achieving massive choir sound

[deetak]

Does anyone have any suggestions on how to get a sound like the male chorus in this song? (I don't mean the proper vocal harmony, 4 part choral harmony or whatever. I mean the sound of having a seemingly huge number of singers like that.):

http://www.youtube.com/watch?v=BKhgxAV_OnY (I am not supporting NK's policies or anything, it was just an example for the audio effect)

I use REAPER and have one mic (no preamp, and not a particularly good one). Would I just have to record the vocal part over and over until I'm at the number of men used in that choir? I am completely new to recording so I have no idea what to do, how/if I should apply reverb on the tracks, or anything. Is this possible if there is only one person available to record (the most important condition)?

Thanks

[jast]

It's possible within reason, yes. You can only get authentic choir sound with many different voices (that's pretty much the nature of choirs, after all) but something similar can be done alone.

First off, yes, you need to record that part many times (simply replicating it via a couple of delay effects will result in garbage). To make the parts sound different, alter your vocal tone in a few takes and keep slightly changing your position relative to the mic. Similarly, pan the tracks differently so it doesn't sound like all singers are standing in exactly the same place.

So that's one thing. The next thing is vocal tone in general. The sound in your example is classical singing (think opera singers). This makes up a large part of the sound characteristics. If you haven't been properly trained in this kind of thing, basically pretend to yawn while singing to approximate the sound (as usual, when doing weird things with your voice, remember to rest regularly and drink lots of water).

Depending on the size of the choir you want to fake it may take too long to record that many takes. In that case you could use delay effects on a couple of them (wet level +0dB or a bit less, delay time in the range of roughly 20~50 ms). Don't fake too many, though, or the result will sound robotic. You can also use a chorus effect on a couple of takes (in the JS section), though I have no experience with those; I tend to hate chorus effects if they're not masked by something else.

Next thing to do is simulating room sound. Choirs tend to be recorded in open spaces, i.e. lots of reverb. The best choice is to find decent reverb impulses of a church or something like that and use them in the ReaVerb plugin (and adjust the wet mix gain to taste).

Good luck with that, then.

[Billy's Little Trip]

I've played around with Clone Ensemble
It lets you add up to 32 voices, bass, alto, macho, feminine, etc. The only problem is that it's REAL heavy on the CPU. You pretty much need to start a separate project just for the vocals, get them right, export them as a wav, then move them to your main project.
The free VST is limited, but if you know the right places to look, you can get a full working plugin.

[jb]

Listen to this: http://johnorama.com/songs/sensitive_pop/the_blitz.mp3

If you think I pulled off a choir sound in that, then here's my response. Sorry Jast, don't mean to dump on your advice. It's basically sound, but in my opinion it's a little overboard.

First off, yes, you need to record that part many times (simply replicating it via a couple of delay effects will result in garbage). To make the parts sound different, alter your vocal tone in a few takes and keep slightly changing your position relative to the mic. Similarly, pan the tracks differently so it doesn't sound like all singers are standing in exactly the same place.

You really don't have to bother with that. Just sing it a bunch of times. Sitting in a different place isn't necessary because you're human, and you're going to be in a slightly different place each time because you're not a machine. You're going to sing in a slightly different tone each time because you're not a professional. Just do a bunch of takes and-- this is the important thing-- have at least three parts. A duet with a bunch of takes probably won't sound choiry enough. It's still not bad though.

So that's one thing. The next thing is vocal tone in general. The sound in your example is classical singing (think opera singers). This makes up a large part of the sound characteristics. If you haven't been properly trained in this kind of thing, basically pretend to yawn while singing to approximate the sound (as usual, when doing weird things with your voice, remember to rest regularly and drink lots of water).

If pretending to yawn works, go for it. If you're not a classically trained singer and you try to get a classically trained sound, you're probably just going to be disappointed-- find another artistic solution to your problem.

Depending on the size of the choir you want to fake it may take too long to record that many takes. In that case you could use delay effects on a couple of them (wet level +0dB or a bit less, delay time in the range of roughly 20~50 ms). Don't fake too many, though, or the result will sound robotic. You can also use a chorus effect on a couple of takes (in the JS section), though I have no experience with those; I tend to hate chorus effects if they're not masked by something else.

Man, just get fast on the whole "make a new track" thing in your software, and sing the part over and over and over. It won't take that long.

Next thing to do is simulating room sound. Choirs tend to be recorded in open spaces, i.e. lots of reverb. The best choice is to find decent reverb impulses of a church or something like that and use them in the ReaVerb plugin (and adjust the wet mix gain to taste).

This one I'll agree with, but don't overdo it. Just having a shitload of tracks will go a long way to making it sound echoey and stuff. Create one f/x track and pipe all the choral takes through it.

That's basically how I do it, when I do it.

JB

[jast]

Listen to this: http://johnorama.com/songs/sensitive_pop/the_blitz.mp3

If you think I pulled off a choir sound in that

Cool song. Too much loudness to say a lot about the choir sound, in my opinion. Sorry.

Just do a bunch of takes and-- this is the important thing-- have at least three parts. A duet with a bunch of takes probably won't sound choiry enough. It's still not bad though.

In my suggestions I was actually going for a choir sound in a single part, i.e. unisono. I'll agree that a multipart piece is probably easier to fake. I haven't actually tried to fake choirs yet.

[Teplin]

I use REAPER and have one mic

I use Reaper too, and I've got a trick for recording a lot of voices fast.

First you set up a track so that you can hear the other voices you've recorded without having to stop and record again. Right click on the track and make sure "Selected tracks are free item positioning enabled".
Select the section you want to record to (left click and drag), click on "insert" (in the menu), and click on "Empty Item". Right click on the item you just created, select Item Settings --> Play All Takes.
Toggle the repeat on (hit r), hit record, and just keep singing until you have enough voices.

Now you don't have to break the flow to record another track. This method will record all the items into one track, but if you want to they can be dragged to their own tracks for panning, etc.

The only other advice I have that hasn't been mentioned so far is that singing an octave up or an octave down from the main part makes it sound a bit more like lots of different men singing. These are the techniques I used to make it sound like more and more singers come in at the end of this week's Howl Down the Chimney track.

Good luck!

[signboy]

Antares Avox Vocal Kit makes a pretty decent fake choir. There's some demo examples in the link, but they aren't really used for what I think you want. I used it for backing vox in Robot Candidate. Although it can be found, it is a bit pricey, so maybe the multiple takes option is a good idea. Besides, it sounds like Jast really knows what he's talking about here.

[jast]

Besides, it sounds like Jast really knows what he's talking about here.

No, I'm actually just extrapolating from other things I've done. All I do is way smaller ensembles and I'm typically going for a non-classical sound.

[Project-D]

Going off J.B.'s advice, if you can get two friends to sing it with you doing what he suggested, it will go a long way to getting the sound you want. I'd imagine if you were in a good sounding, live room using the room's natural sound would be better than artificial echo - stand back from the mic a little more than you normally would to get some room sound. Although you wouldn't have to use a classical singing technique, you would have to take care with your diction, it could get muddy sounding after awhile.

[jb]

Going off J.B.'s advice, if you can get two friends to sing it with you doing what he suggested, it will go a long way to getting the sound you want. I'd imagine if you were in a good sounding, live room using the room's natural sound would be better than artificial echo - stand back from the mic a little more than you normally would to get some room sound. Although you wouldn't have to use a classical singing technique, you would have to take care with your diction, it could get muddy sounding after awhile.

Yeah, that's a good point. One of your problems will be with lots of "s" and "t" and "p" sounds all over the place, either too strong because you've got three takes all hitting an "s" at the same time (sybillance), or scattered like a tommy gun because you weren't precise. You can set up a gate to remove these problems, maybe leaving one track alone. Or you can just be careful when you record. Or you can just cut off the problem areas-- chop the wav on those tracks, and rely on the rest of your arrangement to cover problems.

If you do the "stand back from the mic" thing, you should do like three takes or so close to the mic so you get the proximity effect, which will make it sound like there's more roundness to the sound, emphasizing the lower frequencies in those tracks. Mix as needed, taking advantage of panning and levels to make it sound like the parts are where you want them in the stereo field.

A lot of this depends on how fussy you want to be with the recording. You can be REALLY fussy-- there is a lot of tweaking to be done if you're searching for perfection. Myself, I'm usually pretty happy with messy, as long as it sounds giant and glorious.

Keep in mind that doing this yourself just singing normally, you'll be able to get a pretty close facsimile to a high school, amateur college, or amateur church choir sound. Those kinds of groups are all just normal people singing normally. Not wavery classical types, although the church groups swing older in membership so you'd have those older wavery voices in them. And almost none of them have really great diction.

[deetak]

Thank-you very much for the advice Jast & JB (a whole primer), Teplin (nice tip), Signboy, Project-D. I was pressed for time for this entry so I couldn't really experiment them, but at least there is this thread now for choir related comments.

Billy's Little Trip: That VST was extremely useful, thanks.