Archive for January, 2008

An Introduction to Psychoacoustics

Monday, January 28th, 2008

The purpose of this article is to introduce some of the basic concepts of psychoacoustics so that you can maximize their use in your music and mixes. The information contained herein comes from my personal experience as well as a variety of sources including Mix Magazine, Digital Music Magazine, Future Music Magazine, and much more.

As a guy whose been playing guitar for 35 years, I’m a relative noob when it comes to home recording (last 6 years). Simply put, I am newer to mixing music than to playing it. As a result, these days I spend more of my time studying, learning and practicing my recording and mixing techniques than I do practicing guitar. As my abilities evolve, I’ve come to appreciate the importance of something called psychoacoustics.

Pyschacoustics is the term used to describe the mind’s subjective interpretation of sound. It has also been described as the study of psychological correlates of the physical parameters of acoustics. I find that knowledge in this area can help you create dynamic, powerful mixes that listeners say they enjoy. And whether you are aware of it or not, professional mixing and mastering engineers know the topic well, and intelligently manipulate the psychoacoustics of the music you love to listen to.

Can you read the following paragraph?

“Aoccdrnig to a rscheearch at Cmabrigde Uinervtisy, it deosn’t mttaer in waht oredr the ltteers in a wrod are, the olny iprmoetnt tihng is taht the frist and lsat ltteer be at the rghit pclae. The rset can be a toatl mses and you can sitll raed it wouthit porbelm. Tihs is bcuseae the huamn mnid deos not raed ervey lteter by istlef, but the wrod as a wlohe.”

Of course you can. As stated in the paragraph, your brain doesn’t need to have the perfect spelling of every word in order for it to read, relate to and understand what is written. Essentially, your mind fills in the blanks. There are numerous visual examples of how the brain fills in pieces of a picture or scene in order to create a complete story your mind can relate to and understand. The same is also true in the world of audio.

THE 4 STEPS OF LISTENING

Just as our minds fill in the missing parts of visual stimulus to create an image we can relate to and understand, so our mind fills in pieces of a sonic image in order to provide us with a more complete auditory experience. In essence, how the mind does this is what the field of psychoacoustics studies. And a deeper understanding of how the mind works with sonic images, can greatly improve your ability to create more impactful music and mixes that excite the listener and leaves them with a pleasant listening experience. But for people without a lot of recording or mixing experience, a good listening experience tends to begin and end with their monitors.

It seems as though we are always searching for the “perfect” pair of monitors. Monitors that are so reliable that if you know how to mix, and you mix on those magic monitors, you won’t be able to go wrong. And yet, even pros wrestle with getting the “right” mix that sounds good no matter where you are and regardless of what speakers you are listening to. Why is it that the sound of mixes can appear to change so much from one environment and set of monitors to another? The first and most obvious reason is the quality of the monitors. But mixes can even sound different on monitors of similar quality in similarly configured rooms. What’s up with that?

There are a number of factors that influence the creation of a universally good sounding mix. Monitors and room acoustics are two important factors for sure, but not the only ones. As important is the sonic perception of the listener and engineer. Even great engineers pass through a series of similar steps as they mix.

1.) First, a person sitting down to mix, has their initial listen of the music and naturally form a first opinion.

2.) Second, they start to become more familiar with the music as they work with it and their first opinion goes through some adaptation and change based on a deeper evaluation.

3.) Third, they really starts to hone in on very subtle aspects of the sound.

4.) Fourth, their ears can become “bored” with the sound and as a result, they can begin to have doubts about the precision of their choices, their ear, or their monitors.

These stages are natural, and we all go through them. You don’t have to be an engineer to experience them. And these stages can happen without being consciously aware that they are happening. So one thing great engineers know, are their own listening strengths and limits. They know how to take breaks so as to not get bored with the sound. They know that they are going to go through these stages, and they balance their choices in mixing accordingly.

So knowing the stages our brain goes through as we listen to music is an important step in learning how your mixes can impact the listener. As you develop your ear, learn the strengths and weaknesses of your hearing capabilities.

In an effort to maximize the freshness of your listening capabilities, it’s important to understand how playback levels and ear fatigue impact your work as well.

PLAYBACK LEVELS AND EAR FATIGUE

Playback level refers to the Sound Pressure Level (SPL) at which you are listening to music while you track, mix or master. Although every person is different, it’s worth learning what levels work for many professionals as well as what works best for your ears. Using the right playback levels for your ears will help you minimize ear fatigue, and maximize your ear’s ability to listen critically and discern subtleties.

It all begins with the first time you sit down to work on a session. Paul Elliott, Chief Mastering Engineer for Disc Makers’ Sound Lab says the following about setting the right playback levels, “When I’m first listening, I’m conservative and listen fairly low, around the mid- to low 80s, which is slightly above conversation level. This tends to point out anything that’s a major mix issue with the project, although obviously, with bass, you need to play back a little louder to hear what’s going on there.”

Adam Ayan, a Grammy Award winning resident of Gateway Mastering says, “I operate at a fixed level whenever I’m EQing, and it’s somewhere between 85- and 90dB SPL. That’s not really loud, but once I have my EQ set, I can give my ears a break.”

Other professionals such as Fred Kevorkian, owner and operator of New York City’s Avatar Studios, say, “I don’t know exactly what SPL I’m listening at, but it’s always the same. To me a softer listening level is best when I need to hear details, mix balances, et cetera. Listening for distortion and ticks are surprisingly more noticeable at low levels.”

Additionally, mixing and mastering engineers usually deploy certain practices when taking their first listening pass. Paul Elliot says, “It’s kind of like the 30,000-foot view, where you determine whether there’s anything that I’ll need to talk to the engineer or artist about. After that, I’m looking at the overall tonal character. Something ’soft’ in the top end or ‘big’ in the bass may lead to a discussion about how much of this you want to keep.”

Learn what playback levels work best for your ears. You may want to start by using playback levels between 85- and 90dB SPL. You may find you need to go a bit higher or a bit lower. Once you find your comfort zone, defined by what leaves you with little or now ear fatigue, stick to that playback level and be consistent.

BEYOND MONITORS AND ROOM ACOUSTICS

Think about it like this. Sound really only exists in the mind of the listener. It’s the age old question, if a tree falls in the forest, and no one is there to hear it, did it really make any sound? Well, scientifically speaking it is possible to measure the sound waves produced by that falling tree even if no one was there to see it occur. However, even though we can measure a sound wave, what we cannot guarantee, is what that sound will feel and sound like in the mind of the listener. Great studio monitors don’t change this fact. In fact, professional mixing engineers have confirmed that the way a mix sounds to them can have a lot to do with the mood they are in, and even how well they slept the night before.

One significant variable in how music sounds to us, has to do directly with how it’s arranged. Decades ago, professional studios employed professional arrangers, whose job it was to arrange the instrumentation and playing of the music in such a way so that each instrument sat within its frequency range. This made things sound clearer even before tracking began. For example, if you listen to many recordings by Dire Straits, they tend to always sound good no matter where they are being played. On a great home stereo, in a car, or on an elevator, their recordings sound really good. Much of this can be explained by their arrangement. They are very well balanced arrangements, whose instrumentation nicely fits across the frequency spectrum with a minimum of masking, even prior to recording and mixing.

And herein lies another interesting point. The cleaner the sound signal, the more likely it will sound good on a variety of platforms. Wherein the more distortion a signal has the harder it is to achieve consistency with regards to sound via a variety of different monitors and environments. This has been tested using sine waves and pink noise.

A sine wave is basically the simplest form a sound wave can take. It is a sound wave without any distortion. Scientists have tested sending sine waves via a variety of different monitors and tested the perceived sound that resulted from each monitor. The results where that listeners found very little difference between different monitors when listening to a mid frequency sine wave. However, when the same test is done using Pink Noise, a wide variety of responses were recorded from the listeners.

Pink Noise is a signal with a frequency spectrum such that the power spectral density is proportional to the reciprocal of the frequency. But what does that mean? Basically, it means that Pink Noise is a signal that contains all of the frequencies occurring at the same time. When the same test was done sending Pink Noise through a variety of different monitors, listeners had great variety in their reports of how things sounded to them. What this means in musical terms is that the closer a mix comes to having noisy signals, distortions etc. the greater the tendency that the mix will sound different when listened to on different monitors. Another way of stating this is to say that the more information that a mix contains, the more likely that listeners may hear small differences when listening to the same mix on different monitors. This is not a bad thing, just something to be understood by the person doing the mix.

Another issue that impacts the listening experience is familiarity. Sometimes, we become so familiar with a particular piece of music and its arrangement, that our brains lose their ability to stay objective and discerning. For this reason, professionals know when to take breaks, and the process of creating a modern music CD will often include different engineers for tracking, mixing and mastering.

Bernie Grundman a 42 year veteran of the music industry and owner of Hollywood-based Grundman Mastering says, “After hundreds of plays, mixers can get used to the song, thinking they’re hearing the lyrics louder and end up burying the vocal because they can her it and they can ‘fill in’ the word in their mind.”

Grundman says that similar psychoacoustic effects can occur with surround mixes. He says, “When I hear them I might think that the vocal is kind of low, but when you hear that same cut with the picture, it’s perfect because when you see the artist singing, your brain fills it in.”

A common application of psychoacoustic effects today can be found in how the perception of bottom end is created. Bass has become more and more prominent in contemporary music in recent years. Grundman explains that, “One problem with bottom is that it tends to be very resonant and rings a lot. It tends to cloud a whole mix. If you can get a sense of level out of the high end of the bass, essentially the higher-end transients, then you don’t need so much bottom. You’ll think it’s louder than it is if you can hear the punchy snap of the kick drum and the slap of the bass.”

SUMMARY

In summary, there are many factors that impact the quality of one’s mixes, not the least of which include your monitors and the acoustics of the room in which you are mixing. Our brain’s disposition is equally important in defining a quality sonic image. As a result, it’s best to track, mix and master at consistent playback levels, with a clear head, well rested and in an even tempered mood. Pay close attention to the natural arrangement of the music even prior to EQing, as that too impacts the sound quality going into the mixing and mastering process. Be aware that clear signals are easier to make consistent across different monitors than distorted signals. And don’t let your familiarity with a piece of work get in the way of discerning subtleties. Take frequent breaks, and use familiarity to your favor so as to create mixes that allow the listener to fill in portions of the sonic image without you having to actually have the sound added to your mix. And what about those monitors?

Often times, professional engineers will end up staying loyal to a certain pair of monitors. Not because they are the best, or most expensive, but because they are extremely familiar with their strengths and weaknesses, as well as with their own ears’ strengths and weaknesses. As a result they can consistently produce high quality mixes using the same monitors as opossed to changing monitors every time they notice a discrepancy.

All of this information, combined with the mixer’s knowledge of their own ears and monitors, play an equally important role in producing consistently high quality mixes.

See you around the Forum!

An Introduction to Gain Structure

Tuesday, January 22nd, 2008

What follows is an article I have put together based on research I have conducted, personal practice and experience, and some enormously valuable technical help from John Scrip, A.K.A. MASSIVE Mastering. Although I’ll talk about analog systems here, the majority of this article is written to help the digital home studio.

Eliminating unwanted noise is a fundamental task in any studio, be it a digital in-the-box set up, or a purely analog rig. But perhaps the most important point of understanding your gain structure, is its impact on headroom in your mixes. As John has pointed out to me, everything starts with having and protecting the right amount of headroom at every stage in the audio production process.

Building in the right amount of head room begins with having the right gain structure in your signal chain when you begin tracking. It’s widely understood that mastering too hot is not a good thing. Mixing too hot is even worse, as it will leave less headroom for the mastering stage. But tracking too hot is the worst thing you can do to a mix. Since it’s the first stage, it drives a domino effect of hot and hotter signals throughout the rest of the audio production process.

Digital recording systems that operate at 24 bit resolution are designed to capture a high quality sound while assuring the best signal to noise ratio and dynamic range possible. Having said this, dealing with noise and distortion are a frequent challenge for the home studio operator. In the majority of cases, your equipment is not the problem, but the gain structure of your signal chain may be.

STRUCTURAL INTEGRITY
Using the correct gain structure in either digital or analog studio scenarios assures that the final result utilizes all of the dynamic range available while minimizing background noise and without any undesired signal distortion. This may seem like an obvious and generic statement, but it’s very relevant when discussing gain structure as any added distortion to your signal in an early stage of your signal chain, cannot be later removed in subsequent stage. Independent of the function of any of the pieces of gear (hardware or software) in your signal chain, it’s important to guarantee that the level of the line out of each piece of gear is adequate for feeding into the input line of the next piece of gear in the chain. Every piece of your signal chain (hardware or software) should be adjusted so that each receives the appropriate input signal. As a consequence, all pieces of your signal chain are optimized within their dynamic range.

THE WEAKEST LINK
A good gain structure is based on having every piece of your signal chain operating within the appropriate range even though unwanted distortion in your signal can occur at any point in your signal chain. Preamps are especially susceptible to internal distortion, especially those that are built into lower end mixers. It is also possible that you microphone adds distortion, but it is more likely still that your preamp is creating the distortion. In this case, an accurate measurement can serve as a guide to solving the problem. Once distortion has been introduced into the signal, it cannot be resolved by lowering the volume of the channel. You first have to correctly adjust the gain of the preamp. Just make sure you do so without an EQ, as EQ can influence the signal being measured.

ANALOG VERSUS DIGITAL
The most critical moment to guarantee an optimal gain structure is during the tracking process. If you are working with analog equipment there will always be a certain dynamic margin (the amount of space available between a processor’s line level signal and the point in which distortion begins to occur) available before any distortion becomes evident. But keep in mind that depending on the converter and how its calibrated, analog signals can become distorted well before they will actually clip the digital converter, again assuming the converter is calibrated low enough.

But moving from the analog world to the digital world we eventually reach a point of no return wherein all of the bits are 1s and anything beyond that point corrupts the signal. Twenty four bit recording allows for much more dynamic range than we can ever achieve with analog gear, and it is important to not run an analog signal too hot (+12cBVU for example), so as to leave enough headroom. Smaller analog distortions can add up during a mix. For this reason, the importance of headroom (assuming a converter calibrated to -18dBFS = 0dBVU) is to allow healthy recording signals with lots of room for the occasional errant peak.

A possible solution is to adjust the sensitivity of the line in on your sound card. Instead of adjusting the line out of the mixing console to match the line in of the sound card, it’s possible to configure your sound card to a sensitivity of -10dBV. If the level of the signal of the line outs of the mixing console are +4dBu, you can achieve 0dBFS for the digital line in with a level of +4dB VU on the analog mixer. One tip is to insert a VU meter into your signal chain, wherein you can monitor that input levels are within reason. A -10dB signal at 0dBVU sent unbalanced is going to read 0dBVU. Just as a +4dBU signal at 0dBVU is going to read 0dBVU.

UTILIZING COMPRESSION
To obtain the maximum signal-to-noise ratio for a digital line in, it’s not always easy. With a 16 bit system for example, leaving 6dB of margin reduces the level of signal to noise to 90dB. Some argue to very lightly apply an analog compressor before going into the line in of the sound card. This can smoothen out the transitory peaks and raise the level of the weaker signals. However, others including John Scrip, (A.K.A. MassiveMastering), say a 90dB dynamic range is much better than nearly any input chain. A 6dB margin, even at 16 bits, is hotter than the desired input. Even if your input levels as low as -30dBFS in 24 bit (or around -15 to -12dBFS in 16 bit) that wouldn’t be leaving too much headroom.

DIGITAL DYNAMICS
In the digital world, a poor gain structure can have a much more catastrophic effect than in the analog world. If you saturate a virtual channel beyond 0dBFS you can damage the audio signal. The advantage of those virtual mixers is that the measurement can help you detect and resolve problems easily. All of your channels, busses or aux channels, groups or main outs, have a peak meter that gives you additional visual information about your audio signal. But if they are approaching Full Scale (0dB) the mix is too loud and all of the headroom required for mastering is being used up. Keep in mind, that in the digital world the signal is always controllable. If it’s too hot, you simply turn it down. However, in the analog world, any damage done to the signal on the way in is permanent.

And so I end this article where we began. Everything starts with having and protecting the right amount of headroom at ever stage in the audio production process. Building in the right amount of head room begins with having the right gain structure in your signal chain.

See you around the Forums!