Achieving a convincing crowd-effect is a little tricky but not at
One of the most natural effects used in popular music from time to time is that of the large crowd sound. While a real life crowd with natural reactions is certainly the best way to go, not all of us are currently cutting a live album. Luckily, you can recreate a very similar sound and feel in your home studio without the need of hundreds of people singing simultaneously. But with that said, you will need as many different people as possible to make it more convincing, so no solo jobs here; simply grabbing a couple of bandmates and overdubbing them countless times is not only time consuming but will not sound too realistic.
By far the easiest way to achieve the crowd sound short of a live record is to have at least six or so people and record them simultaneously. Of course, the more the merrier if you have the space and equipment for it. Not only will you get more voices in less time but the result will sound much more convincingly crowd-like because of the natural variations between peoples’ voices. Still though, you probably won’t get a full crowd sound with just one take so you will likely need to layer a few of them together. A good idea is to reposition the singers slightly after each take for added variation. You can try changing mics as well.
If you’re going for a crowd sound more akin to something you’d hear during a sporting event or those atypical athlete songs, they tend to include lower harmonies in order to thicken the texture so if you haven’t added some lower vocal emphasis, consider that an option.
Of course, the first problem with this whole situation is finding a way to get a cue mix to the singers as most of us probably don’t own enough headphones and headphone amplifiers to give each performer their own. A simple solution would involve having the performers listen to a demonstration until they are comfortable with the parts. With any group of singers, there’s always one or two who tend to lead while the others follow, so when you finally do hit that record button, whatever headphones you do have on you should go to them. If your headphones are loud enough, any sound that bleeds through should help those without them keep timing. If getting everyone in sync is still a problem, put on some headphones yourself and signal the timing to the performers instead.
This fix usually works if your performers are more or less confident and warmed up, but it should be said that the most common problem with this solution is that the singers without headphones tend to feel a bit overexposed without any cue mix and will likely perform a bit timidly – not so great when you’re going for a lively crowd sound.
An alternative to this method would be to use a speaker while recording. The problem that this route will give you is obviously some monitor spill, and although you can try and face the speakers away from the microphones in order to reduce the pickup (in which case a figure eight mic would be best, or better yet, something like the MXL REVELATION which comes with variable pattern control, including the figure eight), you'll most likely find some of the cue mix bleeding into the background of your takes. This creates two notable consequences: for one, you’ll need to make sure that arrangement of your backing track doesn’t change all that much after the overdubbing sessions, otherwise the bleed will likely create a strange sounding shadow of any parts that will later be removed; also, you’ll probably need to do some trial and error with the mic distance and speaker level to keep that bleeding within a reasonable limit.
Seeing as how there’s no way of avoiding the bleeding, it is recommended that you record long enough on either side of you vocal tracks that you have ample room to decide exactly where you want to fade the spill in and out during the mixing stage. It may end up sounding strange is the spill cuts out too suddenly at the end of the last phrase instead of, for example, waiting until a song section boundary.
Whether or not speaker bleeding is an issue, you will most likely be better off trying to record the vocals as dry as possible since most small-room sounds probably won’t help in achieving that crowd effect you’re after. By recording dry you leave yourself more flexibility simulate a larger, more realistic sound artificially. As far as what effects to use at this stage, most people will automatically run for some reverb but you’ll have better luck going after a convincing crowd-sound by using some slapback delay instead. Go for something around 100ms. If you want a more aggressive tone, think about sending the delay’s output through an amp modeler too. From my own experience, I found that a decent amount of slapback does enough of the job that you can later use reverb to smooth things over or to give the impression of a large room – both of which can also be done with an effect that has a somewhat quick decay in order to avoid muddying up the mix.