That “1176 Sound”

Why has the 1176 compressor been a staple in hit pop mixes for over a half century? Let’s explore the magic of this classic audio processor.

Besides equalization, automatic volume leveling, or as it’s more commonly called in audio engineering, “compression and limiting” contributes more to the overall character of modern pop production than any other factor.

We mix engineers are interested in compression and limiting because they are crucial tools that many of us have struggled to master. There’s probably not a professional mix engineer who hasn’t regretted compression decisions — it’s one of the obvious signs of an amateur mix.

So, what is the difference between limiting and compression? Well, at minimum, it’s just a matter of how aggressively attenuation is applied. A copy of the 1980 1176LN manual describes the traditional definition:

Although no strict standards exist, amplifiers with compression ratios of up to 8:1 are called Limiters. By this definition, the following graph shows that the Model 1176LN can function as a compressor or a limiter since it has compression ratios selectable from 4:1 to 20:1.

The 1176LN transfer curve printed in a 1980 UREI owners manual

That’s still a reasonably sensible distinction. However, in garden variety plugins we use today, both brickwall look-ahead limiting and true peak (∞:1) limiting are often standard defaults in limiters. So, for most of us, the definition of “limiting” has shifted to mean strict limiting that approaches ratios of ∞:1. The artistic difference is subtle but important in practical applications.

To paraphrase the legendary mastering engineer Bob Katz, when you want to retain the relative musical dynamics of a signal, use a limiter; when you want to alter the musical dynamics of a signal, use a compressor. It’s easy to overlook the nuance of that sage advice, and it’s worth unpacking. Katz is generalizing and is well aware that aggressive limiting with a fast release time certainly has the potential to alter musical dynamics — ITB mixers use aggressive hard limiters all the time on bass or vocals to push the signal “forward” in a mix. Perhaps the best way to understand Katz’ guidance is to look at what a limiter doesn’t allow you to do.

An image of Purafied Audio’s VST emulation of a low-cost Alesis utility limiter from 1988 that was considered “junk” 25 years ago. A rather noisy solid state VCA affordable workhorse for stage and home studios, it was actually a stealthy little circuit design that was not unlike the program dependent response of a classic limiter like an LA-2A. Dead-simple to use and very musical. Now, prized “vintage” kit because Mac DeMarco used one. I remember these well.

The Alesis MICRO LIMITER is about as basic as an analog utility limiter gets. It’s a useful example because it illustrates how a transfer curve achieves musical results. Note that there are no controls for threshold, attack time or ratio. There’s also no indication of what the release time is in milliseconds, so you’re going to have to just use your ears. I find it “cute” that they’ve printed a vague transfer curve on the right without any x or y information, apparently to let the user know that it begins compressing gently before it reaches hard limiting. They knew that their average customers were musicians who’d lose or wouldn’t read the little manual booklet in the box (pre-Internet era). It’s like an attempt to explain the functionality to an alien (or extremely stoned keyboardist). Here’s how Alesis engineers described the characteristics in the 1988 manual.

The MICRO LIMITER can be thought of as a compressor when the green LEDs are lit because of the low compression ratio (2:1 to 8:1). A compressor is usually used to even out the volume fluctuations of an instrument or vocal and sometimes for special effects.

A compression ratio of 10:1 or above changes the compressor into a limiter. A limiter is used to prevent short term peaks (which add little information to the program material) from overloading amplifiers or tapes. It also can be used for certain types of special effects. Once the yellow LEDs are fired, the unit becomes a true limiter with the fourth LED indicating a compression ratio of about 16:1. Since we at Alesis feel that the MICRO LIMITER is more easily and quickly set up by just listening, we have eliminated these designations from the front panel of the unit.

The MICRO LIMITER employs the “soft-knee” approach, which means that there is always some compression applied to all input. The MICRO LIMITER also employs program dependent attack time which allows the unit to be more musical sounding than other compressor/limiters on the market. Because of the unique characteristics of its detection circuitry, which have been especially tailored for use with a rhythm section, the unit can be thought of as a “power box”, adding punch to bass and drums with very few of the undesirable side effects normally associated with compressor/limiters.

A limiter is generally assumed to have an attack time of zero (or as rapidly as the circuit can react), but the MICRO LIMITER purports program dependent attack, which is a hallmark of classic designs. This could simply mean that there is a fixed relaxed attack time combined with soft knee. For now, we can disregard this temporal dimension — the time it takes for a device to react to the control input — and look at how a gentle “soft-knee” transfer curve works in general.

In the transfer curve above, the y axis represents output and the x axis represents input. The Alesis Limiter is a progressive soft knee design (bottom curve) where the ratio of compression begins at very low signals and increases as input approaches its threshold target of 16:1 (which the guys at UREI would agree is limiting). This results in a musical and natural sound because there isn’t a point where the device suddenly and drastically “clamps down” and attenuates the signal. The only substantial control option is release time, which is the amount of time it takes the circuit to release the attenuation and return to zero.

The release time:

  • A slow release time of 250ms or greater may provide a smooth sound with the tradeoff of sudden short power transients noticeably attenuating the subsequent signal. If you’ve ever listened to highly dynamic classical music on FM radio, it will sound like the broadcast limiter suddenly clamping down during a timpani crescendo, followed by a conspicuous volume drop in everything. However, if the input program is already compressed and relatively consistent overall, slow release times have the benefit of retaining the relative dynamics of the material which will sound more natural overall.
  • A medium release time of 50-80 ms is usually the range at which the attenuation can return to zero with reasonable transparency without introducing distortion on low frequencies.
  • A fast release time below 50 ms will be very aggressive and effective for an “in your face” sound with the tradeoff of distortion at lower frequencies. This is the range where we’re either just trimming the peaks with a little gain, or obliterating the signal with a lot of gain. Keep in mind, this may be an acceptable range for most non-bass instruments but isn’t recommended for low frequency instruments or full range mixes. Maximizers, such as those designed by Alexey Lukin and the team at Izotope are algorithmic limiters that seek to solve this problem in professional mastering, with the potential to make your track REALLY REALLY DAMN LOUD.

Now that we’ve reviewed the basic functionality of most limiting amplifiers, we can see that most compressors are just doing the same thing with a wider range of controls. Most modern compressors add the ability to adjust threshold, attack time, release time and ratio. The most advanced systems also allow control of the “knee size” of the transfer curve, frequency equalization of the control input, and root mean square (RMS) analysis of the signal. We’ll look at those latter concepts in a future post that covers more advanced digital compressors.

Practically speaking, the 1176 is a leveling amplifier with a control range optimized for mix both compression and limiting duties. The key technical difference in our modern context is that most limiters feature a fixed attack and ratio. Compressors, on the other hand, allow a variable amount of the transient to pass before attenuation and release achieves the target ratio, which increases the transient. If the design allows (with lookahead and knee), it may also do the opposite: immediately attenuate the transient and increase the subsequent decay to be louder than the transient, which is usually reserved as a special effect.

Enter Attack and Release!

The range of dynamic manipulation possible with a compressor is intimidating and difficult to master. I’ve been using compressors for thirty years and I still get frustrated and impatient trying to dial in the perfect settings. To the neophyte, both compressors and limiters are usually assumed to just “level out” an uneven signal so it remains consistently loud and forward, when in fact, compressors are a much more advanced dynamics manipulation system. The wrong settings on something like a vocal can result in increased plosives, sibilants, and compression artifacts like “pumping”, where the attenuation reduces the volume in unnatural-sounding ways. On a percussive instrument like a single drum mic channel or drum mix, compression settings can dramatically alter the sound, variably described as “punchy”, “thin”, “small”, “boomy”, “explosive” or “splashy”.

To understand how a compressor’s variables manipulate the dynamics of a signal, let’s take a look at a few signal envelopes from the wonderful free Audio Compression Visualizer. This visualization tool assumes a very straightforward design with no program dependent behavior or knee, which is typical in standard VCA or stock utility compressor plugins.

Above we have an unaffected signal where the transients are noted. Output signal is always denoted in blue.
Reducing threshold engages the compression, but the attack is extended to allow the initial transient to pass through unaffected. The pink shape at the top represents the attenuation curve. Green indicates original input. The resulting waveform now has more “bite” or “thwak” relative to the subsequent signal. This is often used to give a percussive signal more attack and reduce boominess such as the initial strike of a drum or strummed guitar. However, on a vocal, this would result in more “puh”, “th” and “ss” sounds.
With a rapid attack and release at a high limiter ratio, we can see that the transients have been removed, which would usually be considered over-compression. This gives you a sense of what aggressive limiting with a fast release can do to a signal. If this were a tom drum, it would dramatically accentuate the boomy resonance of the instrument.
Here are settings that achieve a middle ground where the overall envelope shape is retained but leveled. This would be a common setting for many sources, including vocals.
A slower attack and faster release brings the entire signal forward for a louder “in your face” sound without sacrificing transient attack.

Myth: attack time is the time it takes for a compressor to start attenuating after exceeding the threshold.

While there’s no agreed upon standard for electronics or software engineers to measure the behavior of attack time, I’ve rarely seen a compressor that doesn’t begin compression as soon as a signal exceeds the threshold. It’s not like a “pre-delay” for gain reduction — that would work horribly on highly transient signals — applying sharp sudden gain reduction patterns on the signal 200ms after it was triggered by a previous event? Not good. The attack time is the time it takes for the compressor to achieve approximately .66 of the target ratio (and this varies by compressor). The works well for two reasons:

  • instantaneous attenuation without at least some curve would result in distortion, clicks or pops
  • allowing for another third of gain reduction after the target attack time builds in some program dependence and sounds better.

In other words, attack time is the rate of reduction which allows you to control the smoothness of gain reduction, distortion, and forgiveness of transients. Of course, an ultra-fast attack, like the 20 microseconds of the 1176 is still going to be extremely aggressive and prone to distortion, and this has always a been a challenge with aggressive limiting.

This demonstration above shows how most all compressors react to a signal in the temporal domain. The settings in this example have an aggressive ratio of 30:1, attack at 200ms, release at 500ms, and a hard knee in order to clearly demonstrate the slow 200ms rate of attack and release on clean blocks of noise. The large blocks exceed the threshold, the small blocks do not. Note how the attenuation continues to release at the left of the small blocks below threshold due to the slow release time.

The simple answer to why classics like the 1176 and LA-2A are so popular is that their program dependent designs minimize common undershoot and overshoot problems and just sound good. Before we explore why, let’s look back at the history of limiting and compression in the first half of the 20th century to understand why the 1176 took the audio engineering world by storm.

Limiting amplifiers were originally developed to control and protect critical equipment and radio transmitters from exceeding levels that would damage them, distort them or violate FCC regulations. Early models through the 1930s and 40s were large bulky and relatively primitive automatic level attenuators made of vacuum tubes and transformers with limited capabilities beyond automatically turning down the volume when it exceeded a specific threshold, and usually did so by releasing the attenuation slowly and gently. However, they immediately attracted the interest of recording engineers who needed solutions to automate level control of unpredictable live performances.

A rare 1939 Collins limiting amplifier used by the Army Signal Corps. Photo credit: Reverb.com

By the 1950s two revolutionary compressor/limiters were developed that changed the future of recording: the Fairchild 660 and the Universal Audio 176. The former established the gentle, natural sound that engineers use to “glue” mixes together, imparting a pleasing cohesion to a blend of signals. The latter was designed by “the father of audio engineering” Bill Putnam and his colleagues for more aggressive and surgical possibilities. Being all tube and transformer-based designs, these compact (for the era) products and control schemes introduced the foundation of the technology we still use today. Both the Fairchild and UA 176 remain highly coveted and expensive units.

The legendary Fairchild 670 employs 20 valves and 11 transformers.

Anyone who’s worked with vacuum tube equipment knows they are delicate, unpredictable and challenging machines to maintain. For folks in the 1930s through 60s, tubes were a frustrating and failure-prone component that had a limited lifespan and required frequent replacement. Following the successful demonstration of the solid state transistor on December 23, 1947, at Bell Laboratories in Murray Hill, New Jersey, the path was established for a future of stable, reliable and compact electronics made of semiconducting materials that would make the large and fragile vacuum tube essentially obsolete.

Top view of a vintage Universal Audio 176, the tubey ancestor of the 1176. I recorded vocal tracks for the Vibraflux album through one of these units in 1996/7.

By the mid 1960s, Bill Putnam and his team at United Recording Electronics Industries (UREI) started experimenting with various solid state components to develop a “modern” semiconductor successor to the acclaimed 176. The challenges of reliable, reasonably transparent solid state input and output amplification were already established by circuits they had designed for mixing consoles, but Putnam’s lab also needed a transistor-based solution to the automatic attenuation control. They discovered an approach where a component called the field effect transistor (FET) could be used as a variable gate resistor to rapidly increase the amount of audio input sent to ground (bypassed) as more voltage was added to the control, thus reducing the gain peaks. The response was so efficient that the circuit was, like the 176, capable of microsecond attack (start of attenuation) and very rapid release, which approached or reached clipping distortion. Their design tamed this potential for excess intermodulation distortion with a transfer function curve (knee) on the attack and a longer variable release time that reduced egregious distortion at lower frequencies.

The result was a compressor capable of variable attack times that ranged between very fast and insanely fast to reach the target ratio. Where the 176 already achieved a virtually instant 100 microseconds, the 1176 reaction time approached 20 microseconds, which provided recording engineers a reliable tool that could control dynamic range with unprecedented precision and less “tube” color than its predecessor. Where most compressors struggled to clamp and release on transient signals, the 1176 was capable of bringing the dynamics of a transient-rich signal forward with a hyperreal “in your face” character, teetering on an “edgy” clipping distortion that would define the future of pop music. By accident, British sound engineers later discovered that depressing all the ratio control buttons bypassed the ratio circuitry altogether, revealing the unique non-linear character of the raw FET attenuation circuit: compression of the transient and even higher ratio limiting following the transient which is often described as “program dependent.” It was a completely new sound that made drums sound explosive.

An early 70s advertisement for the blackface 1176LN – the initial low-noise revision that dominated studios throughout the 1970s and beyond. Image credit: Preservation Audio

Despite this lightning-fast attack time, the 1176’s handling of transients is actually quite nuanced. Even in its most aggressive 20:1 ratio setting, it’s not a hard true peak or brick-wall limiter. Compared to most stock DAW compressors available today, which are “feedforward” designs, a crucial difference is the 1176 is a “feedback” topology. This means that the control signal feeding the attenuation control circuit is not the input that goes into the box, but the signal that has already been compressed. While this might sound counter-intuitive, and there is a slight tradeoff in sudden transient attenuation, it’s part of what made such an otherwise hyperfast compressor more relaxed and less prone to over-compression artifacts, which is ultimately a more natural and satisfactory sound. By comparison, my first compressor, a late 80s Alesis 3630 feedforward voltage control amplifier (VCA) design, when dialed in with similarly aggressive sub millisecond attack and fast release times, even with a modest 4:1 ratio in soft knee mode, would sound jumpy and conspicuously unnatural — like an over compressed signal — which isn’t usually desirable.

A stack of UREI 1176LN “silverface” editions. This revision H followed controversial revisions that had changed the output amp to a push-pull configuration from the original class-A design and removed the input transformer, replacing it with a differential amplifier. Collectors consider revisions prior to these changes to be more desirable for their saturation and color.

Compared to those later, highly predictable linear transfer functions of voltage control amplifier designs from the 70s onward, the FET’s nonlinear behavior in ratio and frequency response was quite possibly a “happy accident.” Each of the major design revisions are known for having their own character. Beyond the behavior of the FET control, the amplification stages of each revision’s components and transformers also imparted various blends of subtle harmonic saturation that added more or less midrange density and high frequency presence, depending on vintage, even when no compression is occurring. In this respect, the 1176 is also a pleasingly colorful solid state line amplifier. Tyler at Top Sound Design conducts a harmonics test on various revisions of the Universal Audio emulations where we can observe that most of the harmonic structure is below the noise floor until the unit is set to extremely aggressive limiting.

The original 1966 version with the blue Stripe faceplate had a higher noise floor and midrange saturation that was “corrected” in the “definitive” 1176LN (low noise) revision that dominated every pro studio in the 70s. While the LN revision had superior transparency for a wider range of source applications, twenty five years later, famed mixing engineer Chris Lorde Alge would popularize his particular vintage blue stripe’s density and grit as the secret sauce of his aggressive modern vocal sound. It was so celebrated that Waves partnered with Alge to carefully analyze that exact unit (twice) to develop one of the best selling digital audio plugin series of all time bearing the “CLA” signature branding.

The Waves CLA-76 developed by Chris Lorde Alge and Waves audio.

Now that we have flogged the proverbial dead horse of the mythology surrounding over half a century of 1176 use by leading mix engineers, let’s consider the practical and subjective reasons why people love and continue to choose this device and the many digital emulations of it.

Austin Moore at the University of Huddersfield authored an extensive paper on both the technical and subjective characteristics of the 1176 that goes into great detail. His analysis was a key source for this writing and I found it a fascinating deep dive. As a singer having my own voice recorded through a vintage 176 and 1176 blackface over the years, and having experimented with various DSP emulations extensively, I’ll summarize what I think are the reasons why the 1176 is beloved.

The key to “the 1176 thing” is its simplicity and predictability. Thanks to the combination of program dependent behavior and the relatively limited range offered by its controls, the 1176 is foolproof and reliable. Experienced engineers even sometimes use the term “1176ey” as a shorthand description for fast aggressive feedback compression. The transfer function curve and feedback design allow a user to hit it hard without completely neutralizing the transient character of the signal envelope, resulting in unpleasant artifacts. It’s not uncommon to process a vocal with extreme gain reduction, and it just sounds great and reasonably natural. Compared to many modern compressors, it’s difficult to get bad results from an 1176 because the variations in settings don’t drastically alter the results.

To quote Moore:

According to an article in the 2003 edition of the Universal Audio Webzine entitled “All Buttons Mode” the attack, release and ratio are all source dependent. Shanks points out that the 1176 will faithfully compress or limit at the selected ratio for transients, but the ratio will always increase a bit after the transient. To what degree is once again material dependent. This is true for any of the 1176’s ratio settings, and is part of the 1176’s sound...”

…part of the 1176’s punchy sound is due to how it compresses the attack portion of the transient lighter than the rest of the volume envelope. This program dependent nature of the device may also be one of the reasons it is often listed by producers and engineers as a compressor that can be used on practically any source with good results.”

Moore’s analysis of the circuit and inherent behavior of the FET reveal that much of the ratio control topology was possibly Putnam et al trying to add some variability to an inherently musical but inflexible transfer curve to make it operationally comparable to the 176. Changing the ratio does change the transfer curve, but more accurately, it also raises the threshold and still allows transients through in a pleasing musical way. When aggressive but “musical” compression is desired, the 1176 is a more of a wholesale choice. In fact, most engineers in the 70s, despite being surrounded by control systems that looked like spaceships, were not overwhelmed with choices. Before SSL popularized very “usable and forgiving” VCA compression directly on every desk channel, most of the time a tracking or mix engineer was deciding between an 1176 for more aggressive compression or an LA2A optical limiter for more relaxed, even and natural dynamics control, or often both in series (still a popular approach). The modern neophyte working in a DAW is presented with a stock compressor often featuring an intimidating array of widely tunable parameters, or they may download scores of free and paid options just to manage dynamics: a paradox of choice that can easily derail creative flow or prevent the user from understanding how to achieve a good sound from the wide range of settings a stock compressor is capable of. Similarly, the cheaper “hyper versatile” VCA DBX and Alesis compressors I had access to in my salad days in the early 90s were easy to overuse and ruin signals with, and I often did. Instead of diving right into the 3630 (mine was used and lacking a manual), I would have been better served by getting my feet wet with the foolproof Alesis Micro Limiter. It’s no wonder that modern producers still reach for an 1176 or 76 emulation, because it’s so widely recommended and reliable.

Beyond the simplicity and reliability of the 1176, numerous famed producers have praised it for its pleasant saturation profile that is variously described with esoteric words like “presence” “sheen” and “density,” which means that unlike mere EQ, the device adds new mid to high harmonic energy to signals. When that saturation is combined with aggressive yet “natural” leveling, it sounds like that “magic” we associate with the familiar “pop radio” sound: consistent, clear, forward and excited.

In the DAW era, there are practically unlimited options to sculpt a compelling vocal sound, but if a good vocal take, a little EQ and an 1176 is a reliable old recipe for a satisfying pop and rock sound, then it’s like ordering pizza — easy and most everyone likes it. So, go grab one of the many 1176 clones and use it. If you have a UAD interface, it will offer a faithful emulation by the engineers who build the hardware version, and the company is still led by Bill Putnam’s son, Jr. Apple Logic has an excellent stock “FET” emulation that reliably performs like a 76. The aforementioned CLA 76 by Waves (already considered a legendary plugin) is usually available for the cost of a decent restaurant meal. The most accurate free version I’ve found is FETish by Analog Obsession, but it’s a CPU hog that doesn’t dabble with nice predictable and efficient formulas like many older plugins. It is likely a component by component emulation of the entire 1176 circuit and attempts to use as much of that spare Moore’s law CPU headroom on your computer as possible to do so. Bouncing channel process choices is advisable for archiving anyway, so find that sweet spot and commit to a render.

Since our YouTuber above already provided a comprehensive demonstration of UAD’s “faithful” emulations of the 1176 with test tones. I thought it would be interesting to analyze the saturation characteristics of FETish – a very popular free alternative to the 1176.

Typical settings that approximate the “Dr. Pepper” sweet spot on a normal 1176 on a 440 sine wave.
The plugin with any abount input gain immediately generates a prominent 3rd order and other odd harmonics, regardless of ratio. This means that the modelling of the amplifier stage is probably adding some “warmth” independent of compression.

To contrast, PreSonus’ ultra CPU efficient Fat Channel FET compressor, a plugin I find myself reaching for increasingly often for quick mixes, does not appear to model any line amplifier saturation. Given the low CPU footprint, I wouldn’t expect it to without expensive oversampling. However, with aggressive limiting at fast attack and release settings, the plugin does generate harmonic distortion similar to UAD emulations, even at 4:1 ratios. I’d expect any compressor to saturate with such aggressive attack and release settings. Fast compression and limiting are essentially clipping behavior.

PreSonus FET Comp only generates harmonics at very aggressive attack and release settings. In comparison to FETish, no harmonics are generated at typical Dr. Pepper settings.
Saturation with odd harmonic patterns: 3rd, 5th etc.

Meanwhile… for those who enjoy more challenging puzzles, there is a completely free “vintage DSP” broadband compressor engine developed nearly twenty years ago with practically unlimited parameters and a lot of intimidating slider controls and numeric input boxes that can do far more for precision dynamics control with supremely efficient computational elegance. It’s a good example of the polar opposite of the 1176: digital, surgical, sterile, unforgiving but incredibly powerful if you dare to unlock its secrets. We can dive into that vintage software treasure in my next post.