Review: Zynaptiq Unchirp

Audio Restoration Plug-In
Publish date:
Social count:
Audio Restoration Plug-In
Image placeholder title

The German software maker Zynaptiq, creator of plug-ins such as the highly acclaimed Pitchmap and the reverb-remover Unveil, has a well-deserved reputation for innovation. Its latest offering, Unchirp, continues this tradition (see Figure 1) by providing a powerful and complex plug-in whose primary purpose is to restore audio that’s been degraded by FFT (Fast Fourier Transform)-based processes such as lossy codec encoding or spectral denoising. But its prodigious abilities also come in handy in both the mixing and mastering environments.


Unchrirp’s main screen is divided into three sections. The largest is a graph-like display that functions both as a real-time analyzer and for controlling the Bias feature, which I’ll explain a bit later. Below it are five large 3-D circular parameter knobs (graphically the same as the ones found in Pitchmap), each with accompanying sliders. Across the bottom of the screen, five buttons and two sliders comprise the Global Parameters.

To understand Unchirp, start by looking at the five sections of the middle portion of the screen, each highlighted by one of the five large color-coded knobs. With the exception of the Gain section, each section controls one of the processes available in Unchirp.

On the far left are the controls for the Musical Noise reduction section. Musical noise is a type of artifact that’s often generated when encoding audio with a lossy codec, especially at low kbps rates. It’s also generated when using a spectral denoiser with the settings too high; it can give audio a swooshy, underwaterlike sound.

The term musical noise is often used more broadly to cover all such artifacts, but Zynaptiq defines it in Unchirp as artifacts that are completely extraneous from, and not correlated to, anything in the recorded sound. Correlated artifacts are defined by Zynaptiq as “chirping,” which can be reduced by the Dechirping section.

Both the Dechirping Amount and Musical Noise Reduction knobs are next to sliders (labeled Window) that adjust the maximum length of the artifacts to be removed. The shorter the maximum length, the less the desired signal is affected by the processing. You can use I/O Diff, located in the Global Parameters section, to hear only the part of the audio being processed: This can help you set a window length that is long enough for the processing to be effective, but short enough not to adversely affect your recorded material.


In addition to creating the above-mentioned artifacts, lossy codecs and spectral denoisers can also smear transients. Unchirp has a section dedicated to restoring transients that includes the large Threshold knob and the Orig and Synth sliders. The process allows you to create synthetic transients to replace smeared ones, or boost or cut the original transient’s level, among other possibilities. This section is also very useful for creative processing of drum and percussion parts.

The Solo switch allows you to hear just the transients in the signal (including the original and any synthetic ones you might have), which can be useful when setting it.

The Treble Boost process lets you replace highfrequency information lost during encoding and de-noising. The Mix section has a Wet/Dry control, which is where you adjust how much of Unchirp’s processed signal gets routed to the plug-in’s output.


Three of the four processes—Musical Noise, Dechirping, and the Transients—can be further fine-tuned with the Bias feature, which gives you frequency-dependent adjustment of the processes and parameters, adjusted through creating breakpoint curves in the graph.

Each controllable parameter has a check box under the graphical editor, matching the color coding for that process; when you click on the name beside the checkbox, a line appears in the middle of the graph. Clicking anywhere on any frequency in the line adds a breakpoint (you can add multiple ones), and by moving the breakpoints, you can create a curve.

Five different curve types are provided. If the curve is above the center line, you’re adding to the process in the frequency range defined by the curve; if it’s below the line, you’re subtracting from it. In essence, the Bias feature, used in conjunction with the corresponding process control, lets you apply processing only in a particular frequency range.

If this sounds confusing, frankly, it is. And it’s not easy to get your head around the concept. The manual is fine for explaining what the knobs do, but considering how complex Unchirp is, it could really use some online tutorial material. As of this writing, the video on Zynaptiq’s site only shows what the plug-in can do, not how to use it.


In addition to the previously mentioned I/O Diff button, other Global Parameters include a Side button that switches the processing to only the sides of an M/S configuration; a Limiter button that activates a brickwall limiter set to 0 dB to prevent overs; and a Bypass button. The Sync function, controlled by a slider, is designed to add clarity by synchronizing the processing to the transients in the target audio. The Warmth slider reduces the buzz and metallic ringing that can be made audible by the Unchirp process.

The final Global parameter is HD Mode, which adds additional quality to the processing but increases the already hefty CPU load of Unchirp— one of the main drawbacks of this plug-in.


Unchirp has a wide variety of applications, and it includes a large set of presets that are organized into categories including General Purpose and Sound Design, Mixed Music, Mixing Tools, Post, and Psycho-Acoustic Processing—perfect for getting you started. I tried Unchirp on a variety of sources and for a number of different applications.

For restoring the sound of a file compressed with a lossy codec, I encoded a song as an MP3 at a paltry 80 kbps, and the resulting audio was rife with artifacts. Using one of the Mixed Music presets, I was able to significantly improve the sound.

Unchirp is also touted as being able to reduce artifacts in audio files that were over-processed by a spectral denoiser. I tried it out with both spokenword and music files that had denoising artifacts. When I used settings from the appropriate presets and tried out my own settings, the plug-in seemed to have very little effect. There was no dramatic improvement as there was with the MP3 file. According to Zynaptiq, there are two limiting factors. First, the longer the FFT, the longer the artifacts; at some point, they become greater than the window of time available in Unchirp's real-time processing. Secondly, some denoisers have their own musical noise suppression logic, which attempts to mask the noise in a way that makes it difficult for Zynaptiq to reduce.


Fig. 2. A psychoacoustic preset that emphasizes Bark frequencies. This setting can add punch to a weak-sounding snare drum. I was intrigued by the possible uses of Unchirp for mixing, and found that its Transient section worked great for making drums and percussion more punchy. It also added sparkle and life to an acoustic guitar track, using, believe it or not, the Unchirp and Musical Noise processes, along with the Treble Boost. In fact, it was great for making a variety of tracks sound better. The problem is that it’s so CPUintensive that I couldn’t use it on too many tracks simultaneously in a mix. The workaround is to process tracks and bounce them in place, or bounce and re-import them, depending on your DAW.

Image placeholder title

I could also see Unchirp being a useful tool for mastering. Its preset section features psychoacoustic settings that emphasize the 25 “Bark” frequencies (see Figure 2), at which the human ear is most sensitive. Energy in those bands, particularly transient energy, makes audio sound louder and crisper. I tried some of the presets on recorded music and found that in many cases, it made them sound brighter and more punchy.


Although Unchirp is CPU intensive and complex, there’s a lot to like about it. Its collection of presets definitely helps demonstrate settings for particular applications, or at least gets you in the ballpark; I would urge Zynaptiq to post more tutorial information and videos on its site. I also hope they can address in an update the artifacts created by current-generation spectral denoisers.

Image placeholder title

That said, Unchirp is a very powerful plug-in, and there is nothing quite like it on the market. Its ability to salvage bad-sounding, lossy encoded music is impressive, as is its ability to adjust transients and add punch and definition to both individual and mixed tracks.

If your production work includes lossy encoded music that needs improvement, Unchirp will be a no-brainer for you—especially if you’re a techie or someone who is not afraid to roll up your sleeves and take on a challenge. This plug-in gives you processing capabilities unavailable elsewhere.

Quick Tip: Resuscitating a Wimpy-Sounding Snare Drum

A great Unchirp function is its ability to add punch to a weak-sounding snare drum. Here’s one way to make it more snappy; this setting will work best on snare recordings that are round-sounding, and don’t have enough transient energy.

1. Set Unchirp to its default setting, then turn the Musical Noise and Unchirp sliders all the way down.
2. Set the Transients threshold a little bit lower than the center position, which will allow more synthetic transients to be created.
3. Set the synth slider a little below halfway.
4. Set the Orig slider a little over the center position to slightly boost original transients.
5. Compare the sound by using the Bypass button.
6. Try moving the Sync slider left and right. Moving to the right will make it sound tighter. Pushing the Orig slider up will boost the original transients.

Eliminates artifacts. Restores transients. Bias feature.

Complex. Manual and Website lack tutorial material. CPU-intensive.


Mike Levine is a musician, producer, and music journalist based in the New York City area.