VCO vs DCO Oscillators Objective Differences
VCOs vs. DCOs - Objective Differences in Sound
VCO = Voltage Controlled Oscillator
An analog voltage controlled clock controls the frequency for an analog waveshaper. The analog clock takes the input (key strike), and outputs the frequency/wavelength to the analog waveshaper. The waveshaper determines the shape and harmonics of the output (ie: Saw, Triangle, Pulse/Square)
DCO = Digitally Controlled Analog Oscillator
A digital clock circuit controls the frequency for an analog waveshaper. The digital clock takes the input (key strike), and outputs the frequency/wavelength to an analog waveshaper. The waveshaper determines the shape and harmonics of the output (ie: Saw, Triangle, Pulse/Square)
Similarities Between DCOs and VCOs
The waveshaping in both VCO circuits and DCO circuits is analog. The waveshaping is where the unique peculiarities and imperfections of waveshape come in, with their unique harmonic output. Since both have analog waveshaping, the general difference here is negligible.
Differences Between DCOs and VCOs
The difference is how the analog waveshaper is controlled - either via analog voltage clock, or a digital frequency clock circuit.
The Reason that DCOs Were Created
VCOs have notoriously bad performance when it comes to accurately rendering out a specific pitch, especially when you take measurements over multiple octaves and try and match up multiple different voices. As electronic manufacturing has progressed over the past few decades, things have gotten better, however even modern VCOs have inaccurate performance across multiple octaves and voices.
In the late 1970 to early 1980s, as the polyphonic (multiple voice) synthesizer market blossomed, this was a major issue. With Mono Synths (1-Voice Synths), the poor pitch tracking of VCOs was not as much of an issue. Synthesists could tune their mono synths for the general octave range they were playing leads for, and the synths fit right in the mix of songs.
When building multiple voice synths, like the 6-Voice MemoryMoog, 5-Voice Prophet 5, or 8-Voice CS80, the poor VCO frequency perfomance became a major problem. With multiple voices, each containing multiple oscillators that were nominally supposed to be in tune, the problem was very evident. DCO circuits solved the problem for developing multiple voice poly synths that were in tune across multiple voices and over multiple octaves.
Objective Audible Differences of VCOs vs DCOs
There are some misconceptions of the exact nature of differences between the two Oscillator circuits. Many people only rely on subjective terminology, or say the difference is "oscillator drift". Below, we will breakdown what "oscillator drift" really is. In reality, "drift" is not the best terminology to use, as there are multiple factors, and some involve motion, while some are more static. Also, there is variance based on intonation. There are two primary characteristics that can be measured when comparing polyphonic VCO synths vs polyphonic DCO synths. They are both related to the frequency that is output:
1. VCO Frequency Jitter (High Frequency Motion)
VCOs exhibit a high rate, but low amount of frequency jitter. The total range is very small - usually less than two cents (2/100ths of an octave) movement, but at a fairly high rate - usually between 5-12 hertz. The exact amount of this frequency jitter varies from VCO synth to synth, and is sometimes more fluid, and sometimes more random. We have taken measurements from dozens of different VCO synths - both monophonic and polyphonic synths dating back from the 1970s, up to modern day VCO synths. All of the synths we tested exhibit this VCO frequency jitter. It is universal to VCO synths. The audible effect of this VCO frequency jitter is very subtle, and probably not noticable to most people. At the fundamental its nearly impercetable, however, if you view a spectral graph, showing all the harmonics generated, you will notice that in the upper harmonic range (ie: 5khz and up), the effect of the frequency jitter is more noticable. There is a significant amount of movement in upper harmonics. Again, the audible effect of this movement is very subtle -due to amplitude dropoff in higher harmonics. This movement does "blur the harmonics" a bit. Rather than having perfectly crisp upper harmonics, there is more motion, which softens the sound in the upper frequencies.
2. VCO Frequency Accuracy Over Multiple Octaves and Multiple Voices (Stable Offsets in Tuning)
When you strike a key on the keyboard, a signal is sent to the frequency clock circuit to determine the target output frequency for given oscillators. If you strike a A4 note, the nominal frequency that should be output is 440.00 hertz. With a DCO synth, you're going to get that nominal pitch output to the waveshaper. With a VCO synth, you can tune it and generally get pretty perfect pitch for a given note. However, there may be a small imperfection to the output, even if freshly tuned. You may get 439.4 hertz output, or 441.1 hertz, where 440.00 is the nominal amount. Small differences in temperature or humidity may cause this imperfect performance.
Intonation Offsets (Detuning Offsets that Vary Based on Distance from a Nominally Tuned Point)
The bigger issue is that VCO voices are generally tuned for a specific section of the keyboard. You may get close to perfect tuning around A4 note and surrounding octave, however, if you play up and down the keyboard, and get down to A1 or up to A8, you will most likely notice that the oscillator is more significantly out of tune. In older synths, this effect is especially noticable. We tested several classics, like the Yamaha CS-80, MemoryMoog and Oberheim OB-X. These older poly synths may have up to 20 cents or more detuning range, when comparing a single oscillator over many octaves.
This is known as "Intonation Detuning" - the amount of detuning varies +/- based on how high or low on the keyboard you are playing. You can tune the oscillators, but generally you'll get a "sweet spot" somewhere in the middle, with detuning as you play different sections of the keyboard. This intonation based detuning performance was the most common type of detuning we experienced from testing various VCO synths, classic to modern. We found that most VCO synths get progressively more sharp as you go down the keyboard (relative to nominal target pitch), and relatively more flat when you go up the keyboard (relative to nominal target pitch)
Some poly synths have different voice allocation schemes, which varies the measurable output depending on how many notes are currently held down, or which specific notes are held down. For instance, the Roland Jupiter-8 features a couple voice allocation modes - rather than a round robin voice allocation, the Jupiter-8 can be set to always start on Voice 1 and progressively use up additional voices as new keys are held. The Prophet 5 and Prophet 10 feature voice allocation based on the actual keys struck - each key on the keyboard is assigned to a specific primary voice, and if you re strike the same key over and over, you'll get the same voices triggered each time. Other synths like the Oberheim OBX, and many others, use the traditional round-robin approach, where each consecutive key strike is assigned the next available voice in a chain, which loops through all the voices on the instrument, then back to the first voice.