Model Behaviour: LA-2A Emulations vs Hardware

The questions

So the questions before us now are these:

1. Does the unique LA-2A magic have a place in the production of modern dance music?

2. If so, do the new plugins from Universal Audio offer any of that magic to those who don’t have the £3,000 for a single channel of the analogue box?

The short answers, for those interested in putting this article down and spending their money as quickly as possible, are ‘yes’ and ‘yes’.

The long answers are more interesting, though, especially for those of us who actually have a budget with limited funds that mandates we choose our tools more carefully. Because, as with all uniquely magical audio processors, the LA-2A does not work for all things all of the time, and in fact it is much more limited in use for electronic music than, say, an SSL or other VCA-style compressor. Whether you will find its mojo useful depends on the types of sonic textures you gravitate towards and, most importantly, the kinds of sounds that define your productions.

The good news is I’ve done enough legwork to give you a good sense of whether you need to check this tool out, so read on, because if you haven’t noticed, I like to talk about this stuff… a lot.

One Leveller, Three Models

Modelling is a relatively new technology made possible by the sheer power of today’s computers (and I say this knowing that in ten years this statement will be laughable). The idea is that you use test equipment to analyse all the things an analogue processor does to the sound, and you create algorithms which transform sounds in a way that is as close as possible to the hardware.

The three new UAD LA-2A plugins

If you’ve spent two minutes on Gearslutz you know that there is an ongoing and always-heated debate about how close plugins come to the hardware they model. Nearly everyone admits they’re in the ballpark; most would also agree that there are sonic differences. Some say those differences matter, some say they don’t.

What is not up for debate is that more and more artists, engineers and producers – even the ones who think plugins come up short – rely on them more and more with every passing year. They offer immense bang for the buck and a workflow that makes high-quality music production a possibility for tens of millions of people worldwide who otherwise wouldn’t be able to afford a basic recording rig. For a grand or three you can get one or two channels of coveted analogue processing; for anywhere from £0-299 you can get a plugin that can be used on as many tracks as your computer can handle, with total and instant recall of every instance in every session. Whatever else is true, plugins represent extraordinary value for money.

So if the idea behind modelling is to recreate the effects of a hardware processor, why did Universal Audio create three new models of the LA-2A? If you know anything of the history of this box, you know that it has been through several design revisions (i.e. parts and spec changes) and been owned and manufactured by several companies. In addition, because most LA-2As out there are roughly 20-40 years old, a great many have been modified, upgraded, and changed by technicians along the way. And old parts have a way of aging and degrading in ways that make them sound and react differently from when they were new.

All of which is to say that, while the new reissue hardware LA-2As are very standardised from one to the next, most vintage units sound, compress, and distort in subtly unique and different ways. UA chose three units that they felt best exemplified the different sonic and musical range of the circuit, and they modelled them up. In use, there is no doubt that the three models behave differently, both in the way they compress and the way they distort, so having all three really does give you different options to play with. The differences between the three models are both subtle and meaningful; but by and large, the LA-2A either works on a sound or it doesn’t, at which point choosing among the models is akin to choosing different varieties of coffee bean. One’s a little bolder, one’s a little smoother, one has a bit more kick, but in the end they all taste like coffee.

The LA-2A in dance production

Enough of the back story. Let’s get right to it and break down how the hardware LA-2A and its plugin emulations fare on the various elements that go into typical modern dance music productions. Attack asked me to investigate how well my hardware LA-2A and the various software models work on dance music sound specifically. Here are my thoughts on some of the most commonly used types of sound in dance productions. For now let’s talk in general terms about all LA-2as, whether they be vintage or modern hardware, or a software emulation. We’ll come to the specific strengths of each different model shortly.

Kick, Snare, Hats, Individual Hits

As in the world of acoustic instruments, the LA-2A can add exactly the right amount of grit, growl, and snap to a drum – whether it’s sample-based or a synthesised electronic sound – or it can squeeze it in a way that just makes it sound smaller and lose its impact. That makes this category of sounds hit or miss, and very subjective. You just have to pull up an instance or two and see what happens, which is generally my favoured approach to mix engineering anyway. In the sound clips we’ll come to on the next page, I’ll show an example of when the processor works (to my ears), and when it doesn’t.

Bass

The LA-2A has always been one of my favourite bass levellers for a few reasons. Firstly, the tube harmonics (i.e. distortion) it adds allow the bass to cut through a mix better with more apparent loudness but at quieter levels. Secondly, something about the attack and the knee (the intensity with which the compressor first ‘grabs’ a signal crossing the threshold) is close to perfect for the lowest notes in the spectrum, frequently making it punchy and tightly defined even when the source is a little flabby or indistinct. Only having to adjust one knob also makes it dead easy; the trick is knowing when you’ve made the movement of the bass consistent but still interesting, and being able to identify when you’ve gone too far and simply made it flat and lifeless. In the sound examples we’ll hear the nice, gentle rounding of these plugins doing something lovely to a bassline, and an example of the same LA-2A model pushed too far and crushing the life out of it, killing the size and definition.

Drum Bus/Drum Loops

This is the one area where I generally found the LA-2A to come up consistently short for dance music. If used solely as a distortion box, depending on the style of music you’re doing, it can be a great source of warm dirt. But for compression, most modern styles of dance music place great emphasis on movement in the drums, and the LA-2A tends to have more of a stiff, flattening effect on drum loops. This is one area in particular where the preset nature of the beast works against you; the fixed release time on the LA-2A was designed to make for relatively transparent compression, and in my experience most producers today want their drum compression to sound more aggressive, more alive, and more pumping rather than flattened out. So for distortions it can be a winner, but for drum loop compression the hit-or-miss is mostly a miss in my experience. In the sound examples I’ll show how this plays out, and also give an example of just how differently my own plugin, the UBK-1, fares on the same material.

Electric, Acoustic, and Treated Guitars, Piano, Plucked Strings, etc

As with acoustic instrumentation, this class of sounds is one of the areas where this compressor invariably shines. If you produce any kind of disco, chillwave, sample-heavy IDM, or anything involving samples or models of stringed instruments or plucked sounds, you owe it to yourself to begin crushing things with an LA-2A starting yesterday. It adds an amazing shimmer to these midrange instruments, and you can dig in with absurd levels – 20 dB or more of reduction – and the result is a perfectly clamped wall of sound that parks easily in the front or back of any mix, and always has the right amount of transient pop to keep a sense of life and sparkle no matter how crushed. Any time I need to compress an acoustic guitar or piano in my own productions, my hardware LA-2A is the one of my first choices, and the only time it doesn’t get used there is when it fares even better on the vocal or bass… which, as often as not, it does.

Synth, Arps, Pads, Toplines

Here I found the results still to be mixed, but especially if I’m using the drive to add some dirt and depth to a virtual sound, the LA-2A can put a smile on my face. When it doesn’t work is in the very frequent case of processing loops and samples that are already heavily treated, or that have some pre-applied pumping going on. In those cases, adding more squeeze with this opto leveller quickly becomes too much of a good thing, and the sound becomes so compressed it takes up too much real estate in the mix and is hard to ‘seat’. But for more raw sounds, or for virtual instruments that lack the fatness and textures that analogue synths tend to exude with no effort, getting heavy with the LA-2A can add a palpable sense of realism and grit to an otherwise sterile, boring source.

Vocals

In my humble opinion the LA-2A is one of the sweetest, most foolproof vocal compressors ever made. Much the same as with guitars, its transparency and lack of pumping effects allow you to dig in to unreasonable degrees and, unlike drum loops which get smaller and smaller as you turn the big knob, the voice just gets thicker and denser. It’s not the way to fly if you want scorching, white-hot, in-your-face vocals, but if you just want to take an unruly or wiry vocal and make it well-behaved with a generous amount of thickness, it’s almost impossible to get it wrong with this compressor. You can go too far with it, but it won’t be bad compression, it’ll just be too much compression. For those who aren’t super-savvy about all the nuances of compression and all the controls, the simplicity of this box makes it a no-brainer for vocal processing.