I need a Babel Fish - but I suppose now, maybe "just" similar circumstances, as I have...
> Linear or logarithmic are just ways to display things. The underneath substance
> doesn't change.If you have a low input volume, logarithmic bars will not
> help you. The audio triggers volume bar is the representation of the signal
> power, based on the data coming from the audio input line. I can assure you
> that with a good input signal that bar reaches 100%.
If I'm not deaf till then!
I have to await further next Live-Band chance,
if alike 70-100dB sound pressure at FOH were enough while my capture gain is already 100%...
About Low Level some shots: Audacity in tandem, linear (to close out my hw) - couldn't catch best glimpses but IMO you'll see though...
* just a little bit to sluggish? I don't know - this latency is perhaps in limited extent a tweak worth (by a knob on surface)
* wrong divisor 256? too lazy to search & I think you're too genious for such a blooper - but maybe also a tweak worth, as a kind of gain?
log scaling:
Your words remember me to "Only watch, don't touch!", well-known sayings here...
Sure we're talking about the art of visualisation - but those values you can
display, you can't use? - Tell that to the marines, my Dear!
If this was not enough for the initial spark, Fasten seatbelt, pls - it' s hard enough to explain my trains of thought in mothers language alone,
but my genuine audio engineer could join me at least (and by 2nd beer lately):
We are talking about float values between 0 and 1? Otherwise here even 0 to 100%
If you'd compare, perhaps by help of Audacity the difference:
a) linear has a low --- lets say "standard level of noise" but also a great
level difference to the peaks (which is normally fine, this should give good handlings to the 2 thresholds)
b) log dB has a higher base level - disadvantage is the lesser difference to peaks, but maybe though,
on lower input levels the better choice? I can't cover the whole dynamics in imagination
We didn't talk about that formula, 20.0f * log10f( x )
If result at x=1 corresponds to 100%, the remain computation is just proportional, rules of three,
cancel & divide till you're lucky again between 0 and 1 and continue with the trigger itself!
All in all just a little bit algebra, I bet you can make it.
As we're looking for a good beat detection for Live Rock usage:
This is best done in bass freq.bands, maybe bass drum, often snare, toms - but bass guitars have much hum, mids / treble
(who by all needs want use, should do - but here IMO more impractical, too blurred).
And there are "sister" effects (carpenters expression -> results *2 ), to deal with too.
I tried already, to feed a speed-widget tap, hopeful that there's only one signal at upper threshold. Seems to work fine...
(the speed dial itself could contain all my color & movement chasers, also EFXs by another factors, "nice to have": therefore also a "live" knob maybe by script))
Perhaps those "canned Light" DJs are happy with the various possibilities in assignment,
but I think I'd be more lucky with this my "one beat only Tap" assignment and which freq.band to use as Input, I should have parametric on surface, by live selection available - because most of us can only estimate needed frequencies roughly afore.
threshold(s) of course, too...
"pre adjustment" in the count of bands or bandwidths, in the properties is IMO already ok like it is (a "nice to have": live adjustment here, too)...
The "attack" has to be immediately, but also an adjustable "release", decay or hold (mostly less than the strokes, but in moods sometimes also await 2 or 4 strokes...), to get near a natural beat I want to use. -
Cheerio, Controllore di luci - If it moves you, try to emerge it!
PS: He, the audio engineer is the man with the hat in front of me, look ska...
