I'm doing a demo for a local act and we've tracked layer upon layer of overdubbed guitars: there are 10 rhythm parts with various chord voicings, and 10 lead parts playing variations on the solo and riff hook. A few of the layers are duplicates, but we had four different guitars, playing bar chords, open chords and power chords for the rhythm parts.
My question is: how many should I use? The lead tracks are mostly duplicates and there isn't much distinction between them, so I'll comp those later; it's the rhythm that's bugging me. The parts are tight and played on nice instruments, so the issue isn't so much of musicality, it's of fitting all the variations into the mix without it sounding like mush. Do I try and fit them all in? Or comp them down to make one or two awesome tracks? It's essentially a bog‑standard rock sound, so double‑tracking the rhythm makes sense, with each part hard panned, but how would you incorporate the other rhythm tracks? We DI'd the guitars, as we can then use IK Multimedia Amplitube to change the tones. I assume I would try different Amplitube settings for each pass?
Via SOS web site
SOS contributor Mike Senior replies: In my experience, unless someone's put in a fair few hours of punching‑in and/or editing, most tracked‑up walls of guitar aren't tight enough to sound punchy. Even if the timing of the individual picking transients is on the money, the lengths of the notes or the point at which the strings are damped doesn't match up nearly as well and also affects the rhythm. So my first suggestion would be to focus on the tightness of any layers of the same basic guitar part. This is especially relevant if you're stereo panning them, because human hearing is extremely sensitive to inter‑ear timing differences. If you're having trouble with mushiness, you may find that there are some tuning problems too, so be critical and ditch anything sour. The closer the tuning of your parts, the more the pitched elements of the part will reinforce, and this will help keep the harmonies of the part clear despite any layering you may do.
For a middle‑of‑the‑road rock sound, it's typical for the stereo field to be balanced by putting double‑tracked rhythm parts on opposing sides of the image, and if you're going to have more than two parts going at the same time it'll probably give you a more satisfying spread if you don't pan them all to the same two places. You'd also expect the guitar arrangement to fill out for the choruses, so you're sensible to think in terms of adding more overdubs for those sections, plus you could add more extreme‑panned layers here while leaving the verses slightly narrower. That said, while added overdubs can help increase the illusion of size, they will tend to make the composite sound more bland and homogenous, as well as pushing the guitars away from the listener: using more guitar parts means that each has to be lower in level to avoid making the rest of the mix sound small. Every producer tends to make their own compromise in this regard, so you can only make an informed judgement by comparing your mix to a few commercial tracks in the style.
Beyond that, if you've got that many parts available to you, I'd use them to bring some light and shade to the arrangement. Most riffs are made up of smaller musical figurations and fills, and you can really bring them to life if you give each of the different sub‑sections of a riff its own character, by altering the balance of the parts from moment to moment. The easiest way to do this is to line all the tracks up together with different modelled amps and then edit different bits away from each track. If you use clearly contrasted settings from Amplitube, this could make life easier. Because your guitars are DI'd, you could also use them to drive virtual stomp‑boxes and amps at different moments, by splitting the DI audio between sequencer tracks with different plug‑in settings.
You have access to the amp settings, so you shouldn't need much processing during the mix, beyond perhaps a little low cut to tame any general woofing around that might interfere with your bass part. However, in my experience, layering up parts that all use the same amp‑modelling engine seems to make it trickier to get a really solid sound, so I'd experiment with other modelling options as well (even comparatively lo‑fi ones) — or, even better, try re‑amping a few of the parts through a real amp and speaker to catch the sound of some real air moving.