You actually said "why are you so obsessed with smoothness as being the key differentiator?" I have no idea what this supposed to mean, so how can I reply to it? I originally made a jokey remark about "jaggy lightning bolts" that has turned into this nonsense. This was in reply that electricity is not smooth. But we both know that AC current is first rectified using diodes and then smoothed using capacitors. It is further smoothed as it travels round a circuit by other components as required. As for "stepiness", this was referring to ancient digital synths like the Korg M1. Here is an AI response about "stepiness"...............Gamma-UT wrote: Wed Jul 23, 2025 2:17 pmI don't think I'd quite be as strident about boasting about how anything explained to you goes, figuratively, in one ear and out the other utterly unimpeded.dellboy wrote: Wed Jul 23, 2025 2:03 pmYep, sitting here with my digital Kurzweil K2700 as a controller for my laptop with VSTs on it. A pair of Presonus monitors and a USB audio interface and headphones. My current setup. No "smooth" analog stuff at all. Completely in the box.
Need I repeat what I actually wrote (not that I don't expect you to ignore it completely like everything else)?
"obsessed with smoothness as a differentiator".
Not to mention that I thought I explained that digital synths are just as smooth. As are USB interfaces into headphones. But that nuance, it seems, completely passed you by.
AI Overview
Old digital synthesizers can exhibit "stepiness" and aliasing due to the limitations of their digital signal processing. Stepiness refers to the noticeable, quantized nature of the sound, especially noticeable in parameters like pitch or filter cutoff when changing rapidly. Aliasing, a form of distortion, occurs when the digital representation of a signal doesn't accurately capture its higher frequencies, leading to the introduction of unwanted, lower-frequency tones.
Stepiness:
Quantization:
Digital synths use discrete steps to represent values like voltage or frequency. When these steps are large (due to low resolution or sample rate), changes in sound can be perceived as discrete jumps rather than smooth transitions, resulting in stepiness.
Example:
Imagine a filter cutoff frequency being controlled by a low-resolution potentiometer. As you turn the knob, the filter frequency might jump between a few distinct values instead of smoothly gliding, creating a noticeable "step" in the sound.
Aliasing:
Nyquist Frequency:
In digital audio, the Nyquist frequency is half the sampling rate. Any frequency component above the Nyquist frequency cannot be accurately represented and will be aliased.
How it happens:
When a digital synth tries to generate a high-frequency tone (e.g., a sawtooth wave with a lot of high harmonics), some of those harmonics might exceed the Nyquist frequency and get folded back down into the audible range, creating unwanted lower-frequency tones that weren't part of the original signal.
Example:
A digital synth generating a high-frequency sine wave might produce a tone with added lower-frequency harmonics due to aliasing, making the sound less pure and potentially harsh.
Solutions and Workarounds:
Higher Sampling Rates:
Using higher sampling rates (e.g., 96kHz or 192kHz instead of 44.1kHz) increases the Nyquist frequency and reduces the likelihood of aliasing.
Oversampling:
Some synthesizers internally oversample the signal (e.g., 4x or 8x) before processing it, which effectively increases the Nyquist frequency and reduces aliasing.
Anti-Aliasing Filters:
These filters are used to attenuate high frequencies before they reach the digital conversion stage, reducing the amount of aliasing that occurs.
Analog Modeling:
Some plugins and synths aim to emulate the imperfections of analog circuits, which can mask the harshness of digital artifacts like aliasing.
Careful Sound Design:
By understanding how aliasing and stepiness occur, sound designers can choose waveforms and modulation techniques that minimize these effects or even use them creatively.
In essence, stepiness and aliasing are inherent limitations of early digital synthesis, but understanding their causes allows for clever workarounds and even creative applications.
