About the decibel (dB)

The light example is a little more abstract in a sense.

It is exponential, but in base 2, not base 10. In base two exponential growth is a linear doubling in base 10.
I'm not following you. Linear is linear, exponential is exponential, logarithmic is logarithmic—regardless of base.
 
...an you please explain why some level meters show a doubled signal as a 3dB increase and use 10*Log(dB2/dB1) but others like SPL Meters show a 6dB increase and use 20*Log(dB2/dB1)...
Cliff touched on that in the fourth paragraph oh the OP. dB describes a power ratio. If you double the power going into a circuit, you get a 3 dB increase in the signal. But some people measure voltage to measure dB. As it turns out, if you double the voltage you apply to a circuit, you're putting four times the power into the circuit, and that's a 6 dB increase in power.
 
Cliff touched on that in the fourth paragraph oh the OP. dB describes a power ratio. If you double the power going into a circuit, you get a 3 dB increase in the signal. But some people measure voltage to measure dB. As it turns out, if you double the voltage you apply to a circuit, you're putting four times the power into the circuit, and that's a 6 dB increase in power.

Which is why power increases so rapidly relative to the volume increase perception. 6 dB SPL = perceived doubling in volume = 4X power.
 
Bingo.

But if you plot the first with the vertical axis being linear you get a nice exponential curve, 10, 100, 1000, 10000.

For the second you get a straight line, 2, 4, 8, 16.

No you don't. They're both exponential series. One is base 2, one is base 10. Neither is a straight line when plotted; one accelerates faster than the other. The series 2, 4, 8, 16 as you wrote it above isn't linear. Remember: a linear expression is ax+b. For a=2, b=0 the linear series yields 2, 4, 6, 8 for the series, not 2, 4, 8, 16.
 
The light example is the way:*2=*2*4=*4*8=*8*16=*16and so on...Right?A linear graph with exponential scales?
 
No you don't. They're both exponential series. One is base 2, one is base 10. Neither is a straight line when plotted; one accelerates faster than the other. The series 2, 4, 8, 16 as you wrote it above isn't linear. Remember: a linear expression is ax+b. For a=2, b=0 the linear series yields 2, 4, 6, 8 for the series, not 2, 4, 8, 16.

I was wrong.

Posts edited.
 
Great stuff, Paco. I learned something new just playing around with the frequency-ratio-to cents calculator.
 
This is funny because I remember getting my first digital audio device at work and trying to align it so that +4dBu=0VU was 0dBFS. I mean a meter is a meter, right? ;)

I figured out pretty quickly that -15dBFS with my reference level was about as high as I could go without clipping and didn't understand why for literally years. I just experimented with it until I didn't get any clipping. It makes me jealous the knowledge that is out there today vs. when I was coming up in this stuff. If I knew half of what I know today then I'd have known twice as much as I did back then (linearly). ;)

But what really sucks in my opinion is that you can get material that is -22dBFS, -20dBFS, -18 dBFS, -15dBFS....and they're all technically acceptable because no one bothered to specify a true analog to digital reference level standard. I never understood that. Hell if memory serves we had a system that tried to use an analog representation for the metering, but it was scaled for like -10dBFS or something so you'd constantly clip. Then we had to record and play it back and convert it several times to figure out what unity was and then figure out the offset for the bullshit meter.

I just set everything for -20dBFS = +4dBu these days and if it's off I live with it. Nobody seems to care anymore.

/old angry guy rant
 
This is funny because I remember getting my first digital audio device at work and trying to align it so that +4dBu=0VU was 0dBFS. I mean a meter is a meter, right? ;)

I figured out pretty quickly that -15dBFS with my reference level was about as high as I could go without clipping and didn't understand why for literally years. I just experimented with it until I didn't get any clipping. It makes me jealous the knowledge that is out there today vs. when I was coming up in this stuff. If I knew half of what I know today then I'd have known twice as much as I did back then (linearly). ;)

But what really sucks in my opinion is that you can get material that is -22dBFS, -20dBFS, -18 dBFS, -15dBFS....and they're all technically acceptable because no one bothered to specify a true analog to digital reference level standard. I never understood that. Hell if memory serves we had a system that tried to use an analog representation for the metering, but it was scaled for like -10dBFS or something so you'd constantly clip. Then we had to record and play it back and convert it several times to figure out what unity was and then figure out the offset for the bullshit meter.

I just set everything for -20dBFS = +4dBu these days and if it's off I live with it. Nobody seems to care anymore.

/old angry guy rant

It is strange there was no standard.

I use mostly -18dBFS = 0VU.
 
Many times we hear about "headroom" ....this was the first time that someone really explained what headroom is... Thanks a lot Cliff!!!!!!
 
Here's the thing. Even Cliff didn't quite get it right.
We need to first coincide that the understanding of vanadium phosphate first and only, but not a third, -db persay, is only congruent to an asficiated scale. Casein point to a more lateralixed comprehension.
Haha I'm only kidding. I wish I was smart enough to understand all that db stuff in the first post but it causes me a migraine so i gave up trying to get it.
 
Difficult to understand, yes. But not NEARLY as difficult to understand as how an incredibly busy gear designer finds the time and inclination to write a remedial Audio tutorial for his gear's users... instead of just saying "google it".

Very impressed.
That was "remedial?? :D
 
Back
Top Bottom