Friday, July 29, 2011

Recordingtip 28 Be cautious using brickwall limiters like the L2. If your mixes resemble a brickwall, back it off

The advent of digital recording brought in a whole new playground of toys with which to play for mixers. And in my view with all the myriad of processors now at our disposal, nothing has really changed the sonic landscape as much as the so-called “Brickwall” limiter. I remember the first time I loaded a modern, mastered CD into Protools. It was a project I’d mixed in an analog console and mastered by Bob Ludwig. The move to “in-the-box mixing had yet to really happen, though most of us were recording in a digital DAW of one kind or another, most mixing was still analog. So I loaded it in. And there it was...the waveforms cut like a finely cut hedgerow garden in England. ALL the transients lopped off and looked like nothing I’d ever seen before.

Of course Ludwig is one of the most respected mastering engineers in the business and it sounded great. But I was amazed at what I saw and wondered if such an approach made sense on all types of music.

It’s been a good many years since that day. And since then mixing in the box isn’t that uncommon at all. In fact, almost every mixer uses some form of it and with that comes the opportunity to put an L2 or any other limiter which can smash ever single transient to smithereens. The trick is knowing how much to apply.

I’m fortunate that most every mix I make is mastered by a really great mastering engineer. So all I really need to worry about is making a great mix and letting the mastering engineer do their magic. The caveat of this is most labels judge mixes before mastering. So there’s the temptation to put a hard limiter on the mix so it gives some semblance of what it will be when mastered in order to please the label gang.

This is especially a problem when an album is mixed by several different mixers, such as the last album I did. One would think that professionals at the label would understand that mastering will even out the levels of the different mixes and apply their processing to them. Unfortunately that’s often not the case. And if you give your mix some dynamic range and the other guy smashes his to death, it’s often that mix, sounding louder and more aggressive, may make your mixes seem not as good. You’d especially think that professional folks at major labels know and understand all this, but it’s rare when they do.

This reality puts many engineers in a bit of a quandary. Do we do what we think is right, and know the mix will sound good once it’s mastered and hope the folks at the label realize this? Or do we make a “label” mix, squashing it a bit for a sense of what it will be like when mastered, and turn into mastering the version we want? I can see that as a good option, provided that mastering makes it better. But what if your version of what mastering will be isn’t that good? After all, a good mastering engineer will probably do a much better version of that than you ever will. You could in the end not helping your cause.

One of my mentors, Jack Joseph Puig commented in a recent Waves webinar that he prefers to get his mixes as close to a mastered version as possible. So he’ll go ahead and hit the limiter as hard as he feels right for the final result. But let’s face it...very few of us are JJP. Jack is a very serious student of many things, including how his mixes end up when mastered by Bob Ludwig and the like. So I give Jack full reign with such things. But my suggestion to all those learning the ropes is to remember that JJP has many years of work and is much better able to know how much is right, and how much isn’t right. Remember, compression is the most permanent thing you can do. It’s better to leave some room for your mastering engineer.

So while I why understand why mixers want to compress their mixes hard, more often than not I hear mastering engineers complain that they have nothing to work with when they get mixes squashed to death. I heard a reaction to one such mix from one of the guys at Sterling as, “All I can do is turn it down!” That’s not the response of a guy needing to validate his reason for being there. This is a guy who’d be happy to have a great mix show up and he do nothing but a straight transfer. His is the voice of frustration of the loudness wars and that he can’t make an album feel and flow the way he hears it in his head were he given a bit of room to do something with it.

Now while I almost always use an L2 or some sort of limiter as a final stage on the stereo buss, it's purpose is to capture those peaks that a normal compressor misses. I’ll rarely have it do much more than a few dB of limiting. I have it there to catch an irregular drum hit or something like that. I'm not saying you should never use one. Just be careful in how it’s used.

One other thing. Before I begin a mix I ask who will be mastering so I have some idea of what will happen once I turn my mixes in. Knowing their particular mastering style helps me know what I need to do. And more often than not, I’ll send a mix to the mastering house for some helpful comments about my mix. If I can, I’ll even go down there and hear my mixes before I’m done. This is a great thing to do and virtually ever mastering engineer I’ve worked with not only allows this, but encourages it. Just give them a call, tell them who you are and that they are mastering something you're mixing, and would they mind if you dropped by at a time convenient for them to hear it. Sometimes the mastering house may not be in the city you live in. In that case, ask if I can upload a mix to them and then have a phone conversation about it. That works too. You’ll be surprised at the things you learn by doing this, and your project will be better for it.

That being said, I realize that many of you won’t have the opportunity to have your mixes mastered by the likes of Bob Ludwig, Andrew Mendelson or Doug Sax. If that’s the case what is the best thing to do?

The best approach is still to not overdo limiting when mixing. This is even more important if you don’t have a A level mastering engineer. At least the top level guys can probably handle over-compressed and limited mixes better than the others. One of the great secrets of a great mastering engineer is knowing what NOT to do. A lesser experienced mastering engineer may feel it's their calling in life to limit the hell out of every mix, thinking that is good.

Which brings me to my final point. As much as possible, attend your mastering sessions. You will learn a lot. Back in the old days of Abbey Road, young engineers began in the mastering room so they could learn what things were like when mastering good and bad mixes. I've learned as much at mastering about my mixes as I have anywhere else. And you just might save your mix from sonic annihilation!

While I understand the desire to have a really great sounding and loud CD, if you whack it too hard it'll probably sound not nearly as good as you’d hoped.

And remember, any fool can hit the limiter hard. But only then great ones know when to stop. You’ll have one more stage at mastering. Give some room to do the crush then.


  1. I think the common goal that mix engineers share is to get the best possible sound. The best possible sound is NOT the loudest mix. I think that mix engineers should focus on how instruments interact with each other, being creative about effects, plug-ins and use dynamic processors at each channel when required, panning, etc. If tighten dynamics at the master bus is a must, then use it with gentle settings to let Mastering Engineers do what they do best.

    When analog recording was the only way to go, the need to record at hot levels had two main purposes: to get tape saturation harmonics over some instruments that sounded better with it (bass, drums, etc), and to keep tape noise as low as possible -compared to the recorded signal level-.

    Early days of digital recording, the common way to go was recording/mixing/mastering at very low levels due to several reasons but mostly because industry didn´t know exactly the consequence of going over a relative point. As time passed by, the first time I heard a justification on getting hotter recording levels in digital domain was that “quantization noise” was annoying, somewhat a valid point those days. But right now, with all the technology that most of people can afford, 24-bit recordings that bring more headroom, state-of-the-art A/D and D/A converters or even those you find on a pro-tools interface are far from introduce quantization noise like the early days, so I don´t see a point to record or mix that high.

    I can understand the pressure some mix engineers should feel when they work for an artist and the label is asking for hotter mixes, but they should keep in mind that one of the beauties of music is to listen to it and feel the different dynamics that blends together into a song. Hear the details, even on “heavy style” music. Don´t ruin it just because it’s a trend to use extreme compression/limiting. Cheers!

  2. First of all I'd like to thank David for this forum where we can post our thoughts and interact. Sharing stuff that we probably learned in the College of "hard Knocks" rather than from any professor. that being said I'd like to add my 2 cents in when it comes to limiters. My knowledge of them could be somewhat out of date since I was most active in the early days of digital recording. Heck I even remember Stevie wonder coming into Sigma Sound with a Sony PCM-1000 A/D converter that recorded to Video Tape. Primitive , Huh? But, I'm off on a tangent, back to limiters. To tell you the truth, I've always hated them!!! Unless the settings are exactly right it's been my to my horror that when I went back to a fresh listen, that they would cause all kind of problems. Unwanted artifacts that were awful especially the "Ping-Pong effect.
    In other words when a loud sound or transient would trigger the limiter you could actually hear the "Choke" coming in and then returning to normal. This created a very nasty type of coloration which ,in my case , would cause listener fatigue. What you would get would be comparable to someone constantly turning a volume control up and down. Therefore I avoided them like the plaque. I would take a level of the loudest sound , IE; drummer, could prodice and use that as my "redline." That was the early days, people and since then a friend has demonstrated the Omni processor to me. It's totally controllable, and unlike the early days can be set by software to do almost anything. But, and i do mean this, it is not a set it and forget it type of equipment either. You have to make and take the time to make sure it's set correctly. And if I were to use it I would have two tracks going with it on one and the other without. Well, as i said, this is a viewpoint from my days as an asst engineer. Now that I'm not actually playing "on the road" fulltime I'll be going vback into the studio to see and catch up on lots of the new stuff.

  3. Great points Juan and Larry and thanks for the comments.

    Yes Juan, we don't have the analog noise floor these days. And the bit rate is good enough for almost any pop music.

    And Larry, thanks for the TSOP references. Great to hear that