| Author |
 Topic 
|
weskola
Location: USA
Posts: 24
|
 Posted - Wed May 07, 2003 2:14 pm
|
|
|
When finding the Average RMS Power for the purpose of performing some normalization or limiting, what should be entered as the Window Width? I see references to 0, 1 and 50 (the default in 2.1). The results will vary depending on the Window Width setting.
|
|
kylen
Posts: 290
|
|
Posted - Sun Jun 29, 2003 11:32 am
|
|
|
bump...
I'm wondering this same thing about the meaning and use of window width. If I find out I'll post back here unless someone comes in to clear it up!:)
kylen
|
|
SteveG
Location: United Kingdom
Posts: 6695
|
|
Posted - Sun Jun 29, 2003 12:04 pm
|
|
|
It's easy. Well, easy to explain what it means, and why time comes into it, at any rate. If you look at the definition of RMS, you will find that it is the Root of the Mean value Squared. Now, to get a mean value, you need to define the period you are measuring over - the mean of 1mS of a complex waveform is not neccessarily going to be the same as the mean value of ten seconds of it! So to compute the mean value, you have to specify the measurement period. If your waveform is assymetrical, then the RMS value is going to be rather more use than the peak value, because it will tell you rather more about the whole of your waveform.
Incidentally, the purpose of this is severalfold. With a continuous sine wave current, the RMS value tells you what the equivalent DC current would be for to give the same amount of heating effect. But with a signal voltage, it has other useful features as well - like it's always a positive value, because it's a squared number, and all squared numbers are positive. That's why you never specify + or - with an RMS value.
Anyway, that's why the window size is important.
_________________
 |
|
|
|
kylen
Posts: 290
|
|
Posted - Sun Jun 29, 2003 3:11 pm
|
|
|
| Quote: |
| If you look at the definition of RMS, you will find that it is the Root of the Mean value Squared. Now, to get a mean value, you need to define the period you are measuring over - the mean of 1mS of a complex waveform is not neccessarily going to be the same as the mean value of ten seconds of it! |
Oh, thanks SteveG. I'm trying not to go down the wrong rabbit-hole here so I'll see if I can repeat it back as I understand it now.
The mean value is the average over some time period, you define what that time period is for a single measurement by using the window width. Depending on the current wave selection in the editor window there will be multiple measurements of [window_width] ms each till the end of the selection area.
With a 50 ms window width, for any one time period, it is possible to come up with a fairly accurate RMS value for any wave in the audio spectrum of 20Hz thru 22.5KHz . This is because the period of a 20Hz wave is 50 ms while the period of a 22.5KHz wave is .04 ms so that range of frequencies would fit within a single 50ms window.
Finally all of the "means" will be collected together, depending on the length of the wave selection in the editor window on the monitor. That is how Average RMS Power is accumulated thru-out the entire wave selection. Minimum RMS and Maximum RMS are simply the lowest and highest of any single RMS measurement at [window_width] ms.
Thanks for the info!
kylen
|
|
|
|
SteveG
Location: United Kingdom
Posts: 6695
|
|
Posted - Sun Jun 29, 2003 3:42 pm
|
|
|
That's more or less it. But if there's any justice in the world, for any given window size, the Total RMS power for any given selection should always be the same, regardless of the window size, since ultimately, the 'area under the curve', which is what's finally being determined, will still be the same... but the maximum, minimum and average certainly will vary. You get a better idea of how it changes by viewing the 'Histogram' window, but you have to think carefully about what it's showing you.
_________________
|
|
|
|
kylen
Posts: 290
|
|
Posted - Sun Jun 29, 2003 4:37 pm
|
|
|
| Quote: |
| the Total RMS power for any given selection should always be the same, regardless of the window size, since ultimately, the 'area under the curve', which is what's finally being determined, will still be the same... |
Right, of course! Thank you for that too. I'm getting a pretty good visual if I don't shake my head around too much before it all hardens...:D
kylen
|
|
| |
Topic
|
All Forums
Average RMS Window Width