Fig. 1
In your guitar pedal dealings, you may have heard the phrase “component tolerances.” Nearly every component in a pedal is marked with a value, and ideally every component in your pedal would be that exact value, not one bit more and not one bit less. So, every 1k-ohm resistor would be exactly 1.0000000000000k ohm and every 10 µF capacitor would be exactly 10.0000000000000 µF. In this supernatural circuit situation, every pedal would sound identical. There would be no deviations from each component’s intended value, and there would be no deviations from the effect’s intended sound (all other things being equal). Unfortunately, we cannot hope to achieve this sort of metric perfection in the real world. While perfection may not ever be attained, it is also not often required, and all the circuits we interact with day in and day out can tolerate some sort of variation in their components’ value.
Your pedal’s designer specifies every component value in a design to result in a particular timbre or function and needs to know how much each component employed might vary from that specified value. When manufacturers make parts, they specify the nominal value and a particular tolerance value. Often, this tolerance is specified as a percentage of the nominal value. So, a 10-percent tolerance in a 1k-ohm resistor could be anywhere from 900 ohms and 1100 ohms. The higher the tolerance figure, the more variation you can expect in the value of a given component.
Fig. 2
While more variation may seem like strictly bad news at first blush, it does have some benefits—principally, cost. The machines and processes required to make a one-percent resistor are cheaper and faster than those required to make a resistor of arbitrarily higher precision. Consequently, a run-of-the-mill one-percent tolerance resistor may cost five cents while its 0.005-percent laser-trimmed counterpart costs $12. If vintage pedal prices are starting to make you queasy, know that it could be much worse. Demand plays a major role in cost as well, as it is pretty rare that you need a 0.005-percent resistor.
So, what difference does tolerance make and how do we know when we need to splurge for the caviar components? Without getting too far into the mathematical weeds, here’s a couple examples. Take Fig. 1, where a very simple resistor/capacitor low-pass filter is shown. This filter’s corner frequency [Fig. 2] is determined by the value of the resistor and capacitor, and that frequency has a certain sensitivity to variations of those values. [Note: The corner frequency is also known as the cut-off frequency—frequencies above this point will be attenuated by the low-pass filter.] In this particular circuit, if we wiggle the capacitance value by 10 percent, the corner frequency will move by approximately 10 percent.
Fig. 3
The corner frequency of the inductor/capacitor low-pass filter in Fig. 3 has a different sensitivity to changes in the value of the capacitor. If we increase the value of the capacitor by 10 percent, the corner frequency of the filter moves by approximately five percent. So, we can say that the corner frequency of the Fig. 3 circuit is less sensitive to changes in the capacitance than the Fig. 1 circuit. If you want to build circuits that are more forgiving of changes in component values, you can make some design decisions that will help!
You can also quantify what difference component variation will make in the context of your particular application. Let’s assume we’re employing the circuit in Fig. 1 as a pedal power supply filter. Let’s set resistance at 470 ohms and capacitance at 220 µF. We know we’re primarily wanting to filter 60 Hz hum from our power line, and at this nominal value of R (resistance) and C (capacitance), we can attenuate 60 Hz by approximately 32 dB. If we choose a 20-percent capacitor, in the worst case, our C drops 20 percent to 176 µF and we only reduce that 60 Hz noise by 30 dB. In practice, that difference of 2 dB probably won’t result in a dramatic difference in performance. This tolerance to higher tolerance parts is one of the reasons why we see 20-percent capacitors in big power supplies. When bigger is better, some amount of overkill can make variations in value a non-issue, practically.
Whether you’re bending your own circuits or just trying to figure out why you can’t find a backup that’s quite as good as your No. 1 dirtbox, you might consider how the imperfections in those little devices inside our devices add up to make something special.
Guitarists are used to coveting cranky, old gear, but if you’re not a rack user, the H3000 might not be for you! If you ever see one of these in a studio, be sure to take some time to check it out, and maybe give the plugin a try.When it debuted in 1987, the H3000 was marketed as an “intelligent pitch-changer” that could generate stereo harmonies in a user-specified key. This was heady stuff in the ’80s! But while diatonic harmonizing grabbed the headlines, subtler uses of this pitch-shifter cemented its legacy. Patch 231 MICROPITCHSHIFT, for example, is a big reason the H3000 persists in racks everywhere. It’s essentially a pair of very short, single-repeat delays: The left side is pitched slightly up while the right side is pitched slightly down (default is ±9 cents). The resulting tripling/thickening effect has long been a mix-engineer staple for pop vocals, and it’s also my first call when I want a stereo chorus for guitar.
The second-gen H3000S, introduced the following year, cemented the device’s guitar bona fides. Early-adopter Steve Vai was such a proponent of the first edition that Eventide asked him to contribute 48 signature sounds for the new model (patches 700-747). Still-later revisions like the H3000B and H3000D/SE added even more functionality, but these days it’s not too important which model you have. Comprehensive EPROM chips containing every patch from all generations of H3000 (plus the later H3500) are readily available for a modest cost, and are a fairly straightforward install.
In addition to pitch-shifting, there are excellent modulation effects and reverbs (like patch 211 CANYON), plus presets inspired by other classic Eventide boxes, like the patch 513 INSTANT PHASER. A comprehensive accounting of the H3000’s capabilities would be tedious, but suffice to say that even the stock presets get deliciously far afield. There are pitch-shifting reverbs that sound like fever-dream ancestors of Strymon’s “shimmer” effect. There are backwards-guitar simulators, multiple extraterrestrial voices, peculiar foreshadows of the EarthQuaker Devices Arpanoid and Rainbow Machine (check out patch 208 BIZARRMONIZER), and even button-triggered Foley effects that require no input signal (including a siren, helicopter, tank, submarine, ocean waves, thunder, and wind). If you’re ever without your deck of Oblique Strategies cards, the H3000’s singular knob makes a pretty good substitute. (Spin the big wheel and find out what you’ve won!)
“If you’re ever without your deck of Oblique Strategies cards, the H3000’s singular knob makes a pretty good substitute.”
But there’s another, more pedestrian reason I tend to reach for the H3000 and its rackmount relatives in the studio: I like to do certain types of processing after the mic. It’s easy to overlook, but guitar speakers are signal processors in their own right. They roll off high and low end, they distort when pushed, and the cabinets in which they’re mounted introduce resonances. While this type of de facto processing often flatters the guitar itself, it isn’t always advantageous for effects.
Effects loops allow time-based effects to be placed after preamp distortion, but I like to go one further. By miking the amp first and then sending signal to effects in parallel, I can get full bandwidth from the airy reverbs and radical pitched-up effects the H3000 can offer—and I can get it in stereo, printed to its own track, allowing the wet/dry balance to be revisited later, if needed. If a sound needs to be reproduced live, that’s a problem for later. (Something evocative enough can usually be extracted from a pedal-form descendant like the Eventide H90.)
Like most vintage gear, the H3000 has some endearing quirks. Even as it knowingly preserves glitches from earlier Eventide harmonizers (patch 217 DUAL H910s), it betrays its age with a few idiosyncrasies of its own. Extreme pitch-shifting exhibits a lot of aliasing (think: bit-crusher sounds), and the analog Murata filter modules impart a hint of warmth that many plug-in versions don’t quite capture. (They also have a habit of leaking black goo all over the motherboard!) It’s all part of the charm of the unit, beloved by its adherents. (Well, maybe not the leaking goo!)
In 2025, many guitarists won’t be eager to care for what is essentially an expensive, cranky, decades-old computer. Even the excitement of occasional tantalum capacitor explosions is unlikely to win them over! Fortunately, some great software emulations exist—Eventide’s own plugin even models the behavior of the Murata filters. But hardware offers the full hands-on experience, so next time you spot an old H3000 in a rack somewhere—and you’ve got the time—fire it up, wait for the distinctive “click” of its relays, spin the knob, and start digging.
