One of the most basic, yet fundamental questions I ever get asked as a sound engineer, whether talking about live or recorded sound, is:
What makes sound GOOD?
This will be a longer post, going deeper into the different aspects of live and recorded sound, and what makes it good or bad.
TOTAL READ TIME: 15-20 MINUTES
WHY DO WE EVEN CARE?
Whether you’re recording audio or amplifying sound for a live performance, the quality of the sound is one of the first things an audience will notice if it’s bad. If your intent is to grow an audience or a fan base, or have audience members come back again, you can’t have bad sound. It’s within this context that we’re covering these following topics, specifically, the qualities of good sound and how to overcome the problems that cause bad sound.
I've created a free 5-day mini-course about the principles behind becoming a full-time creator. You can sign up with just your email.
The most important thing when it comes to sound is Intelligibility. If you can’t understand what someone is saying, than the sound system or the recording has failed. While what follows is not a comprehensive list, these are definitely the main offenders:
1. Too often, sound that would otherwise be perfectly intelligible is ruined simply because the mic was placed incorrectly, or in a suboptimal position. When you are recording someone’s voice, or amplifying it in a live setting, it is important to get the mic as close as possible, but not so close that you start getting distortion or proximity effect. Proximity effect is when the lower frequencies are amplified much more than the mid and high frequencies, causing the tone of someones voice to sound boomy or low end heavy. If the desire is to record or amplify someone’s voice as accurately as possible, this is not something you want to have happen.
2. Secondly, if the mic is placed “off axis”, it may be rejecting part or all of the sound coming from your target’s mouth, changing the tone and the intelligibility of what they are saying or singing. Microphones are built with different polar patterns, which determine how the microphone picks up sound. For example, this is what an omnidirectional polar pattern looks like:
What this means is that the microphone picks up sound equally no matter what “axis” you are addressing it from. Front or side, top or bottom, left or right, it all “technically” sounds the same. Now some mics accomplish this better than others, but the idea is that there is no incorrect way to speak or sing into that microphone.
Another popular microphone polar pattern is the cardiod pattern. This pattern is created so that it picks up sound directly facing the capsule, but rejects sound from certain angles. This is beneficial when you are on stage, for example, and want to pick up the sound of the lead singer’s voice, but not the sound from their monitor on the floor.
The further off axis you address this mic, the more the tone will change, and the lesser the intelligibility.
There are variants on the cardiod pattern, such as the supercardiod and the hypercardiod. These pickup patterns make the pickup pattern smaller, making it better at rejecting sound that is off axis. A common example of this is a shotgun microphone used on film sets and other recording, allowing it to capture a single voice and reject any unwanted sound.
To fix this, make sure that you are addressing the mic properly, and are not too far away from the mic.
While your mic placement may be impeccable, the end result can still come out unintelligible if other recorded sounds are too loud in relation to the vocal. Whatever is most important – the dialog, the lead singer – should be the most prominent part of a mix. If the level is too low, the other instruments or elements in the mix will then be too loud in relation, and will clutter up the mix making it hard to hear what needs to be heard. This can be prevented as long as you have proper signal to noise ratio with each of your microphones or channels.
2. SIGNAL TO NOISE RATIO
Signal to noise ratio, simply, signifies the difference in level between what you want to hear and what you don’t want to hear. In this sense, noise could be anything from the wind, to traffic, to other people or instruments, handling noise, electronic noise, or any other unwanted sound. All electronic equipment has an inherent amount of noise, often called the “noise floor”. Expensive gear tends to have a lower noise floor, and often touts it as one of the features of a particular product. And, as expected, less expensive gear tends to have a higher noise floor. How can you tell? What does it sound like?
Here are two recordings. The first was recorded with the input level as low as possible, the level meters barely registering as I was speaking. Then, in an audio editing program I brought up the level to a suitable listening level. Listen closely for the electronic noise of the pre-amps, even though my room is relatively quiet (50dB).
Now, the same dialog, recorded at a proper level, meaning that the average level was close to −12db on a full scale level meter, and that the volume never went above 0db, or clipped.
The level is the same after I raised the first recording, but was very different when it was recorded. When you raise the level of a recording that was recorded too low, you are also raising the noise that was recorded with it. This is why you want proper signal to noise ratio when you record or amplify anything. Proper signal to noise is a combination of proper mic placement, as discussed earlier, and proper level adjustments on your recorder or mixer.
3. TYPES OF NOISE
Speaking of noise, there are a few we should talk about so that we know that 1 – they exist, and 2 – how to correct or prevent them.
If you’re recording or amplifying sound outdoors, you should be aware of the risk of wind and the effect that wind has on a microphone. Essentially, a gust of wind will distort the microphone element, causing a terrible sound, and unintelligible dialog. Notice that when we recorded this video, we didn’t have any of these things.
The best way to prevent this is to use accessories made to prevent the wind from hitting the microphone’s element. Windjammers, blimps, and foam windscreens are all options that help stop wind without drastically changing the tone of what you’re recording. At minimum, you should have a foam wind screen on the mic. Blimps and Windjammers are heavier duty options, mainly used on film sets to record dialog for film and video.
When you walk into a room, you often hear lots of things. The AC or heater running, fans, electrical equipment hum, refrigerator hum, ceiling fans jingling, florescent bulbs that hum, and the list goes on. All of these things are considered part of the room tone, and some of it can be controlled, and oftentimes some of it can’t. The best thing you can do before trying to run a live event or record audio in a space is to “scout” out the space beforehand. All these problems are readily apparent. At that time you can assess what you can and can’t control. You can unplug or turn off loud electrical equipment. You can turn off ac or heater units. You can replace bulbs that are flickering and making noise.
In addition to these noise generators is the actual “tone” of the room. Ever walk into a small bathroom and talk or sing and you can hear the room? What you’re hearing are the reflections of the room. What about when you go into an empty gym and bounce a basketball. Hear how long it takes for the “echo” of the bounce to go away? It could be upwards of 3 or 4 seconds in some rooms. Sometimes, this is desirable. If, for example, you want to find a good place for a choir or orchestral performance, a room with a long decay time – the time it takes for a sound to not be heard after it is made – might be favorable. It will make the sound more immersive and pleasant.
However, what if you’re doing sound for a rock band? Snare drums, bass guitars, and amps don’t sound their best in a big echoey room. The sound of the drums starts to take over. The low end from the bass is all you can hear and it builds up and seems even louder because the room is amplifying those low end frequencies. You can’t even hear the vocals because they are completely unintelligible, due to the note 3 seconds ago still sounding over what they’re singing now. You get the picture.
Here’s an example of a really echoey recording.
This was recorded in a room with a very hard ceiling, parallel walls, and big windows that didn’t provide much noise reduction from the traffic below. All of these things equal bad sound. It’s a strain to try and hear what she is saying, because there is so much echo and noise. Even if you were to record with proper mic placement, you would still hear the room.
Now, put that same person with the same level of speech in a room that has a much lower room tone, and this is what you get.
As is readily apparent, you can hear the difference the room makes in a recording. So if you’re filming, or recording a podcast, or doing a live show, spend some time before hand going to the location and listening to what the room sounds like, seeing what you can do to improve it, and make sure the recording sounds as good as it can
Noise is generally described as unwanted sound that is added to the desired sound source – so wind noise, hum, room tone, etc. Distortion however is an alteration of the sound you are recording. A normal sound wave looks something like this:
A clipped soundwave, however, looks more like this:
And a closer look at that same distorted, clipped audio.
Distortion is normally caused by either recording at too high an input level, so that the microphone or the input is distorting because it cannot handle that level, or a problem with the equipment, like a bad cable or bad connection. The waveform is flat on the top, rather than smooth and rounded, because the signal is hitting a ceiling, signifying the gear can’t handle that much level. So it just kind of craps out, and you get clipped audio.
This is part of the reason why live engineers do sound check. It is equal parts checking for proper level and tone, and making sure the signal coming down each channel is clean – free of noise, hum, and distortion.
It is ESSENTIAL that your recording or amplified signal is free of distortion. You may need to replace a microphone or cable, lift a ground to remove hum, or otherwise position the mic so it isn’t distorting.
Tone is one way to refer to the frequency response of a recording or a sound signal. We’re getting into a more technical aspect of sound quality, but it’s very important. Once you have achieved a proper signal to noise ratio and eliminated all the noise, now you can focus on the finer points that really make a recording or sound system sound great.
An unbalanced tone is obviously undesired. Too much high end and the signal will sound “thin” or “bright” or “tinny”. Not words generally used to describe good sound. Same with too much low end, that relates to sound that is “boomy”. If the sound is too unbalanced, intelligibility will start to suffer as well. Let’s take a signal and start pushing it around so we can see what I’m talking about.
The more unbalanced a signal, the more intelligibility will suffer, but before that, the sound will become undesirable to listen to, and you’re going to lose your audience.
We’ve covered this a bit so far, but it deserves to be expanded on just a little more. Let’s split this up between live sound, recorded sound, and sound for film.
Live Sound Volume Levels
When at a concert listening to a performance through a sound system, you as an audience member should be concerned about the sound level coming out of the system. The level of sound is measured in decibels, or dB, and this level is referred to as the Sound Pressure Level, or SPL. For example, the SPL of a person talking to you from a distance of a few feet is around 50dB. The sound of a lawnmower is closer to 75dB. The sound of a jackhammer at a few feet away is 100dB, and the threshold of pain, the point where you actually feel pain in your ears from a sound hitting your ear drum is 130dB. So it’s safe to say that we would want the sound in a concert to be under 100dB, and definitely not approaching 120 or 130dB. However, sound systems are capable of producing sounds at these SPLs. As an audience member, it is a safe approach to bring ear plugs or ear buds that can lessen the SPL reaching your inner ear, especially if you plan on being close to the stage or the sound system. It is common for the front of the stage to be 110dB or higher, especially for large events. Often the sound engineers will raise the speakers into the air, so that the front rows aren’t getting blasted while the back rows can barely hear. This provides for a more even dispersion of the sound.
If you are the engineer, you are typically mixing from a position that is half or 2/3 the way back into the room. A good rule of thumb is that as you double the distance, the SPL drops by 6dB. So, for example, if the speakers are 120dB from 5 feet away, they would be 114dB at 10 feet, 108dB at 20 feet, and 102dB at 40 feet. If you’re in a small club, this is obviously way too loud. You ideally want to be around 85 or 90 dB at the mix position – again, 1/2 or 2/3 back in the room – which would place your speaker volume in our earlier situation at just over 100dB at a few feet from the speaker, so that at 40 feet back it is 85dB. The reason for this is because sustained listening at high SPL levels has been proven to cause hearing damage. You can listen safely to an SPL of 85 for close to 8 hours without hearing loss. With every increase of 3dB, that time is cut in half. So 88dB for 4 hours, 91dB for 2 hours (how long is your average concert?), 94dB for 1 hour, all the way up to 115dB for about 30 seconds before you start hurting your ears. It is your responsibility as the sound engineer to not be inflicting pain on your audience.
When it comes to levels between audio cd’s and radio signals, there is something that has happened over the last 20 years or so that is commonly referred to as “The Loudness Wars”. Go ahead, GOOGLE that and see how many people have written about this subject.
Essentially the loudness wars are this battle to try and have the loudest recording. This generally is achieved by compressing the mix of a recording as much as possible, and boosting specific frequencies that simulate the sound being closer than it actually is. This is all at the sacrifice of dynamic range, which is the difference between the lowest and highest end of the SPL of the signal. Live sound tends to have the largest dynamic range, followed by theater sound, DVD sound, CD audio, and Cassette Tape. This is largely dictated by the ability of the medium to reproduce the larger dynamic range, and the environment where the recording is listened to.
If you are a recording engineer, you’ll want your recordings to sound at least in the same ball park as these professionally recorded and released albums. However, go too far and you’ll compress out all of the dynamic range, making your mix seem amateur for a different reason. Here’s an example of two recordings of the same song, one recorded in a college course as a class project, and the other recorded a few years later as part of a professionally recorded album. The music in both examples is “Where Are You” by BENTON PAUL, a good friend and super talented musician.
Notice the difference in everything we’ve discussed. Noise level. Tone. Loudness. All of these are much better in the professional recording. And it’s what makes the professional recording just that – professional. Nice job, SCOTT WILEY.
Sound For Film
Film is one of the only mediums that has very strict level requirements that have to be met in order for a sound to be allowed to be played in a theater. Average loudness, peak loudness, dynamic range – all of these are measured using expensive equipment and measurement tools. This is to allow consistency between playback systems, but also to protect the ears of the listeners.
If you are doing sound for film, you should spend some time on the DUC and Gearslutz forums, especially the post production forums, and the posts that go into detail about room measurement, mix levels, and the like. It’s much bigger than the scope of this post, and much smarter people than I have already taken the time to describe these processes in extreme detail.
One thing that is important is the sound between edits. It’s such an easy but often overlooked part of the editing process that it boggles my mind that people don’t fix this.
I’m referring to the sound that occurs between two edits when the editor cuts into a sound wave. If you don’t edit at what is called the “zero crossing”, you will hear a pop in the sound as you cut to the next shot:
The easiest way to fix this is if your editing or audio software has the option to “snap to zero crossings”. What this means is that any time you cut an audio region or event, depending on what your software refers to it as, it will only cut at zero crossings, where the sine wave intersects with the zero axis. It will automatically prevent any pops from occurring.
The second thing you should do is not cut into the middle of a word or a sound. It’s as if I -ere -o -rite -ike -his. Not very good. If you must cut into a word, then at very least put a small fade at the end of the previous clip and the start of the new clip and have them overlap by a frame or more, so that you don’t hear the pop between the edits.
Something else to consider is the difference in level and in noise level between shots. This is why it’s important to eliminate as much noise as possible when you record audio on set, so that it makes for a smoother edit. Cutting a noisy shot in with a quiet shot makes for a really jerky sounding edit, the noise popping on and off with each edit. If you do have noise or different noises, you can try to do a slight noise reduction with a plugin, or use a multi-band compressor, or do some creative work with your cross fades between edits so that the change is less noticeable.
This is a great way to go about getting better recordings and better sound in your live performances. Spend some time experimenting with the things discussed in this post, and start getting better sound!
I've created a free 5-day mini-course about the principles behind becoming a full-time creator. You can sign up with just your email.