Andrew R. Brown andrewrbrown.net.au Real-time Music and Sound with Pure Data vanilla A series that explores the basics of electronic music through visual programming.
I suppose for musical applications it doesn't matter, but in telecommunications, when AM is used for transmitting information, AM without the DC offset added to the modulating waveform necessitates coherent demodulation at the receiver (due to the envelope of the modulated signal not being proportional to the modulating signal anymore), which is more technically involved than direct/envelope detection.
I think that it is because the waveform is being used to modulate the amplitude and the amplitude conventionally goes from 0 (no wave) to 1 (full volume), so you want the waveform to go from 0 to 1 as well. If you use a bipolar wave to modulate the amplitude, the amplitude will go from -1 to 1 instead. There's nothing wrong with this, but for a sine wave, this will result in the amplitude going to zero twice per wave (once at the beginning, and once in the middle, at each node), which doubles the rate of tremolo, as you observed. But for different waveforms, multiplying by a negative amplitude can drastically change the sound (to see this, try multiplying the sawtooth phasor wave by a negative amplitude and use it as an LFO). To avoid this confusion, people often use the convention that amplitude stays between 0 and 1. hope this helps!
thank u!
Was there any particular reason why you translated to a unipolar waveform (6:00)? The only thing I noticed was that it halved the rate of the tremolo.
I suppose for musical applications it doesn't matter, but in telecommunications, when AM is used for transmitting information, AM without the DC offset added to the modulating waveform necessitates coherent demodulation at the receiver (due to the envelope of the modulated signal not being proportional to the modulating signal anymore), which is more technically involved than direct/envelope detection.
I think that it is because the waveform is being used to modulate the amplitude and the amplitude conventionally goes from 0 (no wave) to 1 (full volume), so you want the waveform to go from 0 to 1 as well. If you use a bipolar wave to modulate the amplitude, the amplitude will go from -1 to 1 instead. There's nothing wrong with this, but for a sine wave, this will result in the amplitude going to zero twice per wave (once at the beginning, and once in the middle, at each node), which doubles the rate of tremolo, as you observed. But for different waveforms, multiplying by a negative amplitude can drastically change the sound (to see this, try multiplying the sawtooth phasor wave by a negative amplitude and use it as an LFO). To avoid this confusion, people often use the convention that amplitude stays between 0 and 1. hope this helps!