Softube Clipper is a dual stage clipper that consists of two stages, RMS and Peak, that both distort the audio, but in very different ways.
The idea is to use the Gain to boost the audio and gently distort in the RMS section before it hits the hard Peak clipper.
The RMS stage is a soft dynamic saturator that boosts the audio while increasing the energy (the RMS) more than it increases the peaks. It's very transparent for small amounts, and when driven hard it breaks up into a thick type of distortion.
The Peak stage is a hard clipper that shaves off the transients of the remaining audio, like a limiter with infinitely fast attack and release.
First use: With the default settings, increase the Gain until you hear audible distortion, then back off a dB or two. On percussive sounds, like a drum bus, you can push it even further. Level match the audio by pressing the Set button on the output panel.
More than twenty years ago, we started a long research project on emulating analog distortion, mainly guitar amplifier distortion, that resulted in the formation of Softube as well as a long partnership with Marshall Amplification (and a couple of patents, for example US8165309B2). We've since been geeking out on all kinds of distortion: from the exaggerated Marshall amps or Intellijel uFold eurorack modules, to more subtler mixing distortions, such as Overstayer M-A-S, a sweet SSL 4k console distortion or Softube Tape. (Or the sweet distortion from the low pass filter on an analog synth!)
But we've never made a clipper before.
Distortion can be divided into many different types and have lots of different use cases, but most of our previous distortions, such as the Saturation Knob or the thick and creamy distortion in the Summit Audio EQF-100 equalizer, have all been based on modeling of existing gear. A clipper is something different. It's a more analytical approach to distortion, and its main purpose is to control the audio, to tame the peaks, beefing up your bus, track or mix, or simply make things loud. But it has to do it without any other coloration! It should be possible to use a clipper on a master bus without accidental side effects. A guitar amp might have a really nice and smooth distortion but it also has a lot of tonal color that you most likely wouldn't place on your master. But a clipper is ok.
We wanted to design Clipper to be an analytical clipper with a vibe. The lesson from twenty years of modeling distortion circuits was that good sounding distortions are typically very dynamic, but dynamic distortion is also something that's harder to control and typically adds tonal coloring. While a well-designed pure clipper is a fun and interesting maths project it doesn't offer much "vibe", it's just very static. So the tricky part was to add the vibe, but without adding any other unwanted side effects, especially tonal coloring. The solution was to design a new type of distortion, based off the Bus Processor's "Enhanced Harmonics" mode, but with a strict mathematical approach that allows us to completely control how it interacts with the final clipping stage. Tuning that distortion, which became the RMS stage, was the final puzzle piece that gave Clipper its dynamics, while still being tonally neutral and mathematically analytical.
Clipper can be used just as other distortion or clipper products, but was designed for three main use cases:
The first is making something sound lively and punchy, much like you would use an 1176. The RMS stage adds energy while the Peak stage controls it. Over-do it, and use the Dry/Wet to dial it back.
The second use case is to increase apparent volume. The RMS stage will lower the crest factor (the peak to average level) which will make it sound louder, while the peak stage should shave off the remaining peaks so that you can increase the output level further. Knee should be in Hard mode for minimal coloration. Adjust the Headroom control to adjust which of the two stages works the hardest. Make sure that the filters in the Peak stage are turned off.
Lastly, the more creative use-case, for example when you want to change the tonality of an 808 or a drum bus. That's where you start to tweak the Headroom and Analog Color controls, and set the Low and High Frequencies to home in on a specific frequency area you want to distort.
Have fun distorting!
The user interface consists of four parts, from left to right:
- Input panel, with input metering and level. The input panel can be hidden by pressing the
icon in the menu bar. - Waveform view, showing the amount of distortion.
- Main panel with the Clipper's main controls
- Output panel, with output metering, level and auto gain. The output panel can be hidden by pressing the
icon in the menu bar.
Click here to view the full-sized image in a new tab.
From left to right: input meters and level, RMS gain reduction meter, waveform meter, Peak gain reduction meter, main controls, and lastly gain difference meter with automatic gain ("set") and output metering.
The input and output panels are part of Extended Features, a utility which is shared by all Softube plug-ins. In the settings menu you can tailor what kind of metering you want to see (for example True Peak + LUFS) and set the current layout as default. For more information, please see the Extended Features User Manual.
This section shows the gain reduction metering, with the distortion amount of the RMS stage to the left of the waveform (GR RMS), and the distortion amount of the Peak stage to the right (GR PEAK). The top half of the meters and waveform are showing the left channel, while the bottom part is showing the right channel.
Instead of measuring distortion in THD% or something similar we have chosen to show it as "gain reduction” as in a compressor instead. Everyone is familiar with gain reduction and it is a very effective way of showing amount of distortion for all saturating non-linearities.
The final waveform is shown as the blue waveform in the middle, peaks that been removed by the Peak stage is colored orange, and above that you'll see how much the RMS stage has removed.
Blue is the audio coming out from the plug-in, orange is the amount that has been removed by the Peak clipper, and the darker blue is the amount being removed by the RMS clipper.
To the right of the waveform you'll find the main controls, with Input Gain on top, followed by RMS stage controls, then Peak stage parameters and finally a Dry/Wet and Output Gain.
Input
Gain: Increases the gain of the audio, before the RMS stage.
RMS Controls
Headroom: Increases the headroom into the RMS stage, which in turn will reduce the amount of harmonics. Headroom is similar to a threshold control in a compressor, but since the RMS algorithm is a type of distortion that doesn't have a defined knee/threshold we felt it was misleading to name the control "threshold".
Analog Color: Increases the amount of even order harmonics (the sweet ones that sound good) generated by the RMS stage.
(We had lots of discussions about the name of this control, because "analog color" sounds a bit cheesy... At first it was called "asymmetry", but that didn't imply anything about how it sounded and it felt too technical. Then we tried a bunch of different names but ended up with "analog color" because that's how we experienced the sound. But it isn't based on any analog circuits, it just shares the same spice: a pinch of even order harmonics.)
Low and High Frequency: Sets the low and high crossover points for the RMS stage. The distortion will only happen between these two frequencies.
The filters are excellent to sculpt different types of distortion sounds. If you want something resembling transformer distortion, set the High Frequency as low as possible. If you want to tame the high frequencies, for example emulating the sound of a tape machine, set the Low Frequency high so it only acts on high frequencies.
RMS On: Turns the RMS stage on or off.
Filters On: Turns the filters (Low and High Frequency) on or off.
Peak Controls
Ceiling: Sets the level at which the peak clipper will clip. For a soft knee it will start to clip a couple of dBs earlier than the Ceiling level. It is also possible for audio to overshoot the ceiling when the knee isn't at its hardest setting.
Since Ceiling only limits audio between the Low and High Frequency, it will only act as a peak clipper for all audio when the Peak Filters are turned off!
Soft Knee: Sets the softness of the onset of the clipper. A hard knee is inaudible until it clips, while a soft knee will gradually distort more and more. A soft knee will also have more even harmonics than the hard knee.
"No peaks shall pass!" If you don't want any overshoots at all you need to keep the filters in the peak section turned off, and Dry/Wet set to fully wet.
Low and High Frequency: Sets the low and high crossover points for the Peak stage. The distortion will only happen between these two frequencies.
Peak On: Turns the Peak stage on or off.
Audition: Engage to listen to the audio that is being removed by the Peak stage.
If you want to distort a certain part on a track, for example a snare on a drum bus, you can use the Low and High Frequency to zone in on the snare. To better hear what you are doing, engage the Audition function to only listen to what the Peak stage is clipping.
Filters On: Turns the filters (Low and High Frequency) on or off.
Output
Dry/Wet: Increase to blend in the bypassed audio together with the distorted audio.
Output Gain: Sets the output volume.
Clipper uses multiple stages of internal oversampling, with the main core running with 4x oversampling. The RMS stage uses a proprietary method to limit the aliasing, while the Peak stage uses anti-derivative techniques, higher order harmonics filtering and oversampling. The oversampling stage uses minimum-phase elliptical filters to keep a low latency as well as the punch and attack of transients. The elliptical filters were selected to keep the phase shift in the pass band (the group delay) to a minimum, but with the cost of a larger phase shift in the stop band.
The input and output panels are part of the Extended Features, which are available in all Softube plug-ins.
For more information about Extended Features, please see the manual: https://www.softube.com/user-manuals/extended-features
Thomas Andersson – signal processing
Niklas Odelholm – signal processing, product design, user manual
Arvid Rosén – signal processing
Cameron Clark – signal processing
Björn Rödseth – UI programming, signal processing
Arvid Johnsson – UI programming
Anna Pohl Lundgren – UI programming
Tord Bärnfors – UI programming
Johan Bremin – testing