The designers of Mixbus 32c’s channel had the goal of creating the original 32c console’s designer’s intentions when designing the EQ. Mixbus 32c, despite some of the marketing claims otherwise, is a Digital Emulation of the 32c channel strip.
Emulation - Emulation is attempting to simulate each component’s input and output, then assemble these similar to how the analog device is emulated. Modeling (Simulation) - Modeling is when you replicate the behaviour of the device without concern for how individual components work, only the result of sections of components, or the whole result. There are 2 ways to approach replicating an analog device in the digital world: Let’s take a short break here to clear something up. NOTE - This also extends to the high and low pass as well, however in that case there is a fairly significant loss of high end that’s only avoidable by increasing sample rate. Mixbus 32c’s phase response is quite accurate to the analog EQ with very little extra cost. That’s quite a price to pay for a very small difference, especially when many DAWs do not properly compensate for latency in certain routing setups. Pro-Q 2’s Natural Phase mode is even closer to what you’d expect from an analog EQ, however we have to pay for this small difference with nearly a 3x increase in CPU and a 384 sample (8.7ms 44.1khz) latency per instance. Mixbus 32c’s phase response on the other hand is very close to what you’d expect to come out of an analog EQ, even if the frequency response is slightly off. This shows that there is a compensatory filter in Pro-Q 2 that ‘fixes’ the cramped frequency response, but it also changes our phase response. Same frequency response (nearly), but our phase response is totally different when comparing Pro-Q 2 Zero latency and Mixbus 32c! The Pro-Q 2 ZL curve has more positive phase shift and less negative. Set in ‘Zero Latency’ mode (mininum phase) Here is the Pro-Q 2 EQ curve that produced that phase change. Around 150hz it’s only 30° etc… As we learned above, those phase shifts can be audible if we compared it to another signal with the same frequency response but different phase characteristics… So lets do that! In the chart you can see that at 20hz, there is approximately a 150° phase change. Then I measured the phase response of that EQ curve using the Minimal Phase mode, which produces phase shifting at various frequencies (group delay). I took an instance of Fabfilter Pro-Q 2 and make a simple highpass and high shelf. The X axis shows which frequency has that amount of phase change. This plot shows the phase change relative to our original signal on the Y axis from -180° to 180°. That is where the Bode Phase Plot comes in. #Harrison mixbus 32c analog how to
So in order to continue we need to have a way to discuss how to communicate the relative phase changes, at various frequencies, compared to a reference signal.
The difference is subtle, however it is there. The ‘Flip’ file has 440hz and 460hz playing at the same relative phase, but 480hz has the polarity inverted (basically similar to a 180° phase rotation). The ‘NoFlip’ file has all 3 playing at a given relative phase. In the Comparator above I have 3 sine waves at 440hz, 460hz and 480hz. However, when you have relative phase between varying frequencies, things begin to matter. Go ahead and load up a track in your DAW and flip the polarity on a sound. As these sine waves change phase, they also begin to interact with each other differently, which can have consequences to how we perceive them and their physical characteristics. It also means that certain processes can affect the phase of the signals that make up our sound. So if we can view sounds as a summation of sine waves, then we also have to realize that these sine waves may need to start at different times and start at different positions to achieve certain complex sounds.
Not all phase shift is the result of a signal being time shifted, the phase can also be rotated in place. In our image above the red signal is shifted 90° from the blue signal because the red signal starts at 90° on the blue signal. The phase shift amount is communicated in degrees (or radians) relative to the distance traveled to arrive at that phase location in the reference signal. This value is the difference between the position of the 2 sines (of the same frequency). If we have two signals that we’re comparing then we’re communicating the phase shift. When we only have one signal then we’re communicating phase as a position from 0° to some point in the wave.
0 kommentar(er)
