In this article and video, Paul Maunder takes McDSP’s new SA-3 spectral processor for a spin, checking out the new features and sharing his thoughts on how well it performs. Over to Paul…
One of my favourite and most commonly used McDSP plugins is the SA-2 dialogue processor. Based on hardware originally conceived by Academy Award winning re-recording mixer Mike Winkler, the SA-2 is a 5 band active equaliser/ multi-band compressor which is designed to improve the overall sound of recorded speech.
Building on the success of this Emmy Award winning processor, McDSP have taken the algorithms used in the SA-2 and advanced them several steps further in their brand new SA-3 Spectral Processor. Let’s take a look at what this plugin has to offer.
Overview
The SA-3 expands considerably on the SA-2 with over 1,000 bands of processing, a user programmable bias curve, unique Focus, Range and Density control over spectral processing, side chain support and mid/side processing.
In essence, SA-3 is designed to reduce unwanted spectral spikes in dialogue, vocals and other kinds of audio inputs. It works by analysing the input signal and reducing persistent peaks. The gain reduction, final output, or removed signal can be monitored and metered in the main frequency graph screen. Plot scales can be adjusted for gain reduction and output displays.
In Use
The Detect control will typically be your first port of call when using the plugin as this adjusts the sensitivity of the overall algorithm, a bit like a global threshold across the entire frequency spectrum. This goes hand in hand with the Range control which adjusts the maximum amount of signal peak and resonant frequency removal.
I tried the SA-3 on a number of pieces of dialogue and found that, even just by adjusting the Detect and Range controls, I was usually able to get a result which was a notable improvement on the unprocessed dialogue. The Monitor button was useful as I was able to switch between normal processing and listening to the signals being removed. To refine the processing further there’s an adjustable bias curve. This allows spectral processing to be emphasised or de-emphasised as required. I found that increasing the bias curve in areas where problematic resonances in the audio occurred, I was able to eliminate them very effectively and easily achieve a much more pleasing sound with a variety of dialogue recordings. The bias curve provides 5 adjustable frequency bands which essentially provide broad adjustment of the underlying 1,000 bands which operate in the background.
Controls are provided for curve type, frequency, gain, Q and pan. The pan control is interesting as it adjusts how much the bias curve is applied to the left or right channel or when Mid/Side mode is enabled, how much is applied to the mid or side channel. Naturally, this only applies when running the SA-3 as a stereo plugin. The Bias section provides controls for Scale and Shift. Scale does what the name suggests and increases or decreases the total amount of bias curve while shift allows you to shift the bias curve along the frequency axis.
I’m Your Density
A couple of controls in SA-3 whose function may not be immediately apparent are the Focus and Density controls. Focus provides control over the spectral processing, with higher settings only targeting more persistent resonant frequencies. Density is a control which requires some experimentation to understand. For each fundamental frequency spike or resonance, harmonics above that frequency are further suppressed when the Density control is dialled in a bit, creating a more ‘dense’ spectral suppression. This is the sort of thing which is hard to fully describe but after using SA-3 on a number of pieces of dialogue, I was able to use the Density, along with the other controls, including attack and release, to achieve superb sounding results which gave a more polished sound to not only dialogue but also room tone. I used SA-3 on a scene where a particular low-mid frequency resonance in the background ambience was drawing too much attention. SA-3 allowed me to fix this quickly and get on with the rest of the mix.
Side Chaining
At the lower right hand side of the interface is the Side Chain section. As you might expect, when used with the key input, this allows for ducking. It’s a great way to manage tracks that are competing against one another. In my work, I found this to be useful as a way of carving frequency content out of background music whenever dialog was present. A Trim control adjusts the side chain signal level and there’s a Listen button to monitor the side chain. Side chaining with the SA-3 has a distinct advantage over a basic compressor in that it will specifically target the most prominent frequency content from the side chain signal, allowing the affected sound to be attenuated at those specific frequencies, thereby allowing two sounds to co-exist better in a mix.
Verdict
I spent several days experimenting with the SA-3 Spectral Processor and using it on a number of post-production mixes. I’m impressed with the functionality this plugin provides and it’s a highly effective tool for achieving polished sounding dialogue in film and TV mixes. In practice, the usefulness of SA-3 goes beyond just dialogue and extends to vocal and instrument processing too. The included presets allude to this functionality and often provide a great starting point for settings.
If you like McDSP’s SA-2 then you’ll find the SA-3 a great evolution of this. It retains the character we’ve come to love from the SA-2 algorithms but builds on it with an even more useful set of controls which make the control and suppression of unwanted resonances straight forward. I’ve added SA-3 to my dialogue processing chain and I expect that I’ll be using it on most mixes going forward.
The SA-3 is the kind of plugin which you really need to try for yourself in order to fully understand the benefits. McDSP provide 14 day demos of all of their plugins, including SA-3. For a quick overview of SA-3, check out my video below.