[PD] vu-meter

[IOhannes m zmoelnig]

Atte André Jensen wrote:
> Hi
> 
> I'm trying to have vu show something useful, but with partial success. I 
> have this:
> 
> |
> [snapshot~]
> |
> [posetive]
> |
> [rmstodb]
> |
> [- 100]
> |
> [s $to_vu]
> 
> [posetive] is a simple abstraction that... makes negative values 
> posetive (1->1, -1->1, 0->0, etc).

you mean like the absolute value? like [abs]?

> 
> This kinda works, but at lower levels, and esp with only high frequency 
> content the meter falls all the way to the bottom. What would make a 
> better meter?

hmm, a single sample is not really representative for the signal, is it?
imagine you have a signal:
0 0.33 0.66 1 0 0.33 0.66 1 .....
which will have "0" each and every sample that is dividable by 4 without 
a remainder; since you bang the [snapshot~] only at block-boundaries, 
your vu will always display 0 (-inf)
if the signal is offset by -1 sample, your vu will always display 1 
(unity gain; full-scale)

so i don't fully understand your why you don't want to use [env~]?

[env~]
|
[- 100]
|
[vu]



fgadmrt
IOhannes