Skip to main content ↑ ↓

Notice

Please note that most of the software linked on this forum is likely to be safe to use. If you are unsure, feel free to ask in the relevant topics, or send a private message to an administrator or moderator. To help curb the problems of false positives, or in the event that you do find actual malware, you can contribute through the article linked here.
Topic: AM Frequency Analysis (Read 5432 times) previous topic - next topic
0 Members and 1 Guest are viewing this topic.

AM Frequency Analysis

Hello,

I am a newb to audio analysis and I'm hoping someone can advise me on how to solve a problem I've been unable to solve myself. I'm trying to determine the frequency at which a 19.2 kHz tone in an audio recording is being modulated at. I know that the modulation frequency is probably between 0 to 50Hz, and this frequency varies throughout the recording. The modulation waveform is probably a sine, but may be a square wave. I'm just interested in extracting the AM frequency over time, and I preferably want to have an accuracy of at least 0.5Hz.

I can see the 19.2 kHz peak in frequency analysis software, but I can't work out how I can process the data further to get what I'm after.

Any suggestions gratefully received. 

AM Frequency Analysis

Reply #1 –
Amplitude modulation should give you two additional frequency peaks, at the sum and difference of the frequencies. Look for small peaks just either side of the main peak. Locating them within +- 0.5 Hz will require a fairly wide window on your FFT.

AM Frequency Analysis

Reply #2 –
Quote from: pdq on
Amplitude modulation should give you two additional frequency peaks, at the sum and difference of the frequencies. Look for small peaks just either side of the main peak. Locating them within +- 0.5 Hz will require a fairly wide window on your FFT.

Thank you. By zooming in on 19.2kHz, and plotting as a spectrogram, I can see how the 19.2kHz AM frequency changes over time. In fact there are peaks at 10.2kHz +/- n.f (where f = AM freq., and n = 1, 2, 3...) suggesting the modulation is a squarewave.

Thanks again.

AM Frequency Analysis

Reply #3 –
Quote from: HRV on
Quote from: pdq on
Amplitude modulation should give you two additional frequency peaks, at the sum and difference of the frequencies. Look for small peaks just either side of the main peak. Locating them within +- 0.5 Hz will require a fairly wide window on your FFT.

Thank you. By zooming in on 19.2kHz, and plotting as a spectrogram, I can see how the 19.2kHz AM frequency changes over time. In fact there are peaks at 10.2kHz +/- n.f (where f = AM freq., and n = 1, 2, 3...) suggesting the modulation is a squarewave.


Or, the modulation actually is FM, not AM.

If the modulation is strong enough, you can tell the difference between AM and FM by looking for variations in the amplitude of the modulated wave. No amplitude variations in the modulated wave, then it is FM. If there are variations, then it is probably AM. 

This particular call can be difficult if the modulation is small, or there is an outside noise source that is influencing the amplitude of the wave.

The peaks on either side of the main wave are called sidebands. The main wave is called the carrier.

AM Frequency Analysis

Reply #4 –
I thought that if you FM modulate the carrier with a square wave you would get only two peaks, and if you modulate with a sine wave then you get one broad peak.

AM Frequency Analysis

Reply #5 –
Quote from: pdq on
I thought that if you FM modulate the carrier with a square wave you would get only two peaks, and if you modulate with a sine wave then you get one broad peak.


What you actually see depends on your FFT parameters and the modulating frequency.

If your FFT sequence is long compared to the modulating frequency, then you get a true line spectra. That means no broad peaks, just a true spectrum analysis.

If your FFT sequence is short or matched to the modulating frequency, then you may get what amounts to being the average of a series of snapshots of the actual modulated wave, which is what you are describing.