Why Innovate When You Can Emulate? Exploring Contemporary Plugin
Development And Potential Innovation
ABSTRACT: As software driven approaches to mixing audio have become increasingly prevalent there has been a continuous improvement in quality in these digital mixing tools. Mix workflow in the box has developed through the established workflow presented by a traditional mixer, with the DAW offering a combination of the tape machine and the routing and processing matrix to the end user. One of the most common categories of plugin is represented by emulation, with classic hardware modeled often to component level by developers looking to create devices as measurably close as possible to a hardware reference.
The second category of contemporary plug-in development is based around tools, which rather than focusing on copying an established tool instead are targeting specific mix elements or sonic signatures. Some companies such as Waves have built on the reputations of mix engineers, offering users an insight into their sounds through named plugins. The CLA Signature Series and the Tony Maserati Signature Series are two such collections based around element specific processors. The CLA collection offers processing around commonly understood mix processes, with the Maserati drums instead featuring more descriptive parameters such as thump and snap.
The third category of plugin development incorporates plugins which rather than basing their approach on existing hardware tools or established mix practice are looking to incorporate new approaches to mix processing.
In this paper I am looking to analyse current developments, looking at these categories and how innovation can be incorporated into new mix tools. Emulation of tools leads to a mix economy based around knowledge of the tools and the re-creation of existing mix practice. New tools offer a unique opportunity, with direct access to the characteristics of sound. Through analyzing existing tools it is possible to understand the important characteristics, incorporating these into new contemporary mix processors.
Today’s presentation was inspired, Andrew says, by a conversation he had with the head of an [unnamed] Pro Audio company, who stated that he was looking to develop more mass-market hardware products with fewer features in a ‘race to the bottom’. Andrew’s disappointment with this conversation has led him to investigate categories of plugins.
We first see a discussion of emulated plugins and specifically Universal Audio who developed a hardware-based business model (1176 and LA2A) into a primarily plug-in emulation based business model. Andrew addresses the challenges of emulating electronics in code, using the challenges of diode clipping emulation as an example. This leads us to the concept of component modelling, which has become a popular method of coding plugins.
We are now seeing ‘place’ (e.g. not just a Fairchild, but Ocean Way Studios’ Fairchild) as a marketing concept.
Warm Audio, IGS and others are mass-producing DIY designs, providing hardware products that are based on software emulations (and of course, undercutting the cost of the original hardware). Andrew considers this a non-creative approach that is doing little to drive the audio forward.
A further extension of the ‘brand’ idea is the celebrity mix engineer (last year’s ARP guest speaker Tony Maserati is cited as one example among several). We look next at mixer brands (e.g. CLA-2A, CLA-3A and CLA-76). Although Andrew considers these poor emulations in terms of accuracy he notes that they have a distinctive (and implicitly artistically desirable) sound.
Manny Marroquin EQ plugins are cited as a good example of taking a ‘greatest hits’ approach to plugin design, whereby different EQ bands are taken from different products (!). So we’re now starting to see development of the interface itself. Parallel Particles’ ‘Sub, bite, thick and air’ parameters show the development of semantic descriptors; Andrew’s research is questioning why this user-interface simplicity is derided by some in pro audio.
Taking semantics further, the Greg Wells Voicecentric plugin has one control called ‘intensity’ – but at its heart it’s just an 1176. He’s arguing that this simplicity is a good thing – why would you mess with emulated 1176 knobs when you could semantically jump right to the desired sonic outcome?
We look briefly at non-emulation processing tools, such as Fabfilter Pro-C 2, and configurable tools such as compassion, which enables the user to design the way the compressor works – a techie’s dream, as he describes it.
Andrew suggests that the problem with emulated tools is that they behave too authentically (I infer, through accurate component modelling). And today, gain staging has improved, so why would emulated plugins take a ‘warts-and-all’ approach. He is arguing for emulating the sound by all means but modernising the interface (e.g. through semantics) and eliminating some undesirables (e.g. limitations of noise floor or gain stage headroom). The tools we are emulating (originally) expected a workflow with specific gain-staging, so emulating the tool on its own is not enough to give the sound we may wish to emulate. Education in gain-staging (that is, how and where to place the plugin in the signal chain and workflow) is essential for users – the plugin alone will not do this.
Does a new language need a new control paradigm? Emulation should be about the sound of the gear.