Log in

No account? Create an account
So I'm mixing and mastering Cex's new CD 'TALL DARK AND HANDCUFFED'… - an albuquerque not animate be armada. — LiveJournal [entries|archive|friends|userinfo]
Okrzyki, przyjaciel!

[ website | My Website ]
[ userinfo | livejournal userinfo ]
[ archive | journal archive ]

[Apr. 19th, 2002|10:09 am]
Okrzyki, przyjaciel!
So I'm mixing and mastering Cex's new CD 'TALL DARK AND HANDCUFFED' -- he sent me the instrumentals and the vocals as separate stereo files and i'm combining them. This feels like a momentous responsibility, not just because I want to do right by Rjyan but because I think the CD will have commercial potential, so it's my biggest mastering job yet.

So I've over-obsessed a bit on the setup, and ended up redoing it several times. I think I have it right but .... maybe over-analyzing it too much.... it's really a challenge to get that vocal/instrumental balance right.

[User Picture]From: angryrobot
2002-04-19 08:17 am (UTC)


Wanna go into any details on your process? Send some mp3s my way if you want another set of ears on the mixes......
(Reply) (Thread)
[User Picture]From: chaircrusher
2002-04-19 10:08 am (UTC)

My mixing techniques Re: so...

The basic technique I'm using is this: Rjyan sent me the vocals and the instrumental backing
as two stereo files. I import them into Cubase to line them up.

On the instrumental track, I put the Waves Linear Multiband compressor, set up to do some
low level expansion and compression on each band. Then I tweak the makeup gain on the bands to
scoop a few dB out of the 500-1500 hz range, to leave some room for the vocals. I go subtle on
this -- it doesn't sound hollowed out on the mids, but it's down enough in that range to let
the 'intelligibility' band of vocals ride a bit higher.

The vocals I run through the Linear Eq, De-esser, and Renaissance compressor. The linear EQ
cuts everything below 100 hz and has a mild presence bump between 500 and 1000. The De-esser
I try and tune to bring down sibilants just a bit -- Rjyan did a good job recording, but I'm
compressing the vocals pretty radically and without de-essing they get a bit splashy. Compression is fairly low threshold -- -25dB, fairly high ratio 4:1 or so, 10 msec attack, and
90msec release, with the 'automatic release' setting going. The idea is to have the compressor riding the level pretty much nonstop, which really helps bring up words in the rhymes that were a bit low. Overall the idea is to keep the vocals intelligible, and pumped up enough to stay
intelligible even when they're pushed down in the mix. I also have a Lexicon reverb on a
send, and push just a bit of the vocal through that -- 1.5 sec release, short predelay, hf
damped down at 1500 hz. The vocals are dry and this gives them just enough presence that
they won't sound like they were recorded in a closet -- which they were.

As much processing as I'm doing the result is still fairly subtle, which is the point. Rjyan
is a good producer, and did a decent job recording the vocals, which are the hard part. What
I'm aiming at is to not monkey with his mixes any more than I have to to get them built up
to 'professional' levels -- i.e. so that it won't sound wimpy after a Missy Elliot CD.

As for posting MP3s ... not anywhere public. I'll try and put some up privately where I can send yout the URL ...

(Reply) (Parent) (Thread)
[User Picture]From: optic
2002-04-19 10:32 am (UTC)

Re: My mixing techniques Re: so...

reading this is making me suspect i am nowhere near ready to try and offer my production services to anybody, which i was thinking about trying to do.
(Reply) (Parent) (Thread)
[User Picture]From: chaircrusher
2002-04-19 10:44 am (UTC)

Re: My mixing techniques Re: so...

Well, don't sell yourself short. I've heard your productions and they're tight. It's all about using your ears and having some idea of how to manipulate the tools. You know how to get a good tonal balance, and keep the various sounds loud enough relative to each other. Beyond that it's mostly using compression and limiting to bump up the average level. If you have good ears you'll
do fine, and anything you don't know you can learn by experimentation. You can also read one of those home/project studio 'how to' guides to get some starting points... but I think if you DO
read those books you'll mostly realize you already know what you need to know, but didn't have the lexicon to describe it.
(Reply) (Parent) (Thread)
[User Picture]From: optic
2002-04-22 10:34 am (UTC)

Re: My mixing techniques Re: so...

thanks. i think i do need to understand more about, like, "final mixdown" -- as in, what to do to make the final mix sound good on cd or record. levels, overall compression, etc.
(Reply) (Parent) (Thread)
[User Picture]From: chaircrusher
2002-04-22 10:53 am (UTC)

Re: My mixing techniques Re: so...

I've ripped a bunch of CDs and looked at the levels they use. The important number is the overall
average power. That's what Sound Forge and Wavelab will tell you. In pop music (and techno) there
is basically no dynamic range, with an overall RMS average somewhere between -11dB and 13dB. That's what your target has to be. You locate a section of your track that represents its loudest
level, get the RMS level, and use that as a guide of what to do with the Ultramaximizer.

The Ultramaximizer threshold determines the gain change through limiting. If you set the threshold down -6dB, the output is roughly 6dB louder (i.e. twice as loud). If your mix is well
balanced to begin with, all you really need to do is run the maximizer and let it go at that.

Two things can happen with this which will require some remediation -- the bass frequencies can be heavy enough to push the limiter to the point that the mids and highs get sucked out, and the high end can get really 'splashy' sounding and harsh.

The first you can fix with a multiband compressor. This reduces the dynamic range of each frequency range separately and then recombines them. You can adjust the bands' relative volume with the makeup gain control. You'd put this in the signal chain before the maximizer -- the maximizer always comes last.

The multiband compressor can be used to get rid of some of the splash, by putting a band between 3000 and 6000 hz, and lowering the threshold until this band gets significant gain reduction. But what seems to help smooth things out with much less tweaking is a de-esser. The de-esser should come before any EQ or compression (or maximizing), as it changes the signal level by doing frequency dependent compression. It works globally -- i.e. it compresses the entire signal, but the compressor is triggered just by that SSS/SHHH band, so that when those frequencies are loud the overall gain cuts.

De-essers can really smooth out the sound, but like anything else in mastering, it pays to use it subtly.
(Reply) (Parent) (Thread)