Vocals are probably the hardest thing to make "sound good" in a mix, specially if the mix isn't precisely sparse. Specially in rock music: the electric guitar (specially clean) falls in the same main range and it's a win or lose matter.
When several tracks have to "fight for space" in the mix, there's several things you can do:
- Use panning. Moving an offending track to one of the sides makes room in the center for vocals. This works fairly good, but is not enough most of the time.
- Make them fit!
To make things fit, the trick is the EQ. Sounds are made by many frequencies, but not every frequency is "important". What we have to do is enhace the important frequencies and trim the non important, so we create more space in the mix to add more stuff.
If you only use the recorder, and specially if you recorder is small (think of the MBR or the BR-80), I strongly advise you to create your mixes in the computer. There's plenty of free software to do this, and the results are quite good in comparison with what you can achieve with the recorder alone. Audacity (http://audacity.sourceforge.net/) is free and all you have to do is export your tracks individually (with the suited tool) and then import them in Audacity.
You need a parametric equalizer. Most DAWs have one and there are plenty of free ones. A parametric equalizer looks this way:
(http://www.rs-met.com/graphics/screenshots/EasyQ.png)
To bring up your vocals there are no rules, but this helps:
- Do a steep cut of low frequencies. For vocals, nothing below 80Hz should sound. I mean cut this completely. It usually contains unaudible rubbish which brings up the level but offers nothing. (Highpass 12db/Oct)
- Main information is in the 400Hz to 1200Hz area for most vocals. A small bost with a wide "Q" (the with of the bell-like adjustement) will increase the volume in a clearer way than just pulling up the track volume slider. If you have to fight against guitars (specially electric clean guitars), doing exactly the opposite in guitar tracks also helps making room for the vocals. But be gentle
- Articulations are way higher in frequency, from 4000Hz and even 5000Hz. A very thin enhacement around 4.5K brings up the consonants and articulations. If a listener claims he can't understand your lyrics, it means that the vocals need more presence.
I usually make these adjustements, then tweak (each voice is different, so you'll have to tweak!):
(http://www.ojodepez-fanzine.net/almacen/easyq.png)
Next thing you should do is using a de-esser. A de-esser is a processor which helps you limiting those nasty "S"s. I use this one, which is free: Spitfish (http://makeownmusic.org/free-plugins/spitfish-de-esser-free). Just ramp up the controls until the S's are audible but not disturbing.
(http://makeownmusic.org/wp-content/uploads/2010/01/spitfish.jpg)
This usually works :)
Almost done! Now add some compression. Compression helps you level up your vocals, attenuating the dynamic range. Compression is a whole topic so you better read about it to understand how it works.
Hope this help. Ask away if you need further assistance.
This is a very good tutorial!
And the plugins are good too - DigitalFishPhones compression is excellent "Knob Oriented" stuff.
And EasyQ looks good - I use the now free BlueCat Audio, which is very similar.
Thanks, great offering!
This looks good but I still struggle knowing what to tweak. There is EQ settings on the BR800 and you have a setting for High - mid -low and a frequency for each one to adjust, so which one would i adjust on. OR do i just reduce the db on the frequencies which are not required.
EQing makes my head spin.
EQ can be quite confusing - I find the usual low/mid/high basic equalizers quite lacking as they are usually very limited in what you can achieve. They are OK for live performances (when found in mixing boards) and simple arrangements when recording your songs, though.
If you have the ability of manually centering each band and adjusting its width, it's enough and you can follow this tutorial, but most low/mid/high EQs work with fixed values and quite wide "Q"s (the width of the bell of frequencies the parameter affects to).
Lowering the "low" would trim important frequencies, and raising the high would bring up unwanted noise. I really can't work them out. Maybe somebody more used to work with this kind of controls would offer some insight. Anyways, I'd try gently lowering "low" and gently rising "high" on vocal tracks, and gently lowering "mid" on guitar tracks. And if available (I don't know the BR800), try to record everything but bass and drums with a lower end cut (high pass above 40Hz).
That's why I mix and master in my computer. Multitrack recorders are a joy to record your music, but mixing and mastering is another story.
Oh this is really interesting thanks - I just bump up the mid eq on the BR80 toggle switch but will give it a go in Audacity and see how I get on. :o
Please forgive me as I am new to vocals with my MicroBR... you mean export the raw wav for the vocals into the computer to mix it down, and compress, then import back into the BR??
You could do that, of course, but I'd rather export everything and do the full final mix in the PC. There are several free DAWs which are easy to work with once you get used to them. At first, you will just import every track, apply some EQ, move levels, and mix down - but step by step you'll find yourself doing more "pro" stuff.
It may sound intimidating, but I strongly advice to use the computer for mixing and mastering instead of the BR. The BR is *perfect* for recording, but weak when it comes to mixing.
Besides, there's an interesting pro about this approach: you are not required to bounce your tracks to find more space. I've done tons of tracks on the BR just recording a guide track to T1V1 first, and then using that guide track to record as many tracks as needed in T2, T3 and T4. 3*8 = 24 tracks. When you are done, you export everything to WAV and mix it in your computer.
I haven't used the free Audicity program, but when I have the time I'll give it a go and write a full tutorial explaining the easiest, most straight-forward way to use your recorder to lay down tracks and your computer to mix and master them.
Very good information here, thanks for sharing.
Yep, all good stuff - kind of how I've ended up approaching it.
But, if you haven't heard of it, I also use the "Exciting Compressor" technique:
http://www.recordinginstitute.com/R2KREQ/excomp.htm (http://www.recordinginstitute.com/R2KREQ/excomp.htm)
I could never get my rock vocal to cut through a mix and also sit nicely (not too loud) in the band... until I found that article. I do a lot less to the original lead vocal track nowadays.
You need two copies of the vocal (I used to do this mixing on an MBR, stereo band mix on 3/4, lead vocal and lead vocal copy on tracks 1 and 2, so it is possible :))
Track 1 is the "natural" vocal - very little eq, very little compression (mebbe light limiting if it wasn't when tracked), as much reverb/whatever as you want. Set this to the "right" level in the mix - IE not too loud(!).
Track 2 is the "exciting compressed" vocal - first compress the cr@p out of it (INF:1), and then add a sh1tload of top-end. I have two ways I might "add top-end" it a) Boost (shelf) EQ above 5KHz or so, +10 db at least, and cut EQ below 250Hz or so, -10 db at least. b) The other way is to actually hit the frequencies that matter on the vocal - I tend to hit 5KHz with a Q of 1.0 nowadays - this is where vocal "presence" is. 7KHz is good too - but it's where the Ss are. 10KHz is good for light brightness. Again, big boosts on these - we're not being subtle, remember this track is compressed to hell.
You bring track 2 up to "taste" and your vocal keeps it's natural character as you recorded it but it jumps out of the mix like you are sat next to the singer without obscuring any of the rest of the mix. It kicks the vocals through lush mixes and guitar heavy mixes - it even allows you to lower the vocal and therefore get the whole mix louder.
I've also found recently that you can apply the same trick to other stuff - eg DRUMS!!!. There's a trick called the "New York Compressor". NY studios started doing this to drums in the late 70s apparently. It's basically the same motown trick. In this case, on the compressed copy of the drums, you boost the frequencies you want (Q1.4 or so). On the one I'm working on I hit 80Hz for the kick, and 400Hz for the snare. I can't remember whether I did anything at 7KHz for hardness on the cymbals.
Again, bring the compressed track up to taste, and listen to that drummer kicking some serious @rse :)
It lets you get a big punchy drum sound without having to raise the drums too much - and (might be my imagination) it seems to make the BR1600's digital drums sound a lot more natural.
Final last tip on vox - 3KHz (Q1.0) is where the "breath" is. If you've got backing vox and you want the lead distinct from them, slight boost at 3K on the lead, cut 3K on the bvox mix - all fits together much nicer then.
Remember though, all this stuff is just guidelines and other people's experience. I initially used stuff like this as "recipes" - but now I experiment and twiddle the settings. Like Nathan says, every vocal is different. I have basic settings I reach for as a starting point (ALWAYS listen to the track first without any eq and compression though - it might not actually need it!), but they usually need some sort of tweaking to get it nice...
Interesting. I will read your post slowly and try this technique.
I think read the article first - it's probably clearer!! :D
Most of the stuff in my post is how I've used it since...
But yes, give it a go - it's astounding the affect it has, without messing with the original sounds too much.
Andy's advice is good advice ! I usually forget to apply them, but when I do it solves the mixing problem at once ! I've also used a variation of it. After numerous of Beatles covers, I've often adapted the Lennon approach to the main vocals: DOUBLE IT :) that is: record the main vocals twice, getting them as similar as possible. On a couple of occations (e.g. the last one (https://songcrafters.org/forum/index.php?topic=15974.0), where it really solved a problem getting the main vox to fit in the mix) I used the compression/exciter trick on one of the vocal tracks and leaving the other untouched (except for a touch of reverb).
Funnily enough, I've started experimenting with doubling the lead vocals recently. I've not used it in anger yet (at least I don't think I have... it was an accident if I have!), but I'm beginning to see the use of it.
Something I have used on the current thing I'm working on - a huge rock thing - is the idea of sticking a high harmony, above the lead, closely following it. I got the idea listening to Steven Tyler of Aerosmith - there was a period when he seemed to be doing it to everything.
It hasn't quite worked out like Steven Tyler, and I couldn't track the lead as well as I wanted (and the high part went to a top B!!), but I quite like the effect it has. I've not mixed the vocals into the track yet, but it seems to have the effect of making the lead richer and more noticable. And when the high part falls away, the lead on its own becomes a lot more intimate suddenly.
EDIT: Actually, make that a top C!!! It starts on a B, then the next chord is an Ab, and it went up, so it must have been a C - no wonder I was having trouble!!
I think all this is great info! Incredibly thorough na_th_an, I'm actually interested in the de-esser quite a bit but I haven't the foggiest idea about plug-ins and such. I have audacity but I hate doing my mixing on it. I mostly want to record the song , produce it & move on to the next one. I really have never felt I posses the ear to hear the frequencies clearly and try to record the sound I like and do no EQing. That does lead to some compromises on the vocal quality I guess. I have a very dynamic voice and I know my recordings could use some fine tuning but if it's to get done, now at least, it will have to be by someone else. That person has probably read the info provided here , nice link AndyR, I tried to read it but thought my head would explode!
Long story short (HA!) thanks for the info, perhaps one day I'll use it. At least I know where to find it.
Rock On!
i'm a bit like hook on this, i tend to do everything on my recorder. the one thing i do always try to do tho, is pan all the instruments (especially guitars) away from the middle to leave space for the vocals in the centre.
my experiences when collaborating is that when there are guitars in the middle of a track ive been asked to sing over, is i find it difficult to find space for a vocal, it can also be difficult to hear yourself properly when recording which can result in peaking. so i might ask that the instrumentation be repanned. that said bass in the middle is never usually a problem as it is a different fequency from my vocal range.
just an afterthough in many collabs i do, i send wav files of the vocal, so the person mixing the song may very well use some of these techniques. me? i'm too lazy , i just want the song done so i can get on to the next one.
Quote from: oldrottenhead on August 21, 2012, 07:54:24 AMthe one thing i do always try to do tho, is pan all the instruments (especially guitars) away from the middle to leave space for the vocals in the centre.
I always pull the lead vocal slightly (5 -10 maybe) to the left. The only thing I leave centered are the drums, I do pan auxiliary percussion around but whatever track drives the beat I center. In my opinion panning is the easiest thing one can do to give clarity to the parts thus giving unity to the song. Not the best thing (see above) but certainly the simplest.
Rock On!
Yep, I've applied "exciting compression" to you before now, Jim :D (It was the only way to get you nicely through the dense wall of guitars on Muletide so that it could become Outside!)
And actually, you've got a good point about panning the guitars and stuff away from centre. I don't allow ANYTHING in the centre (apart from bass) when the lead vocal is making a noise (I haven't much control over the panning of the drums on the BR1600 but, apart from the fact they've mixed it for a left-hand drummer, it seems ok, kick and snare in the middle, everything else vaguely where you'd expect it).
Unless I'm out to create an effect, I regard centre stereo as belonging to the singer, the kick, the bass, and the snare. Everyone else has to fit round where they can.
I was actually quite surprised you managed to find any room on Muletide to stick a vocal - I can't remember how it was done exactly, but certainly the lead would have been the "vocal", so smack down the middle, I guess. I was dreading having to remix the original instrumental (might have been impossible, actually) to fit you in as your vocal arrived, but the moment I dialled up two orh tracks beside the stereo mix, one orh with reverb, the other with heavy compression and drastic top-end boost... there you was :)
And Hooky, you just reminded me, saying about de-essers. I had huge problems over Ss originally. I even learnt how to side-chain on an outboard compressor (that's all a de-esser is - set it up to monitor one signal and then duck another one). But it's so fiddly.
And then I learnt that for most us, most of the S comes out of one side of your mouth. Record some "TeSSSt" vocals with the mic at different positions around you - you'll soon find where to place the mic so you don't get any naSSSties :) ... then you just learn to always address the mic like that. I haven't thought about de-essing for a couple of years now...
Quote from: Hook on August 21, 2012, 07:43:18 AMI'm actually interested in the de-esser quite a bit but I haven't the foggiest idea about plug-ins and such.
You don't need a plug-in. Your BR-900 has a de-esser built-in. Just select any patch in the MIC bank that uses the VOCAL MULTI or COSM COMP VCL algorithm then press ENTER to edit the patch. Cursor to [des] or [DES] (uppercase means it's on; lowercase means it's off) and press ENTER to adjust its parameters (there are only two - Sibilant and Level, plus an On/Off switch). See page 95 of the BR-900 manual (revision 6).
Quote from: AndyR on August 21, 2012, 08:20:18 AMAnd Hooky, you just reminded me, saying about de-essers. I had huge problems over Ss originally. I even learnt how to side-chain on an outboard compressor (that's all a de-esser is - set it up to monitor one signal and then duck another one). But it's so fiddly.
The BR-1600 has a built-in de-esser too. See page 96 of the BR-1600 manual (revision 10 or 'A').
And there's the Spitfish VST (http://www.digitalfishphones.com/main.php?item=2&subItem=5) for those that don't have this feature built into their recorders. It's been mentioned earlier in this thread, but I wanted to stress that digitalfishphones has the best "knob" based compression VST's that I've used so far. Just turn the knobs and hear the results. Each one has an advanced mode as well if you want to dig deeper. It will definitely keep you from using too much of the effect.
Ok, now my question. Does anyone know how to keep doubled vocals from sounding "phasey" - Obviously each track will not be identicle to each other, that's why you do it in the first place. But it's tough to get the panning and levels right to remove pitch phasing. I have found that EQ reduction in the under <1K helps a lot. I don't want to throw in the towel, because I like the way it sounds 85% of time. I just want the other 15% ;D
As mentioned, I usually leave the centre for the main vocals, the bass drum, the snare, and the bass guitar. Everything else is either side.
Quote from: 64Guitars on August 21, 2012, 09:40:34 AMThe BR-1600 has a built-in de-esser too. See page 96 of the BR-1600 manual (revision 10 or 'A').
Yep :) It's one of the first things I spotted when I got the BR1600, and I thought "hoorah!". But I'd just conquered the mic technique a few months before that - and I haven't ever recorded a vocal on the 1600 that's really needed it. (Who knows? One day I might :D) (Actually, even funnier - it's just crossed my mind that the default vocal patches I use during tracking, mainly for incoming compression and expansion, might actually have the de-esser included and switched on!! :D I'll have to checked when I get home :))
Quote from: launched on August 21, 2012, 10:05:19 AMOk, now my question. Does anyone know how to keep doubled vocals from sounding "phasey" - Obviously each track will not be identicle to each other, that's why you do it in the first place. But it's tough to get the panning and levels right to remove pitch phasing. I have found that EQ reduction in the under <1K helps a lot. I don't want to throw in the towel, because I like the way it sounds 85% of time. I just want the other 15% ;D
Now, this I don't know the answer to... I suspect it might be why I don't often go for doubling lead vocals (I double, triple, quadruple, more, more(!) backing vocals).
I'm also wondering whether slight phasing between the two is meant to be one of the attractions? (Not if it detracts from the thing itself, though :D).
My gut feeling is that EQ reduction is the secret - I understand that boosting EQ tends to introduce "phasing issues". Maybe "complimentary EQ" (take a little off one and put it on the other, identical frequency, Q, and amount) is the answer? That will make the two vocals slightly different, but maybe that's the answer....
And also, this is more of a query rather than an idea of what to look for (but it might be that too!), when people double-track the lead, how do they pan them? At the moment I'd be aiming for as close together near the centre as possible, but with just enough separation so that they, er, separate ever-so-slightly. I've found that with bvox pairs, panning them hard left and right sounds @rse unless I start adding more pairs to them - I personally feel like one voice panned hard left/right sounds terribly exposed and thin,
even if it's twin is panned the other way. Sometimes that's what I want, but not very often...
Quotewhen people double-track the lead, how do they pan them?
a thing i often do, something bowie used to do a lot round about hunky dory and early ziggy period. was to doubletrack the lead vox but sing them an octave apart and then pan each vocal full left an right.
he uses this technique quite a lot on space oddity to good effect
https://www.youtube.com/watch?v=cYMCLz5PQVw
Quote from: AndyR on August 22, 2012, 01:35:15 AMAnd also, this is more of a query rather than an idea of what to look for (but it might be that too!), when people double-track the lead, how do they pan them? At the moment I'd be aiming for as close together near the centre as possible, but with just enough separation so that they, er, separate ever-so-slightly. I've found that with bvox pairs, panning them hard left and right sounds @rse unless I start adding more pairs to them - I personally feel like one voice panned hard left/right sounds terribly exposed and thin, even if it's twin is panned the other way. Sometimes that's what I want, but not very often...
Yes, the whole idea was to keep them on top of each other, but a 5-15% pan seems to sound better and a wide pan just doesn't work well, maybe because it conflicts with guitar parts and other things. I've never panned mored than 50% with any vocal at all, it just loses power which is what I need in the first place because my vocals are weak.
I've tried your 3k reduction tip for bg vox and that worked very well, Andy - I also removed anything below 70-80hz as well. Haven't tried much for lead vox except a 3-5db reduction from the midrange down.
Jim, I've tried and use all the time octave matching. But I tend to keep them together at all panning levels if panned at all. Maybe I'll try keeping them separate...
for drums i have the snare and hi hats slightly of center to left. thats how they are positioned with the kit anyways. but is putting the bass drum and bass guitar in center confliting or a good blending? i put both in the center with vocals, sometimes vocals might be -5 or so, to left or right, slightly off center depending on the song.
Well, the snare might be off-center on a real kit, but depending on your song, it may sound off. I mean, if the snare is providing the groove, it might need the extra "importance boost" that centering it in the mix provides. But the rule is that there is no rules :)
About the bass drumm and the bass guitar sharing the center - well, they shouldn't interfere too much. If you are recording rock, pop, or folk music, the bass guitar should have its lower end cut to give the bass drum space. I mean, cut everything under 40Hz as much as you can in the bass guitar track.
In dance music, the bass tends to be more populated. In such kind of music, you usually use sidechain compression, in which the kick compresses dynamicly the bass track - which creates some kind of "pumping". My band don't do dance music, but sometimes we arrange songs in a pretty much electronic way. For example, this song uses sidechain compression in the bass track (which was generated playing live a Nintendo DS with Korg software). Note how the kick "mutes" the bass track somewhat:
Excellent thread - a MUST to re-read etc - great contributions All.
I personally like to double track lead vox... mostly because I feel my vocals are pretty bad. I usually need a little something extra to make them sound more interesting. I usually don't do much else to them except the eq'ing as mention'd by Nathan... Great Thread BTW buddy.
Can someone explain what pitch phasing is??? I'm guessing it's a warbbling noise you hear when two frequencies clash within a vocal track.
Quote from: Rata-tat-tat on October 10, 2012, 02:11:30 PMCan someone explain what pitch phasing is??? I'm guessing it's a warbbling noise you hear when two frequencies clash within a vocal track.
I like to double vocals because each track has a slightly different identity, and when panned a certain way it sounds good to me. The reason why I started doing it was because my vocals always sounded thin, but I've since found out that compression and eq will accomplish the same thing, only you don't get any cool stereo effects. I still can't produce single tracked vocals very well but I'm working on it and would like to do it that way too, because it's difficult to reduce phasing problems if my pitch varies too much.
I found some form of a written definition of the issue, obtained from this article (http://www.indiana.edu/~emusic/acoustics/phase.htm) :
"Two sounds with a small difference in frequency, say two piano strings of the same pitch, may be perceived as a single sound, but as the waves evolve, they move slightly in and out of phase with each other. The resulting constructive and destructive interference produce a pulsation of amplitude. This pulsation is known as beating. The rate or pulsation or beat frequency is the difference in frequencies. A string tuned to 440 Hz and one tuned to 441Hz will produce a pulsation once per second."
So I guess the proper term is "beating". But I don't like to say "beating", so I'll stick with pitch phasing ;D
been dbling tracks for quite a while. i notice sometimes i have vox all the way up and it is still a bit quiet. i copy the vox and add the exact copy as another track and both at the same vol. it is louder. i sometimes do this for drums making one stero tack a bit trebly and one a bit bassey. it adds depth. sometimes i dble up on bass when i think it is not loud or bassey enough.
on vox i might use one stereo track plain and the other have the reverb or whatever effect. on some songs i have had the 2 seperate vox tracks have 2 seperate effects on it. it is a good way yo add some volume without distorting, provioded you watch your vol. levels.
One tip I read about, (after cutting the lows out of pretty much all mid range trax), is to bring up the most important trax first. EG vocals and the main sense of rhythm usually.
Then start bringing the rest up in order of importance and pan and balance as needed.
As you bring up a track do not listen to it, but to the other trax. (to prevent over hearing it)
Only once the mix is more or less balanced, do you start with FX and only if they don't sit right or if you feel you have to tweak the faders.(compression).
The main thing to watch for during the process is frequencies that are similar because they will blend instead of standing out. (Pan, EQ)
I've often lost a vocal in a mix and found the best way to get it back is to TAKE OFF the FX because every FX pushes the track back into the mix.
IOW the best way to clarify the vox is to pull all the other trax back and then up the master.
I usually mix this way:
1.- Drums first. Make them peak @ -10dB, more or less. Adjust panning and stuff, compress together several drum and percussion tracks.
2.- Vocals. Once drums are OK on their own, bring up the vocals. EQ them and make them fill the center channel.
3.- Bass.
4.- Guitars.
5.- Keyboards.
6.- FX and special arrangements.
In every step from 3 to 6 I readjust the vocals once and again.
This usually gives me good results :)
Normally get the vocals and drums to what I think they should be, and then add 10% extra.
B
Quote from: na_th_an on August 27, 2012, 04:21:58 AMFor example, this song uses sidechain compression in the bass track (which was generated playing live a Nintendo DS with Korg software). Note how the kick "mutes" the bass track somewhat:
=========
You've hit on "ducking" when you feed an aux send of one track to the sidechain or key of a compressor on another track. When the first track has audio, it triggers the compressor to duck the level of the second track.
I was going to mention this as a possible way to bring a vocal up in a mix. You could check the mix and identify which tracks seem to be competing with the vocals the most, then apply some gentle ducking to bring those tracks down- only when the vocals are present. This would prevent having to alter the character of the competing tracks throughout the entire song.
Youze guiz is sharp, you beat me to it! ;D
What you are suggesting will work nice with heavy guitars in a loud hard rock track :)
I learned about that technique in a very very accidental way. Fifteen years ago when I used cheap and crappy casette tape recorders to record my demoes I learned to lean forwards and sing close to the tape deck's mic. That somehow lowered the guitar volume to make room for my vocals while I was singing, but left everything loud and heavy when I wasn't, so to say. I later understood that it was simply the deck's input compressor doing its job :)
Ah, the good ol' times!
what I have come to notice about vox or any track, is less is more, as per vol. I have, for a long time, been putting the levels up high in the mix especially since they were not clipping, but I wasn't realizing that all of the tracks together can collectively make clipping.
so I had to turn them all down more than I was normally doing, I get minimal clipping, and my recordings sound far less muddied (especially when a lot of tracks are involved) less distorted and sound clearer in placements. much of this I learned from reading on this site or was advised by people commenting on my music, and for that I am thankful!