home | news | articles | about | contact | hire me | fm inc.

Having trouble hearing compression? Turn your monitors down.

The speed at which a signal comes and goes as the compressor grabs and releases the program material is controlled by the processor’s attack and release controls. While a compressor’s envelope can be used to shape a signal in creative ways (e.g. make a snare drum punchy or a vocal extra breathy) it can also work against us if used improperly. When first learning to use compressors and limiters, it can be difficult to hear the compressor’s envelope, or attack and release settings, making them difficult if not impossible to set correctly by ear. While there is a certain amount of practice and experience involved with hearing the subtleties of attack and release, especially at low levels of gain reduction, one thing that I find that helps me hear these parameters working is to monitor at lower listening levels.

Our monitoring equipment (speakers and amplifiers), as well as our human auditory system, begin to naturally compress material at very high SPLs (very loud listening levels). While monitoring your mix at extreme levels is not a great idea for a variety of reasons, namely fatigue and hearing loss, it is especially difficult to hear subtle envelopes in your compressor’s attack and release settings when so much gain is competing for your speaker’s (and ear’s) attention. In other words, if your speakers are so loud that your ears and amplifiers are also adding compression to the signal, how can you expect to make critical decisions about compression?

This is not to be confused with discussions over equal loudness contour graphs and the frequency response of our hearing at different SPLs (there is a whole explanation/discussion of that in my Foundations of Audio: EQ and Filters course ). I am talking about the ability to accurately hear subtle gain changes in a signal over fractions of a second. While I might crank up the volume when EQing my bass and kick drum, to take advantage of my ear’s flatter frequency response at higher SPLs, I will work at much lower levels while listening for subtle changes in a signal’s envelope during compression.

Another trick I find useful when trying to get a sense of my compressor’s timing is to initially use deeper threshold and ratio settings, then ease back when I find something that works. Working the compressor harder creates more push and pull and can help you hear the envelope of the gain reduction, because a greater gain change is happening during the attack and release phases.

It is important to note that compressors attack and release cycles (while often labeled in milliseconds) can be radically different from one another, even varying with the type of signal and the amount of gain reduction being performed. In other words, one compressor plug-in with an attack of 10ms may react entirely differently than another compressor set for a 10ms attack, and the same compressor’s envelope may react differently to one instrument versus another. This is why I suggest that instead of trying to acquire and learn every compressor on the planet, you set aside a few “go-to” processors to learn extremely well. This way, you will be able to predict how they will react under different mixing scenarios, allowing you to translate the sound in your head faster.

At the end of the day, manipulating a signal’s envelope using a compressor’s attack and release controls is an entirely subjective task, with one man’s punch and swagger sounding like another man’s unwanted pumping and breathing. The sooner you can clearly hear the envelope working, the sooner you will be in control of your own esthetic, and form your own opinions about which compressors you prefer on specific material.

To learn more about attack and release, as well as other compression and limiting techniques, be sure to check out my course Foundations of Audio: Compressors and Dynamic Processing @ lynda.com

This entry was written by Brian, posted on January 15, 2012 at 12:17 pm, filed under Articles, MixTips, News and tagged , , , , . Leave a comment or view the discussion at the permalink and follow any comments with the RSS feed for this post.

The importance of understanding a compressor’s threshold control (video)

As humans, we have tons of built in “thresholds” surrounding all of our senses. What do you do when the shower gets too hot? As the temperature rises, it eventually gets to the point that your nerves tell your brain “ouch” and your brain tells you “too hot!” and you turn down the hot water. That point of “too hot,” be it 110 degrees or 120 degrees,  is your skin’s personal threshold for water temperature.   In the case of a compressor, the threshold control defines the point at which the compressor will begin to alter the dynamics of the signal, or compress. If the signal breaches the threshold or gets “too loud,” the compressor reacts by compressing or “turning down” the level of the output, just like you turning down the water temperature in the shower. In other words, It’s the threshold control that tells the compressor when to compress and when to leave the signal alone.

All dynamic processors work on the simple principle of a defined action resulting in a prescribed reaction. Think of it as an “if..then” statement of sorts; if the incoming signal level reaches “X” or higher, the processor will react in “Y” way, if the signal remains below X then do nothing. The threshold control in a dynamic processor allows us to define the reaction point, or the “if” part of the statement, setting the level at which the processor reacts, sometimes by reducing (compressing) or increasing (expanding) the level of the output.

I know what you’re thinking, you’ve seen that threshold control just sitting there, but the presets you pick automatically set for you, so what’s the big deal? Indeed, most compressor plug-in presets include a preset threshold value, and while it might be tempting to use that level,  it is very important to understand that the preset threshold may not be affective for your specific compression task. If the threshold is set too deep, it might over compress the signal, removing all the dynamics. If the threshold is set too light, the compressor might not engage at all. Of course, sometimes you get lucky and everything works out, but remember, the preset has no way of knowing what your input level looks like or how much compression your signal needs. Until we have intelligent processors that can adjust their threshold automatically to accommodate the input signal, you will need to adjust the threshold control manually. Even if the processor could detect and adjust the threshold automatically based on the input signal, the preset couldn’t  know how much compression you needed or desired, just like no one else would know just how hot you like your showers on any given day.

Ultimately, whether you set up your dynamic processors from scratch or from a preset, it is imperative that you understand and use the the threshold control to achieve the desired amount of processing. If you want to learn more and see some cool animations, watch the following free clip on understanding threshold from my new course lynda.com, Foundations of Audio: Compression and Dynamic Processing. Be sure to check out the entire course for more tips on using threshold and other compression controls.


This entry was written by Brian, posted on January 1, 2012 at 7:54 pm, filed under Articles, Audio and Video, MixTips, News, video and tagged , , , , . Leave a comment or view the discussion at the permalink and follow any comments with the RSS feed for this post.

Mixing with Transient Shapers to control drum envelopes (video)

When you use traditional compressors to alter a signal’s envelope shape, you often trade one problem for another. Spiking out the attack uses deep threshold settings that alter the tone and shape of the decay, sometimes in an undesirable way, forcing us to use “mults” or parallel chains to achieve the perfect attack and sustain characteristics for a given instrument. Transient shapers allow us to modify the transient and sustain portions of a signal’s envelope in very transparent and discrete ways, without altering the tonal character of the signal. To learn more about this topic, check out the following free video from my new lynda.com course Foundations of Audio: Compression and Dynamic Processing.


This entry was written by Brian, posted on December 30, 2011 at 1:10 pm, filed under MixTips, News and tagged , , , , . Leave a comment or view the discussion at the permalink and follow any comments with the RSS feed for this post.

What is a Dynamics Processor? (video)

Dynamics processors are little more than automatic volume controls, responding to changes in signal level over time. If the input signal’s level meets a specific criteria (threshold) then the processor reacts by changing the output signal level in a prescribed way (turns it down, turns it up, etc). Here’s another clip from my new Lynda.com title Foundations of Audio: Compression and Dynamic Processing. Check it out.


This entry was written by Brian, posted on December 28, 2011 at 4:24 pm, filed under MixTips, News, video and tagged , , , , . Leave a comment or view the discussion at the permalink and follow any comments with the RSS feed for this post.

Using gates and expanders to remove bleed from a drum kit (video)

Fresh from my brand new Lynda.com course Foundations of Audio: Compressors and Dynamics Processors, here is a free video on gating a drum track. You can watch all the videos in this new course by signing up for a free 7 day trial.

This entry was written by Brian, posted on December 27, 2011 at 1:14 pm, filed under Articles, MixTips, News, video and tagged , , , , . Leave a comment or view the discussion at the permalink and follow any comments with the RSS feed for this post.

Stillwell Audio = Awesome

So I’m not sure how I missed these (probably because I am off in Pro Tools land and forget to check out all the amazing developments happening in the VST and AU plug-in world) but I just recently checked out the plug-ins from Stillwell Audio and they are so beyond legit. I am really digging the “Event Horizon” clipper/limiter, it’s freaking incredible. Most brick wall limiters let you get things loud, which is nice, but they tend to pump, jack the imaging and add a lot of color to the signal and sometimes that’s not cool. Mastering engineers will often forgo a brick wall limiter in favor of just clipping the front end of some really hi-end A/D converters (obviously you must do this very carefully and you can definitely tell when you’ve pushed to far). Anyways, Event Horizon kind of emulates this practice in the box, and while you can’t use it like a L2 style brickwall limiter, and you definitely have to know when to stop, it sounds so much better than your standard “look-ahead” style limiter on mixes where maximum RMS is not required (read: mixes that don’t need to be ridiculously loud for “competitive” reasons). Event horizon is also pretty sweet on drum tracks, you can push your kick and snare into it just enough to get the transients to settle into the mix while not eating them alive like most limiters. Recently I was mixing a sort of “Beatles meets Beck” style tune and this really captured the drum sound I was after, a little distorted and ruckus but not like running your drum bus through sans amp kind of ruckus.

Sadly, the Stillwell stuff is currently only available as VST or AU, as is the case with a lot of the little one man development teams that either can’t afford, can’t get approved, or simply have no desire to acquire Avid’s SDK. Using FXpansion’s VST to RTAS wrapper is necessary but seems to be working fine with all of the plug-ins I’ve tried so far. Many of them induce a mean latency that causes the delay compensation in HD to go nuts, so I generally have been using them directly on audio tracks before any TDM plug-ins or on the mix bus with the delay compensation manually disabled (it’s the mix bus, everything is going through it so it doesn’t need delay compensation). Unfortunately, dealing with Pro Tools HD’s piss poor delay compensation system is something I am all too familiar with as the UAD system I use creates some mean delay problems, especially when you start using them on sub-mixes and routing second order sends. Seriously, Avid needs to give us more than 4000 samples of delay compensation, which was cool like 6 years ago when it came out, but that is just not enough for the plug-ins on the market these days. I’m totally willing to forgo some extra DSP if it could be bumped up to over 8000 samples.

At any rate, you should definitely check out the stuff Stillwell is coding, a lot like the Massey gear it is super fresh and costs almost nothing.

This entry was written by Brian, posted on May 10, 2010 at 1:01 pm, filed under News and tagged , , . Leave a comment or view the discussion at the permalink and follow any comments with the RSS feed for this post.