Extract:Dialogue by Acton Digital is a plugin designed to isolate dialogue from prevalent background noises like wind, rustling, traffic, humming, clicks, and pops. This is specifically useful when you record location sound, podcasts, vlogs, songs, etc., and want to reduce the background noise, if not completely remove it.
I applied the plugin to different styles of recordings, ranging from home-recorded microphone recordings to outdoor recordings done on my phone microphone and a boom mic, and the results were quite impressive. I like how the plugin lets you solo the noise and fine-tune it by using its 3-band equalizer.
It is clean, transparent, user-friendly, and with the latest updates, has much-reduced artifacts. For indoor recordings, it worked wonderfully in reducing the air-conditioner, laptop fan, and room sound. The problem with most noise reduction plugins is that they aren’t very good with the gushing wind, low-frequency rumbles, and air sounds, as microphones are very sensitive to them.
Most other noise reductors end up over-cleaning the sounds at the low end and make the output hollow. However, this plugin is good at handling outdoor traffic noise and wind sounds and maintains the natural tone of recordings with its phase correction algorithm.
The plugin packs a spectrum analyzer that displays the frequency spectrum of your sound and lets you visualize your audio in real time. While it can automatically detect the noise, you still get customization controls. The two main controls are the ‘Sensitivity (%)’ slider and the ‘Maximum attenuation’ knob.
The sensitivity slider fine-tunes the relative sensitivity of noise detection. Positive values heighten sensitivity, enhancing noise removal, while negative values reduce sensitivity, allowing more noise to persist.
In various scenarios, retaining a certain level of noise is preferable. You can regulate this by the maximum attenuation knob. This parameter, set in decibels (dB), dictates the maximum attenuation of frequency components, thereby governing the noise floor after processing.
Moving on, you also have EQ options with three bands in the spectrum analyzer called the Sensitivity curve/filters. The sensitivity filter provides control over noise sensitivity in customizable frequency bands. Three filters are available, toggled using buttons in the spectrum analyzer’s upper right corner: a low shelf filter ( ), a peak filter ( ), and a high shelf filter ( ).
These functions function are like parametric equalizers. You can customize filter attributes by clicking and shifting the colored bullets (handles) on the curve, revealing the current frequency and gain settings. The plugin lets you fine-tune filter slopes or adjust peak filter bandwidth by interacting with the chosen filter handle.
Then there’s the preset selector, where you can select presets, search for them, and also import/export them or save your own presets.
Right next to the solo button, you can click-drag the plugin from the right bottom to resize it. Finally, there’s the solo button that lets you listen to the noise in isolation.
Best use cases
Extract:Dialogue delivers satisfactory results in various audio scenarios. Ideal for podcasts and dubbing, it ensures a professional and crisp outcome by effectively eliminating background noises like computer fans, cable hiss, microphone pops, headphones bleed, and other room sounds.
In studio recordings and editing as well, the plugin stands out for its ability to maintain fair audio quality, removing ambient disturbances like air-conditioner hums. It’s also good for cleaning vocal extraction software output as it is good with crackle sounds as well.
It performs decently in extreme circumstances as well. Unlike alternative applications, it effectively addresses noises that have proven difficult to eliminate. Its remarkable performance is particularly evident in handling clothing rustle on lavalier microphones, overcoming a significant challenge faced by sound editors.
For location sound, including dialogues and interviews, Extract:Dialogue proves good in isolating spoken words from environmental noises like wind and traffic. It is equally adept at cleaning phone recordings and addressing challenges associated with mobile audio capture.
Under The Hood: Algorithm Explained
Operating in real-time, the algorithm is rooted in deep learning principles. Trained on a vast dataset comprising numerous high-quality voice recordings and an expansive array of typical noise sources, this technology demonstrates advanced capabilities.
The extensive training lets the artificial intelligence autonomously differentiate between dialogue and noise, eliminating the need for any intervention. The user-friendly nature of Extract:Dialogue is emphasized, as it seamlessly integrates into your dialogue bus, requiring minimal interaction while delivering results.
Version 1.5 further enhances efficiency by demanding less CPU usage and introduces an innovative artificial intelligence (check AI plugins) model designed to rectify phase distortions and destructive interference, which is particularly beneficial when addressing frequency spectrum overlaps between dialogue and noise signals.
Pros & Cons
One of Extract:Dialogue’s significant advantages is its ability to effectively reduce or eliminate background noises like computer fans, cable hiss, microphone pops, and ambient room sounds. The real-time processing, rooted in deep learning principles, allows the plugin to autonomously differentiate between dialogue and noise autonomously, delivering remarkable results with minimal user intervention.
The plugin’s user-friendly interface provides customization controls, including sensitivity sliders, maximum attenuation knobs, and EQ options with three bands. The spectrum analyzer, preset selector, and solo button enhance the user experience, offering visualizations and precise adjustments.
The artificial intelligence model, introduced in version 1.5, addresses phase distortions and destructive interference, ensuring efficient handling of frequency spectrum overlaps between dialogue and noise signals.
Its efficient performance, especially with the 1.5 version updates reducing CPU usage, makes it suitable for a variety of recording environments, including podcasts, studio recordings, location sound, and mobile audio capture.
Despite its strengths, the plugin comes with a price tag of $99, which may be considered a drawback for some of you. However, the cost is significantly lower than some high-end competitors like iZotope RX. Additionally, while the plugin excels in most scenarios, it may not completely replace more expensive alternatives like iZotope RX, especially in professional studio settings.
Next, while the plugin offers presets for convenience, some of you may find the lack of detailed customization options in the presets limiting. Further, in extreme cases or with poorly recorded source material, there might be instances where the plugin introduces some artifacts or compromises the natural quality of the dialogue.
Compatibility & Installation
Extract:Dialogue by Acon Digital is available for Windows 10/8/7 and MacOS X 10.8 or later operating systems and is compatible with software that can host AAX, VST, VST3, and AU plugin formats. It does not load your CPU much, especially with the latest 1.5 version updates, and can even work with just 1 GB of RAM.
Regarding installation, the plugin comes with an installer that will install your selected plugin format.
As a final word, I would say that Acon Digital has developed a pretty solid plugin that stands out and delivers acceptable results in most de-noise use cases. It is also feather-light on your CPU, unlike other more complicated and heavier noise-reduction plugins like Izotope RX and Waves Clarity VX. Thank you for reading. Hope this article provided you the right value.
Shaurya Bhatia, is an Indian Music Producer, Composer, Rapper & Performer, who goes by the stage name MC SNUB, and is also 1/2 of the Indian pop music duo, called “babyface”. A certified Audio Engineer & Music Producer, and a practicing musician & rapper for more than 6 years, Shaurya has worked on projects of various genres and has also been a teaching faculty at Spin Gurus DJ Academy.