Why does PeakMeter and FreqAnalyst show different peak level
-
- KVRer
- Topic Starter
- 1 posts since 11 Jan, 2009 from skopje
The max peak level that PeakMeter plug-in shows, is always higher than the highest max peak in FreqAnalyst plug-in, even when the precision and the speed are set to maximum. What peak level is more accurate, the one in PeakMeter or FreqAnalyst ?
-
Blue Cat Audio Blue Cat Audio https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=39981 - KVRAF
- 5821 posts since 8 Sep, 2004 from Paris (France)
Hi,
both plugins do not actually measure the same thing:
A peak meter measures the peak value of the entire signal. It means that the maximum value of a peak meter corresponds to the maximum value of the audio signal as a waveform (you could measure it by yourself looking at the waveform and visually checking it maximum value).
On the other hand, spectrum analyzers such as Blue Cat's FreqAnalyst show the intensity of various frequencies in an audio signal. The actual sound is the sum of all the frequencies, so the resulting peak value may be very different from the peak values of its components. There is no easy relationship between both values, it is very dependent on the content of the audio signal (the frequencies that are present as well as their phases): let's take a sine wave at 50 Hz with an intensity of 1, and another at 10000 Hz with an intensity of 1. The sum of these sinusoids is a signal which maximum value is 2, and yet on a spectrum analyzer you will see two peaks at 0 dB only (=1), one at 50 Hz and the other at 10000 Hz. This is the reason why the peak value is always higher.
So both are accurate but do not show the same kind of information.
both plugins do not actually measure the same thing:
A peak meter measures the peak value of the entire signal. It means that the maximum value of a peak meter corresponds to the maximum value of the audio signal as a waveform (you could measure it by yourself looking at the waveform and visually checking it maximum value).
On the other hand, spectrum analyzers such as Blue Cat's FreqAnalyst show the intensity of various frequencies in an audio signal. The actual sound is the sum of all the frequencies, so the resulting peak value may be very different from the peak values of its components. There is no easy relationship between both values, it is very dependent on the content of the audio signal (the frequencies that are present as well as their phases): let's take a sine wave at 50 Hz with an intensity of 1, and another at 10000 Hz with an intensity of 1. The sum of these sinusoids is a signal which maximum value is 2, and yet on a spectrum analyzer you will see two peaks at 0 dB only (=1), one at 50 Hz and the other at 10000 Hz. This is the reason why the peak value is always higher.
So both are accurate but do not show the same kind of information.
-
- KVRer
- 1 posts since 16 Jun, 2010
I have something of a related question to this. I'm trying out FreqAnalyst and I love it! I do however seem to get some odd level readings. To simplify things, if I put a frequency generator directly in front of FreqAnalyst with nothing else going on, FreqAnalyst seems to indicate a lower signal level that what is coming out of the generator. For a specific example, if I set a 1k tone at -20db, I can clearly see it in FreqAnalyst and if I zoom in to the peak of that signal it is always at 7-8 db LESS than the generated level.
I'm on PT 8.0.3cs1 on Win7 64. For further testing, if I do the following:
generator at 1k,-20db -> FreqAnalyst -> Sonalkesis FreeG -> PAZ Frequency
Both FreeG and PAZ show me precisely -20db and FreqAnalyst shows me about -27db.
Thanks.
I'm on PT 8.0.3cs1 on Win7 64. For further testing, if I do the following:
generator at 1k,-20db -> FreqAnalyst -> Sonalkesis FreeG -> PAZ Frequency
Both FreeG and PAZ show me precisely -20db and FreqAnalyst shows me about -27db.
Thanks.
-
Blue Cat Audio Blue Cat Audio https://www.kvraudio.com/forum/memberlist.php?mode=viewprofile&u=39981
- KVRAF
- 5821 posts since 8 Sep, 2004 from Paris (France)
Hi,
The value shown by a spectrum analyzer is not absolute, because it depends on how the analyzer has been calibrated. That's why we propose offset and slope parameters in our commercial versions: you might want to calibrate the analyzer for a sine wave, white or pink noise or whatever other signal that you want to use as a reference. That's why you may have different readings on different analyzer, depending on how they have been normalized.
But that's usually not a problem because what you are interested in is the shape of the spectrum and differences, not absolute values (which is of course not the case for peak level meters).
The value shown by a spectrum analyzer is not absolute, because it depends on how the analyzer has been calibrated. That's why we propose offset and slope parameters in our commercial versions: you might want to calibrate the analyzer for a sine wave, white or pink noise or whatever other signal that you want to use as a reference. That's why you may have different readings on different analyzer, depending on how they have been normalized.
But that's usually not a problem because what you are interested in is the shape of the spectrum and differences, not absolute values (which is of course not the case for peak level meters).
