How to mix a vocal with a mastered beat?

[steveschizoid]

I'm not very experienced with rap and hip-hop conventions, but I had two clients today(!) who brought in beats and wanted to record vocals. All of the beats seem like they have been limited. I guess the only way to deal with that is to turn down the beats, mix the vocals and limit again, right?

Here's the problem. One guy brought a part of a song which (he tells me) was done by a national act for a mix tape (whatever that is), and they wanted him to record a part and send it back. They provided a snippet - not even the full track - and they want the new vocal mixed in with that snippet (which I can only assume will be spliced in with the rest of the track). I'm told it was done this way to prevent piracy.

My questions are 1. what is customarily done in this situation? and 2. Is there any way to do this which will allow a seamless splice?

[Gobblebox]

Being a rapper this seems like par for the course. Outlandish. I love rap. So ridiculous. Haha.

[soulofmalay]

You are exactly right, working with a "2 track" that's already mastered, you have to mix the vocal in and put a slight limit on your overall mix.

With the snippet situation...if they end up actually using the verse (the national act) then they will take your session and import the vocals into the real session (with everything tracked out).

If it's situation where you need to combine a small snippet into part of another 2 track mix.... Calculate the BPM of the track and then simply cut/trim and place the snippet where you need to go

Hope that helps

Malay

[steveschizoid]

Originally Posted by soulofmalay

With the snippet situation...if they end up actually using the verse (the national act) then they will take your session and import the vocals into the real session (with everything tracked out).

Of course this is how it should be done, but, according to my client (and apparently there is no way for me to communicate with anyone else involved) they want the new vocal mixed with the two minute snippet - which they will splice with the existing track. If I've lowered the beat's level, mixed the new vocal, then remastered the snippet, it's not going to splice that well. It really makes no sense to me at all, but the client is a nice, talented kid, and I want to do the best I can to help him succeed.

[soulofmalay]

Originally Posted by steveschizoid

Of course this is how it should be done, but, according to my client (and apparently there is no way for me to communicate with anyone else involved) they want the new vocal mixed with the two minute snippet. It really makes no sense to me at all, but the client is a nice, talented kid, and I want to do the best I can to help him succeed.

Ah I see, well you are right on point again, just mix the vocals in with the snippet and if they aren't supplying you with the rest of the song (even a 2 track) then they will have to splice it together. Unfortunately this practice occurs all the time, especially on "mixtapes" which now days have become compilations of (usually) unlicensed songs that are used more for artist promotion than for actually selling.

However, I've actually been a part of a session where similar tactics were used on a major label song then ended up charting in the top ten because the producer (dude who made the beat) lost the original files off his keyboard and had no way to track it out... so they spliced his minute long loop (2 track) and mixed the vocals into it. Horrible horrible, but hey, it sold! (as I sigh in frustration)

Peace,
Malay

[bobsandifer]

I do tons of this type of stuff and even though they seem to know exactly what they want there might be some confusion in the translation.

1. Find the tempo of the track so you can lock that in.
2. Put the section you need to record at measure 5 but make sure you have the track matched to the beat. The 1 is on the 1 and the 2 is on the 2 etc.
3. Record your vocals plus whatever stacks, emphasis and adlibs that need to be added.
4. Export all of those tracks as waves from measure 1 to the end like you would a stem.
5. Mix the track as you see fit and bounce the final work

Send both versions to the "producer". If you do it this way then you will have ALL problems covered. I have had TONS of repeat business by correcting an issue before it actually comes up.

For the record......I really dislike doing this stuff but I take what I can get.
If the instructions are not clear I can make a short video of the technique.

EDIT.......freakin YEAH !~!~!~! My 400th post !

[steveschizoid]

Thanks for the suggestions guys. Here's what I did: I reduced the beat about 3 dB. I had tracked the vocal with compression, so I compressed again, then stuck a UAD maximizer on the buss and volume automated. I also had a maximizer on the 2 buss (instead of limiter because the maximizer has a mix control) chopping off around 2 dB. When the vocal was over I simultaneously faded up the beat and faded down the mix control of the maximizer. That left about 5 more seconds where whoever has to deal with this absurdity can splice with basically unprocessed material.

It's certainly not ideal; I really don't like the sound of all that limiting. I may try to volume automate to avoid the maximizer on the vocal track, but, given the rapid fire nature of the rap, it could be problematic.

I will also urge the client to send a vocal stem; I'm pretty sure he won't do it.

edit: one final inovation, I automated a multiband compressor to step on the periodic, very loud interjections of the words "L-Trek" into the space reserved for my client's performance, which made him more audible and decreased the limiting distortion. I can only guess this is the original artist's name? Anyone heard of him?

[rob101]

good insight there

[kheftel]

When you had to mix the vocal into an already-limited 2-track background, did you add any reverb to the tracks, or just EQ, compression and the like?

[steveschizoid]

I added a bit of a room from my m7 because I use a very dead, very treated room for recording vocals.

[DaKid]

Originally Posted by kheftel

When you had to mix the vocal into an already-limited 2-track background, did you add any reverb to the tracks, or just EQ, compression and the like?

In these cases, you have to do whatever you have to do to get the vocal to sit with the track. I find that a slight delay and slight reverb can help sink the vox into the 2 track.... As previously mentioned, make sure you find out the tempo of the beat... It makes copying and pasting (lets say the hook) a whole helluva lot easier. Sometimes EQ'ing needs to happen on the track to cut a spot out for the vocal.

Luckily the last few times I've run into this scenario, the producer/beat maker who gave the artist the 2 track had the beat mixed down very well and getting a good final product hasn't been a hassle. It's when the beat is poorly mixed where you have to do some serious surgery to make things work.

[beanface]

Originally Posted by steveschizoid

I'm not very experienced with rap and hip-hop conventions, but I had two clients today(!) who brought in beats and wanted to record vocals. All of the beats seem like they have been limited.

Why would anyone master the beats before the vox went in?!

Thats a major schoolboy error.

[tmrstudio]

Originally Posted by DaKid

In these cases, you have to do whatever you have to do to get the vocal to sit with the track. Luckily the last few times I've run into this scenario, the producer/beat maker who gave the artist the 2 track had the beat mixed down very well and getting a good final product hasn't been a hassle. It's when the beat is poorly mixed where you have to do some serious surgery to make things work.

I definately don't care for working with a 2 track. Often times, it sounds like crap(over compressed and/or noisy) and the client doesn't understand why it can't be fixed

[kheftel]

Originally Posted by DaKid

In these cases, you have to do whatever you have to do to get the vocal to sit with the track. I find that a slight delay and slight reverb can help sink the vox into the 2 track.... As previously mentioned, make sure you find out the tempo of the beat... It makes copying and pasting (lets say the hook) a whole helluva lot easier. Sometimes EQ'ing needs to happen on the track to cut a spot out for the vocal.

What kind of a slight delay are you talking about? Like less than 50 ms?

[kheftel]

Originally Posted by beanface

Why would anyone master the beats before the vox went in?!

Thats a major schoolboy error.

So true, but I have this one client who likes to rap to beats he gets from third-party sites online... and the beats he buys from them usually come mastered so loud that there isn't any room for the vox... I guess they do it to make their beats sound good to buyers, especially over the Web, but it annoys the crap out of me. I end up lowering the gain on the 2-track background, mixing in the vocals, and then cranking it back up - obviously losing quality in the process, but what do you do? Anyone have a better idea?

[DaKid]

Originally Posted by tmrstudio

I definately don't care for working with a 2 track. Often times, it sounds like crap(over compressed and/or noisy) and the client doesn't understand why it can't be fixed

It's definitely not my favorite method of mixing... but sometimes you gotta do what you gotta do.

I'll usually try to explain to them why it's better to have the beat tracked out and play them a mix that I've done with everything tracked out... They usually get it by then.

[DaKid]

Originally Posted by kheftel

What kind of a slight delay are you talking about? Like less than 50 ms?

Just find out the BPM of the song so that your delays are on time. Do a stereo delay and play with a 16th note delay on one side (L) and a 32nd on the other (R). By slight- I mean sometimes I'll bring the level of the delay up so that it isn't audible when you're listening, but if you disable the delay you notice the difference.

Don't go too deep into it. I can give you exact numbers and milliseconds all week, but you won't get it til you load up a session and twist some knobs. At the end of the day, it doesn't matter what you do as long as the final product sounds good.

[kheftel]

Originally Posted by DaKid

Just find out the BPM of the song so that your delays are on time. Do a stereo delay and play with a 16th note delay on one side (L) and a 32nd on the other (R). By slight- I mean sometimes I'll bring the level of the delay up so that it isn't audible when you're listening, but if you disable the delay you notice the difference.

Don't go too deep into it. I can give you exact numbers and milliseconds all week, but you won't get it til you load up a session and twist some knobs. At the end of the day, it doesn't matter what you do as long as the final product sounds good.

Ooh, delays timed with the tempo - got it. I just couldn't tell if you were talking that kind of a delay or one that would simulate early reflections. Btw thanks for the tip about carving a space in the 2-track for the vocal with eq - used it today on a mix, with great success.

[u b k]

One thing you can play around with to get a vocal sitting in a leveled 2-mix with is keyed, multiband ducking.

Split the 2-mix into 3 bands, one running mostly lows and low-mids, another the mids, the 3rd the top. Only you can tell where exactly to make the splits. Keep your crossovers wide to minimize phase hash.

Set up a comp on the mid band and key it off your primary vocal, do very light medium release ducking, 1db should be more than enough. This way the bass energy and the treble stay consistent but the midrange breathes a little in response to the vocal, which gives you more room to tuck it without getting swallowed.

This is one of those situations where software makes your life a lot easier, but it's a total tradeoff because software actually enabled this mess in the first place.


Gregory Scott - ubk
.

[DaKid]

Originally Posted by kheftel

Ooh, delays timed with the tempo - got it. I just couldn't tell if you were talking that kind of a delay or one that would simulate early reflections. Btw thanks for the tip about carving a space in the 2-track for the vocal with eq - used it today on a mix, with great success.

No problem!

[scius]

damn

Originally Posted by u b k

One thing you can play around with to get a vocal sitting in a leveled 2-mix with is keyed, multiband ducking.

Split the 2-mix into 3 bands, one running mostly lows and low-mids, another the mids, the 3rd the top. Only you can tell where exactly to make the splits. Keep your crossovers wide to minimize phase hash.

Set up a comp on the mid band and key it off your primary vocal, do very light medium release ducking, 1db should be more than enough. This way the bass energy and the treble stay consistent but the midrange breathes a little in response to the vocal, which gives you more room to tuck it without getting swallowed.

This is one of those situations where software makes your life a lot easier, but it's a total tradeoff because software actually enabled this mess in the first place.


Gregory Scott - ubk
.

[steveschizoid]

Originally Posted by u b k

One thing you can play around with to get a vocal sitting in a leveled 2-mix with is keyed, multiband ducking.

This is one of those situations where software makes your life a lot easier, but it's a total tradeoff because software actually enabled this mess in the first place.

thumbsup on all counts. I have been experimenting with Izotope Alloy for this purpose; it's the only plug in I am aware of that has side chainable multiband.

This trading of crushed beat mp3's seems to be common practice in hip hop. I actually had another hip hop project come in recently (subsequent to the one that provoked my original post), and it's actually pretty cool, but the guys involved seem utterly uninterested in attempting to have the person who created the beats ftp me the wav files. I have to wonder why.

[u b k]

Originally Posted by steveschizoid

...the guys involved seem utterly uninterested in attempting to have the person who created the beats ftp me the wav files. I have to wonder why.

Do you?

If it walks like a duck...


Gregory Scott - ubk
.

[Darm]

u b k
Sorry, I could not understand what you were saying about the "wide" x over points.
As far as I got it you reccomend spreading the 2track instrumental to three tracks, So i Would create three traks, and drop the same .wav file everywhere, with filters on each track, by saying wide you mean using steeper filters? like 48db/Oct?

[E.rOk.stA]

Unfortunately, this is the majority of my business. It sucks but is an art in it's own respect.

Normally, when mixing, you would EQ the instruments and drums to fit the vocal but in this case, I tend to make the exeption and notch EQ the voice in places as well. Just little notches here and there based on what the focal points of the beat are. It also seems that using compressor plugs with an 1176'ish forward push (or waves SSL) really help the vox to fit into the the already limited beat but stay clear and audible.

As far as delay, I 2nd the use of it lightly. Often the Waves Super-Tap does it with minimal fiddling. Just use the ping-pong preset, tap your tempo, LPF around 1-4kHz and it's almost always a winner. By lightly I mean keep your send fader at between -8 to -12 dB. It will sound like alot solo'd but in the mix, you'll barely sense it. I often use more during a hook or hype track (all the "yeeah's" and "c'mon's" in the background).

Not in the OP's case where they are gonna splice this into an already done song, but in the case where you wanna bring a limited beat back to life, lower the fader to -8 or so and try using an expander to add dynamic range. Also, flux's free plug "Bittersweet" and SPL's "Attack" are good for this. Be careful though. It is way too easy to trash something like this too.

[u b k]

Originally Posted by Darm

by saying wide you mean using steeper filters? like 48db/Oct?

Sorry for being unclear, I mean the opposite: shallow filters, so that your overlaps are wide.


Gregory Scott - ubk
.

[phillysoulman]

Who ya gonna call??
The Master Beaters..thats who.

[3rd Degree]

I have only run into this problem a few times as most people I record come to rap over my beats. What I do is simple and fairly effective but doesn't compare to a well mixed beat that has not been processed further than that.

I basically "de-eq" the beat before, or during the tracking stage, depending on how much time the artist has. If the artist has time for me to work, I tend to just cut anything that is conflicting to where it the voice is more audiable so the rapper can better vibe off the music and doesn't "fight" with the beat in delivery. It may seem subtle but many rappers will get louder, more aggressive, or both, non intentionally.

After tracking I start the process of "mixing". I will low cut the vocals first, almost always. I focus on trying to get one clean verse fitting to beat. I find having a thinner, clearer verse to be what works best. I will try to deal with any overdubs/adlibs and keep them low and process them less than I normally would. Then I start the process of slowly undoing the eq on the beat where I cut and re-eqing the vocals in the process, trying to find a good trade off. I find filtering and eq to be the only tools that do much for me at this point, and the fact I am basically sucking some life out of the verse.

I like to do this infront of the artist when possible, atleast for a rough. They see me struggle much more than when giving me a well mixed beat or session file, and end up with "mixtape" quality. They tend to be either more selective in the future, or just go through me. Remember, rappers who go to other people's studios, weather it be a home studio or professional, usually can't do this themselves. Let the know where they went wrong in terms they understand, skip the jargon, do your best, and hopefully they begin to get that their skill on the mic isn't all they need to get a good product.

[lazarus]

I have had a few dudes come in with garbage MP3's wanting me to make them sound like stars. Must be a Utah thing. His name isn't Dre is it?