Universal Audio WebZine
Volume 4, Number 4, May 2006
[UA Universe] [Ask the Doctors] [Artist Interview] [Analog Obsession]
[Support Report] [The Channel] [Plug-In Power] [Playback] [Featured Promotion]
[Graphic-Rich WebZine]
[Back Issues] [UA Home]


Dr. David Berners (left) is the Universal Audio Director of Algorithm Development; Dr. Jonathan Abel is the co-founder and CTO
Ask the Doctors
By Dr. Dave Berners

Q: How is the Neve 1073 different from the Pultec or from other EQs UA has modeled?
A: The shelving and peaking filters in the 1073 are active EQ circuits, built around class A discrete amplifiers designed by Neve. Making these stages active provides a nice way to make well-behaved, symmetric responses with very little interaction between the EQ bands, and to produce filter responses with plenty of boost (close to 20 dB). The high- and low-shelving filters are incorporated in feedback around a single amplification stage. Since they operate largely in different frequency ranges, there is a good deal of immunity from interaction between these two bands, and there is a beautiful symmetry both to the topology of the circuit, and to the filter responses. The midband peaking filters are also built as feedback around an active stage, and again have a symmetric response. Although it would be possible to build an active peaking filter without magnetics, the 1073 employs an inductor and capacitor to generate the resonance for these filters. The responses for the peaking filters have a nice dependence on the bandwidth as the gain is adjusted, and at higher center frequencies, the Q goes up, for a more focused peak.

The New Neve 1073 EQ Powered Plug-In

The highpass filter on the 1073 is a passive filter that also uses magnetics to generate a third-order low-cut response. The highpass is tuned to have a square roll-off without having excessive resonance. This filter is sensitive to the loading on its output, so it is important that the 1073 be used with a fader of the appropriate impedance to prevent warping of the highpass response curves, since the highpass filter looks directly into the fader.

Racked Analog Neve 1073s
In terms of modeling, the 1073 has several features that make it interesting. Most of the subtlety probably comes from the magnetics, whose self-resonances change the filter responses measurably from what might be calculated from the schematic alone. Parasitic effects in the low-cut filter from the inductor slightly reduce its resonances, and in the midband peaking filters, the responses are drawn towards unity gain at high frequencies by parasitic capacitances in the inductor. On the shelving filters, the low shelves have varying degrees of resonance as different capacitors are switched in for various shelving frequencies. The high shelf has a very gradual slope for both boost and cut, because its poles and zeros are interpolated. This gives it a very warm sound, and it is never strident, even at maximum boost. Both the high and low shelves feature a small "divot," a region of inverse gain slightly before the shelving frequency. This is similar to the feature of the Pultec EQP1-A when low boost and cut are employed at the same time. As far as the overall response of the 1073, the high and low ends are mostly flat, with a tiny bit of roll-off on the high end. On both the high and low end, there is some phase shift associated with the finite bandwidth of the unit. The discrete amplifiers in the signal path are highly linear, with good frequency response, and give a tight, neutral sound when discounting the effects of the EQ and transformers.

Do you have a question for the Doctors?