Menu collapse

Reviews for Invisible Limiter

Matthew Zouhar Lewis — Sound Designer

Japanese manufacture A.O.M. have established their name within the audio industry for their range of top quality M/S tools, so when I came to learn they had developed a brick-wall limiter, they certainly got my attention. Entitled 'The Invisible Limiter', A.O.M. have come at this challenge with a perspective of transparency and simplicity. The GUI design is stripped back, no frills, equipped with 10 controls used to sculpt the dynamics of source material. Lets be honest, there is no shortage of dynamic processors available on the market nowadays, so to stand out you need to come with something special and I believe A.O.M. have done just that.

For those who like to be greeted with response controls such as attack and release may be initially put off to find the developers have opted for an automatic design. In my initial testing, for the variety of source material I threw at this plugin, its adaptive algorithm has been very well tailored to provide optimum results in the majority of circumstances.

The plugin loads its default state in L/R mode whereby the Left and Right hand channels are processed separately, this can be adjusted to M/S mode whereby the Mid and Stereo signals can be independently processed. The audible result of this is that the Mono content is limited harder than the stereo image so for the majority of music, the weight of the track in mono receives more limitation while subsequently stereo content has more room to breath, the end result is a slightly wider sound with more 'air' and less high frequency attenuation.

You have two shape modes for the limiter, logarithmic and linear, as the name suggests, the logarithmic curve is slightly more rounded and less aggressive whereas the linear curve is a lot more precise. I found the logarithmic curves better suited for acoustic material at slower tempos and the linear curve better for faster tempo electronic music. One thing I did notice was slightly more transient distortion in the linear mode, presumably this is caused by the nadir in the reduction curve and how the signal stops its downward motion and returns back to increasing amounts.

The oversampling feature allows you to change anywhere from 1x to 16x oversampling, higher values result in cleaner sound quality but substantially more CPU with each incremental jump. To give you an idea of this, I am currently testing on a 2012 Mac Pro, 12 core 64GB ram machine. On 1x my cpu was riding at around 4%, when I increased the oversampling to 16x my CPU usage meter went up to 76%.

There are 2 different latency modes to choose from Normal and Low. In normal mode my system was reporting latency of 52ms which in a mastering environment really isn't a problem but in a mixing environment this is just a bit too much in reality. So at a slight compromise in audio quality you can choose a low latency mode which reduced its computational processing time down to 7ms on my system.

There are three different overshoot modes, these change the way the limiter deals with digital peaks that slip through the net, you have a choice of suppression which removes overshoots by adding small amounts of gain reduction to the output ceiling, clipping which literally clips overshoots and a thru option which just allows them to pass through. As for which one is the best to use generally, I would say its contextually dependent. If you are not riding close to digital 0dB then using the thru option should be fine also this settings works if you were to be chaining limiters together. The Clipping option is fine as long as you are not hitting the limiter particularly hard otherwise the audible result can be slightly unpleasing. As for the suppression mode, this is kind of the best of both worlds as long as you don't have a very strict volume standard that you are trying to comply to.

In summary:
I am always on the hunt for dynamic processors that take audio quality to the next level, in my pursuit for finding the ultimate brick-wall limiter with the lowest distortion characteristic I have tried a number of pretty impressive pieces of hardware and software. Many plugins available today combine multiple different DSP processing tasks into one unit, where as others concentrate on simply doing one thing and doing it really well. There is a overriding sense when you use 'the invisible limiter', that A.O.M. have designed their product with a very clear intention; to be the cleanest dynamic limiter with the lowest point of distortion. If your goal is to slam source material to make it super loud or simply cap any dynamic peaks to avoid digital clipping, the invisible limiter is as the name suggests, 'transparent'. This truly is one of the most impressive pieces of dynamics processing technology I have ever used.