Every now and then you get clients come in wanting to record to a two track because they can't get the stems for what ever reason.

I have heard engineers say in tutorials they can make it sound good with a lil bit of "voodoo" I have looked up every tutorial for this situation and recommendation on how to approach this that there is on YouTube and reading various forums on here and gearslutz.... Unfortunately there isn't a lot of info on mixing in this situation. How are they doing this?

I would guess you have to almost treat it like mastering since you don't have the stems right?

 

For the mixing side it is really a job like any other, should be very easy. Make a great vocal recording and then balance the vocal in well not too loud, not too quiet and everything intelligable, nothing sticking out annoying to the ear, a good frequency balance. Mixing a vocal to perfection can take a while, you may want to ensure every phrase is bang on, with great monitoring it will be easier of course.

Get the basics right, a good fundamental recording with the right choice of mic specific to the vocalist, dop a quick test recording for 2 or 3 likely mics, no distortion, no pops, no irritating sibilance, good eq to suit the vocal. You could try various types of EQ on the backing track to help the vocal sit well. For example you may wish to cut a little 3.5kHz from the mid channel to help the vox cut through or maybe it sits fine already, you just need to listen and apply your tool set.

There is no magic really, just your skill, good engineering practice, and application of a quality toolset to get it sounding great.

 

I am no mixing/mastering engineer, so take my words as mere idea...

I'd make vocals sound good first, and then start adding back volume of the instrumental backing,
voodoo I'd probably experiment with would be Trackspacer on instruments with sidechain feed from vocals, so it would gently push back music whenever vocal would kick in,
you shouldn't trouble yourself much with claritiy of vocal, but rather the overall feel of how things sit together

 

If i'm understanding the question you are wanting to make "mix" adjustments to a stereo 2 track due to lack of stems availability? I think you are definitely talking about mastering. The voodoo is masterful use of mid-side/parallel processing.

 

Here are a couple things that are easy to miss, but make a big difference.

1. Rhythmic coherence between the vocal and the backing track. No EQ setting can fix this, but if it's not right then no matter what EQ is applied, the vocal will never sit quite right with the backing track. The best way to hit this problem is head on, with Melodyne (where you're going to be melodyning the note onsets, not the actual pitches). Main thing we're hunting for are the stresses and phrase beginnings in the vocal, which will probably need to hit square on the "one". The other parts of the phrase can also be adjusted, but are less key since it sounds quite normal for the vocal to syncopate in these parts of the bar. Be creative with it and listen to the different effects you can generate by changing where the vocal is rhythmically (per phrase). Doing it like a robot and snapping all vocal onsets to the grid (or backing) does not sound that good, either.

2. Zynaptiq Unfilter, which is a blind realtime deconvolution tool. It can be used for two things in our context, we can somewhat remove unwanted shitty mic and room colorations from the vocal the client sent, which they probably made under less than ideal conditions. Get back a cleaner vocal, easier to mix with an already mastered backing track. Second application for Unfilter is to let it listen to the backing track, save the impulse response, and then apply that response carefully to the vocal track to some degree. This makes it sound more like the vocal and the backing were recorded in the same room with the same gear, more or less. It is not just EQ, it also "copies" some of the harmonic character over. Useful, and works best when applied onto an already "purified" vocal.

After such things are taken care of, use your ear and listen to how compressed the vocal is compared to the backing. A great disparity here (vocal uncompressed, backing heavily limited, for example) can sound odd and prevent the vocal from gelling with the rest. So you'd most likely be compressing the client's vocal track carefully as well, till it has a consistency/thickness that matches or complements the backing nicely.

 

Last edited: Dec 17, 2019

thank you all for your replies!