Digital guitar amplifier simulation has been getting better and better over the years and Apple’s Garage Band is no exception. However while sound-wise these simulations may be getting very close to analog amplifiers, the affordances of an all digital system – particularly on a computer – are still very different.

I still prefer the directness and simplicity of a real amp, but there are situations where simulation is the more practical option. The most obvious is during recording: it is much easier to DI the guitar than set up an amp and mic in a separate room.

The other obvious advantage of the digital domain is the range of amps you can use. Many options that would be both expensive and far too loud to be practical to use real versions of are available. I would love a big tweed amp for instance, but even my 1 watt tube combo is almost too loud for home use.

While amp simulation is easier to record with than a mic and real amp, I've found I do need to do more messing about than I do with a real amp. There are two key facets of the virtual rig that I have found most useful to tweak: amp presence, and mic choice and placement.

Presence is kind of tricky to explain, I found Bjorn Juhl‘s description enlightening:

The output transformer does many things, but basically it’s an interface. … Most often the transformer colours the EQ. The internal feedback network of an amplifiers’ power amp tries to correct these losses. The presence control in an amplifier reduces the amount of feedback in the treble frequencies. Upper frequencies are now limited by the output transformer. The Treble ceiling is now set by the limits of the transformers. During light overload the transformer will compress since its unable to reproduce transients rapidly. This softens sharp edges but as overload is lessened, the transformer continues to compress a while.…

Adjusting the presence of the virtual amp is my first stop to correcting harshness, or overly bright response. In fact I often leave the EQ relatively flat. At the volumes real amps operate at I have found it difficult to really tell exactly how a presence control effects the sound, but at the controlled levels of digital its very clear.

The choice and placement of the virtual microphone is the other useful filter for shaping the tone of the amp. The common choices and placement are all available, and I spent quite a while trying out different things suggested in Premier Guitar’s recent mic placement tutorial video.

Because the microphone and cab have such a big part to play in the tone filtering, I have found that the amp’s EQ section is best used as a fine tuner after everything else is in place.

One area amp modelling seems to be an order of magnitude easier is recording heavily overdriven guitar. Avoids the small sound that is so easy to produce miking an amp.

Garage Band’s skeumorphic amp interfaces obscure a nice feature: the EQ and reverb sections are configurable. The labels above the controls are actually drop down menus that allow you to choose alternatives to the default. Great for taming a splash spring reverb or using a modern EQ stage on an old amp model.

One frustration I have currently is it appears that effects loops, and post amp or microphone effects are limited to what comes with the preset amps in the library. For instance, the Clean Guitar library includes an amp called Dyna-Trem. This amp includes what I believe is a post-distortion stereo envelope following tremello.

The envelope trem appears in the amps summary interface, but is not in the amp head or effects. It appears the only way you can use it is to start from the Dyna-Trem amp in the library. Similarly I have not yet found a way to apply, for example, delay after amp or mic without starting from a library amp with that effect on it.

This seems to be a regression too; earlier versions of Garage Band – with worse amp models, and much worse interfaces in general – did allow you to have greater control over the signal path.

The overdrive, distortion and fuzz stomp boxes seem a bit rubbish. Ironic perhaps that the digital simulation of tube amps is pretty good and simulation of little transistor based circuits is average.