Music Articles From

Back to articles index | Home |

Secrets of Mixing - By John Vestman
Rule #1. There are no rules. Be creative. Create a major "Wow!" factor. Create something unique with a twist. Ask yourself, "What is different about our music?" Imagine the unimaginable and bring it into being.

Book a pre-mix clean up session. Take an hour or so to come in with the engineer, and erase all the throat-clearing, the guitar clicks, the out-take solos, etc. You'll feel fresher at mix time, because you can concentrate on the creativity, not the housekeeping.

Allow for more time than you think you need to mix. There's nothing worse than thinking it will take 3 hours to mix a song, and it ends up taking 5. You're under pressure, the engineer's under pressure, and the studio's next client is pacing back and forth in the waiting room. Have extra money (if you're paying the studio) on hand, so that if you go over budget (always the case) you aren't sweating bullets.

LEAVE THAT COMPRESSOR OFF THE STEREO BUSS! Ok, every Mix Magazine and Keyboard Player and Home Studio Addict magazine has got all the cool ads for high-end compressors. Ah! THAT'S how to get my home studio to sound like Ocean Way! Nope. Save your money on that high-end compressor, or just use it on your individual tracks that need compression (particularly bass and vocals). Mastering is the best place for more strategic compression and limiting. Overly compressed mixes box the mastering engineer into a corner, reduce the openness of the mix, and lower the number of enhancement options.

If one mix is very compressed, and several others are not, guess what? I have to jump through hoops to make things sound consistent. In most cases, when I have the artist or producer with me, we end up wishing that the mix wasn't compressed. When I master, sometimes I compress, sometimes I limit, sometimes I do both, sometimes I level-correct in the Sonic, and sometimes I leave well enough alone! I can bring more musical consistency to your cd if you give me a bigger range of options in this area.

Mix "alternate" mixes. After you get the mix just the way you like it, mix another version with the vocal up 3/4 dB. Mix another version with the kic up 1dB (a good option to have). Mix another with the guitars or the backup vocals up 1/2 dB - you get the idea. Whenever you have doubts about a certain instrument or vocal, make an alternate mix and then let the final choice be made in mastering. This is becoming more of a common practice due to the "hot cd" game. The hotter we're being asked to master, the more the mix will change slightly as everything level's out toward the top. That's when you very well might want that kic hotter than you would have liked it in the studio - but now it can be heard in the mastering room - where all that should count is going to count.

Flash! Cd mastering levelling practices have changed and we've been engaged in a level contest to see who can cut the hottest cd. So given that CDs have become so loud, and many prefer volume over more natural (and lasting) dynamics, there is another element to mixing that can be helpful - I'm going to call it "pre-limiting." What you do is start your mix in the normal way, get it to sound great without compressing the stereo buss - the CDR copies you make should sound good musically, but they won't sound as loud as newer commercial CDs. That's OK! Don't worry about the volume right now. Keep this version as a master mix.

Next go at your mix again but insert a limiter (not a compressor) over the stereo buss (limiters are fast, compressors are slow). Crank the output of the limiter so you can now make a hot CDR copy, getting closer to the level of newer CDs. This is just to pre-test how the kic, snare, vocals and instruments start to blend when the tops of the peaks are cut off, which is required to make a hot CD. Listen to the hotter CDRs again to see if there's enough kic punching through the mix. You may have to bring it up... more than you expect.... Get this version to where you like it, and keep it as another master mix.

Now once you are hearing the kic more like you did on the original non-limited masters, go back and remove the limiter and keep this as a third master. Note: By removing the limiter, you may have to bring down the overall mix level, but that's ok. Keep the overload lights OFF. Digital clipping (on any system) is not your friend. Now when you submit your mixes for mastering, include all the mixes: Normal dynamics, limited dynamics, and non-limited 3rd mix (exaggerated kic, trimmed bass and whatever other changes).

This gives you more options at mastering time. First you have a regular mix that sounded the way you liked it in the first place. Second, you have a mix that's geared for hot CD levels with the needed compensation for the limiting that packs everything into a fatter package. Third you have a non-limited but compensated mix that could possibly be the one used for mastering... but... you're using the concept of taking out that limiter so that a more precise level of limiting can be chosen at mastering time.

This whole technique really wasn't needed back in the mid-'90's because the labels and major artists weren't pressing the volume so far beyond normal... as they are doing now. We are really not hearing pop music in a normal dynamic context anymore... and most mastering engineers aren't very happy about it. But... we're here to please our customers, so if level is what you want, level is what you'll get.

Gads.... back to mixing....
Bring in a few commercial cds with you to the mix. But first, know your market. What radio station would play your music? What are the CDs they play often? Which music sounds good over the air? Who's drum sound do you like? Who's vocal, guitar, string, piano sound do you like? Your idea of a big sound may be different from your engineer's, so if you bring in a cd, hand it to him, and say, "Check out cut 5 for the vocal sound." he/she knows exactly what you like. "Put in this other cd and listen to the guitars." You get the idea. For every 4 hours of mixing time you spend, ONE of those hours (spread out during the four) should be listening to other cds when you begin using this technique. This is also cool at tracking time. You have an abundance of material in the form of commercial cds to get ideas from.

NOW HERE'S THE CATCH - CDs have been mastered! I recommend that you audition some older cds as well as newer ones. Why? Because the older cds haven't been compressed and limited as much as the newer ones have, and this gives you a truer sense of the dynamics you hear in a mix. Your goal is to compare the sounds and tones of the cds as a guide during your mix. For instance, you may be used to hearing the bottom end of group "A" sounding great at home. You may have heard the guitars of group "B" on the radio sounding rad. But since the studio monitoring environment is different, your impression won't necessarily translate into the mixing room exactly the same.

The characteristics of the studio speakers will be different. So when you bring in those cds, you now hear what impressed you in the real world right there in the studio next to your mix. If your mix doesn't impress you as much when you first A-B to a cd, don't rag on your engineer! It's just a process, and being diplomatic will save you time and increase the creative flow. Just say, "I like a lot of what we have now, and I'd like to get a little more of [fill in the blank]. I'd like to listen to these to get some ideas." Be sure to check out my page on commercial CD references, and see Studio Monitor Madness for more info about the actual speaker system and it's effects on mixing.

Quick tip: Keep any paper labels off your master CDRs - they inhibit the rotational balance and can cause the player's error correction to work harder. Tip 2: Only write on the top of CDRs with a soft felt-tip pen prior to burning the CDR, not after. The top is more fragile than the bottom! ...and here's even more mixing tips on bass/drums/vocals/de-essing, and some great EQ and compression suggestions but in the meantime...

HEY...MIX TO ANALOG TAPE! The vast majority of projects do not need the hiss-less format of digital, and the bottom is so much better on analog! There is just a "hole" that is hard to describe in digital audio. For some reason, the extra thump that analog has (or holds onto) is just great. There are an accumulation of budget components in some consoles and digital machines that adds to the problem. I highly recommend BASF tape. Ampex (or Quantegy as it's now called) is not as good sounding. 456 is very cloudy sounding tape, and a great way to lose some transparency right off the bat. While 499 is better sounding than 456, it still isn't as clean or live or transparent as BASF 900. BASF is physically cut better, has less print-through, and is just better sounding. An archivist I know has said that some 3 year old 499 is starting to get sticky, just as old 456 did. Quantegy GP9 is really the old formulation of 3M's 250, which was a nice sounding tape (it's not intended for +9 elevation), but my vote's still for BASF.

I don't recommend elevating your level above +6dB. Why? Because of a lot of things. Marketing hype has made the overload capabilities of modern tapes overrated. There's a lot to consider about the plus' and minus' of tape saturation vs. signal-to-noise vs. print-through, etc. Take print-through for instance: Tape machine heads pick up magnetic signal, and the stronger the signal (louder you've elevated the tape) the easier it is for the adjacent tracks to pick up what's recorded. Result: more crosstalk, especially from 500 hz down. That means that all the low end will bleed slightly from track to track to track. At +9, track 5 "hears" more of track 4 & 6 than if you elevate to +5. All that low bleed makes for mush in your mix. You'll have no hiss, but the bottom will be tubby and slow sounding.

Trick: If you don't mind breaking the rules, align your machine so that you set 1K at -2 (using an NAB 250 nW/M alignment tape) and 10K at -3. That way you have to elevate the high end more. The tape can handle the extra high end level, and it doesn't mush up the bottom. It's not enough to saturate the highs, and it's not dangerous enough that if the tape goes to another studio people will faint. Think of this trick as a broad-range, simple form of noise reduction (which is the whole goal of tape elevation, anyway!) Now you get the hiss reduction of a +6 master with the clean bottom of a +5 master! Voila! (Or just use IEC (CCIR) equalization instead of NAB. It's a standard, and it's reproducible and accomplishes the same noise reduction effect.)

Ok, so you don't want to use analog.... the next best thing is a Masterlink at 96k or 88.2k 24 bit, or a 24 bit AIFF (WAV is ok too) file - the higher the sampling rate the better (and remember to stay a couple dB under clipping).
A couple more things about level and then we move on... Competing for level is an old trick that dates back to vinyl, but with vinyl, there was a different reason for cutting a hotter lacquer. Since vinyl inherently had surface noise to it, the hotter the sound (and therefore the wider and deeper the grooves), the less you'd hear the surface noise. Also, if the song come on strong, level-wise, it seems more exciting right out of the gate. (You never get a second chance to make a first impression, right?) Vinyl is an analog medium, and it is a flexible medium, in that there is an acceptable range where the signal can be increased depending on the dynamics of the music.

But with digital, there is NO fudge factor. A one is a one and a zero is a zero. There's no one-and-a-third. So when the level hits digital "zero-VU", that's IT. There are no more numbers. So we mastering engineers, when presented with the hotter and hotter cds of our associates (I don't use the word competitors - there's plenty to go around), we've used our expertise to compress, limit, de-ess, level-correct, and multiband compress the sound right up to the max. What we have to control is two things: (1) The RMS level, or average volume and (2) The Peak levels, those things that can light up the 'digital over' light like crazy.

With analog tape, we engineers used to be aware of the tape hiss, and adjust our levels upward to eliminate that hiss, keeping our eyes on how much headroom we had to work with on the tape, so as to avoid (or utilize) tape distortion. With cds, it's the opposite. We watch the very top peak of the song, and gear everything around how close to digital zero we're going to get with that peak. How much RMS level ends up on the cd depends on how we manage the peaks of the music.

Now I'm perfectly happy cutting a loud cd for you. Cool. Just know that the problem is that all the transients take on a different shape and sound when we do this. For instance, many musicians like punch. Well, think about it. The punch you feel from the bottom or mid-bottom comes from the speaker excursion. The cone moves forward a certain amount and then moves back, and so forth. When we limit/compress the peaks, we are able to bring up the body of the music (the non-peak stuff) higher. That's what gives you that louder, RMS level on a cd. BUT THE RELATIVE DISTANCE THAT THE SPEAKER MOVES IS LESS. That means that the over-all sound is louder, but since the speaker doesn't push the sound wave forward as far, there is less impact from the movement of the air. (Unless you turn it up to glass-shattering levels, in which case the sheer intensity creates the impact.)

Give yourself some slack at first. Group "C" may have had a $50,000.00 budget for their mix alone. Mix so that when you push the cd-player-button, they sound great, and when you push the stereo buss button, YOU sound great too, in the context of your music and the tools you have to work with.

And remember - have fun! Stay fresh, take breaks, go look at girls (or guys ...ya know, whatever).... take vitamins...

John Vestman is a veteran mastering engineer with over 26 years in the industry. His credits include: Hole (Courtney Love), Juice Newton, Ambrosia, Andre Crouch, The Wynans, Great White, Candyman, Billy Davis Jr./Marilyn McCoo and more. John Vestman Mastering is located in Orange County, California, and his web site offers over 40 pages of information about successful studio recording techniques and sound philosophy.

Back to articles index

Copyright © 2001 Galaris LLC. All rights reserved.