wide range (in linear terms) of frequencies in
the top end. For example – a hi-hat stills sounds like a hi-hat if you
low-pass filter it at 10k, the higher components aren’t really doing
anything different for our purposes.
…
So limiting your graph to say 60 > 12k or so would still be representing all
the zingy top and hefty bottom visually, but would do it more accurately (by
doing it less accurately, if you know what i mean!?!?).
So expect to see these ideas appearing in BetterBars.dll in a few weeks or so when I next get chance to touch it.