I've always been amazed at how nice, powerful, and up-front Thom Yorke's vocals sound in Radiohead's song 'Exit Music (For A Film)'. Any ideas on the effects used on his vocals? I'd love some thoughts on how to achieve similar-sounding vocals — besides cloning Mr Yorke, that is!
SOS Forum Post
SOS contributor Mike Senior replies: The vocal in this track is a fairly unusual one, as I discovered when comparing it to various others for the 'Using Equalisation' article back in SOS August 2001 (on-line at www.soundonsound.com/sos/aug01/articles/usingeq.asp). There's a lot of low-frequency energy there, which gives the vocal a real weight, but there are also a lot of extreme high frequencies as well (above about 16kHz), which pulls the voice right to the front of the mix. (The high-frequency damping effects of the air mean that you'd not normally hear lip noises like that from a singer unless they were pretty much singing right into your ear!) Of course, it also helps that the mix level of the vocal is unusually high compared to the acoustic guitar, although that level imbalance seems to me to get less stark as the track progresses.
You could do a lot towards achieving this kind of sound while recording. You should look for a condenser mic with an extended low-frequency response and an extreme high-frequency boost. This is likely to be a large-diaphragm model, as they tend to have more of an on-axis high-frequency peak (by nature of the design) than small-diaphragm models. Bear in mind, though, that a lot of large-diaphragm condenser vocal mics concentrate their HF boost a bit too low for this application (around the 6-15kHz range), so it'd be worth trying different ones out if you can, to find the best fit. If I had to make a blind decision on which mic to use, in my experience I'd say that an AKG C12 (or roughly equivalent C414B TLII or C414B XLII) or MXL V77 mic would do the job. The C12 and V77 are both valve mics, too, and these might provide a little of the high-end enhancement/distortion that can be heard in that Radiohead sound.
If you've ever seen Thom Yorke play live, you'll have noticed that he tends to open his mouth quite wide and tip his head back so that he's roughly pointing his nostrils at the mic's diaphragm. By doing this, he's brightening the sound of his voice reaching the mic, because high frequencies are bouncing off the roof of his oral and nasal cavities and beaming towards it. While your singer might not want to allow you to inspect his fillings in this way, you can clearly just set the mic up slightly below nose level to take advantage of this effect as well. Working directly on-axis to the mic is another way to help maximise the high-frequency capture, because large-diaphragm mics tend quickly to get duller off-axis.
The prominent low frequencies suggest that the original 'Exit Music' vocal was recorded very close to a cardioid mic's diaphragm, where the proximity effect would give this kind of bass boost, and the distinct lip noises corroborate that view. You might also find that Nigel Godrich (Radiohead's producer at the time) dispensed with a pop shield, to avoid impeding the lip noises and other extreme HF in any way. I believe it was Michael Jackson's long-time engineer Bruce Swedien who said that he once even took a mic apart to expose the capsule, in order to achieve the most intimate sound possible, but I'd not try this personally, for fear of damage from dust and moisture reaching the diaphragm. If you're not going to use a pop shield, though, you'll want to discuss with the vocalist how they might be able to contain their plosive 'P' and 'B' sounds, to prevent too much wind-blast noise on the recording.
As for processing the recording, I think you're probably looking for emulated analogue EQ and compressor plug-ins, as I believe Godrich is a bit of an analogue gear nut and likes his esoteric rack units. Again, modelled valve units might help to get a bit more fizz at the high end in some cases. Although you should be able to get pretty close to the 'Exit Music' tonality just with mic choice and technique, I suspect that you'll still need high-frequency boost above 16kHz to get all the way there. It's also worth noting that lots of low-frequency content on a lead vocal can be difficult to blend into many mixes, so either leave enough space for it in your arrangement, as Radiohead have done, or be ready to filter out some low end, otherwise you'll risk a woolly overall production sound. After that, you'll certainly need a fairly hefty dose of compression, which will help to bring up the lip noises, and maybe a little de-essing as well, given the brightness of the sound. If by this point you've not got enough of that high-frequency distortion, you could always add it with a psychoacoustic enhancer or some kind of saturation plug-in.
Thinking about effects, you can hear a fair bit of reverb applied to the lead vocal, but this doesn't really have the consequences you'd normally expect; namely distancing the sound from the listener. The very dull-sounding reverb provides a real contrast with the prominent high-frequency content in the source vocal part, thereby emphasising the apparent closeness of the vocal. This is a clever mixing tactic which attempts to work around the adaptive capabilities of the ear. When you listen to a very bright mix for a while, it begins to sound less bright as the ear compensates for the excessive brightness by decreasing its high-frequency sensitivity. (This is one reason to reference your mix against commercial recordings when mixing.) By putting a very dull reverb on the lead vocal, Godrich has encouraged the listener's ear to interpret what it's hearing as a bright vocal in an otherwise tonally balanced track, rather than just an over-bright track, for which it would try to compensate. The result is that the vocal remains subjectively bright (and therefore up-front) even as the ear acclimatises to the overall sonics of the track.