Fixing Audio Noise Problems in Mixes or Post-Production

Do you have some audio problems to Iron out in your mixes or post production? Are clicks, pops, hums, buzzes driving you nuts? Well you have come to the right place : "Our She'll be right mate, just give em earplugs school of audio engineering”, where we think the problem is all in the minds of the listener.


Nah seriously, no one wants their mixes or audio-visual productions to sound substandard, so here are a few tips I hope will help to give them a bit of a hand up. Firstly, it's best if you don't have any flaws that need fixing by the time you get to the mixing, and then mastering phase of your project. Avoid the "fix in the mix" attitude. A well recorded song, voice-over, dialogue or foley is always going to sound better than one with poorly recorded audio, no matter how many "fixes" that you attempt to apply later. So try to use good quality equipment, good technique and good settings when you record. Aim to capture the full range of frequencies of a voice or instrument, as it is easier to attenuate them later than to try to add what is not there. You may decide you do not want to use EQ during the recording stage, but often using a high pass filter to remove unwanted low bass frequencies is a good idea when recording voice, cymbals or other instruments and sounds that don't need those deep bass frequencies. Hopefully you won't need to fix too many issues later, if things are done right at the source during recording. However, if any problems do emerge at the mix or post-production phase that you were unaware of before, try to fix them on a track by track basis, rather than attempting to fix them once you have a 2-track mix. The earlier you can deal with things that are potential difficulties the better. I will have a bit of a run through a few of the potential hazards that might trip up and give you some tips to help alleviate them.

DC Offset

Cheap sound cards or recorder inputs often add a DC voltage offset to a recorded sound. There are sometimes other factors that can cause this as well. It is at 0 Hz, which we can't actually hear, but it offsets the wave so that it is centered above the zero line. This means you have less headroom before clipping occurs. Another thing that may make DC offset an issue is when doing cuts and edits, even in an apparently silent passage you may get clicks or pops at the beginnings and ends of edits. Check out your tracks, and if they look like the wave is not centered around the zero, then you need to filter it. Or if when you were recording, the meters were reading something even when you were not sending them any signal: that may be a giveaway. Most recording programs have a function called something like "remove DC offset". Use it if the problem is evident. Or filter out everything below 10Hz. Do this prior to any compression or EQ'ing. It should get rid of any offset.

Hums and Buzzes

It is very easy to get hums and buzzes in your audio, especially from things like guitar amps. Try to avoid them during the recording process. Check your cables, experiment with different proximities and angles of the guitar in relation to the amp. Use balanced cables where possible, and avoid long runs of unbalanced cables. Also try to avoid having power cables and transformers near audio cables. Keep them at least a foot (30 cm) apart, and if they must cross, try to do it at right angles, and better still sit the power cable on something above the audio cables so they don't touch. Hums can be dealt with by some programmes and plugins such as The Waves X-Hum, but it is very difficult to remove buzzes from a recording.

Hiss and white Noise

Hiss is often a result of incorrect gain settings when recording. Maybe you had too small a signal, and the gain set too high. Perhaps you just have a noisy piece of equipment. Whatever the cause, use your de-noiser, de-hisser or noise removal plug-in, or program function before you compress or otherwise tamper with the audio. If you compress before you use noise removal, you will likely be boosting the noise along with the signal, but if you remove the noise first, the compressor won't find any noise to boost! Higher settings on these programs or gadgets often seriously mess with the signal and change the tone considerably. Try to avoid using them if at all possible, try mild settings first

Clicks and crackles

These problems can be a real bane of digital recordings at times. Some suggestions I can offer in avoiding them are as follows:

Defragment your hard drives to ensure the computer runs fast enough to keep up with the demands.

Adjust your program's buffers to avoid clicks

Use the latest drivers for your software

Use cross fades at edit points

Avoid clipping any equipments inputs while recording

If you are unfortunate enough to get clicks in your mixes, use your program's or plug-in de-clicker, but try to get as close to the area of the waveform as possible by cutting the region either side of the problem area, and selecting just that small area of the region for processing. Sometimes you can manually remove clicks by re-drawing a wave, with the pencil tool supplied in the software.

Using a de-esser

Here is a useful tip I discovered a while back. Sometimes I found it hard to get enough headroom in my vocal recordings to cut through a mix. There would often be short unexpected transient peaks that would shoot way above the rest of the signal, and into the red zone. Now, I could have used a compressor or a brick wall limiter to squash these hard, but I discovered that they are often just centered around one frequency band, and most likely in the 6-8 khz range where the sibilance of an "S" sound resides. By using a de-esser, I not only made the "S" sounds less harsh, and more pleasing to the ear, but I found I now had much less problem with the transient peaks, and was able to boost the level of the vocals up more before clipping occurred. Kind of like killing two birds with one stone.

This technique is not only limited to vocals either, but frequency dependent compression using a side chain or a de-esser can also be applied the same way to such instruments as an acoustic guitar.

Well I hope these tips are of some use to you in the mixing, mastering or post-production process. Have fun with your audio! .
Rocksure Soundz
(c)April 2011

Using Compressors when recording

I originally wote this article back in 2004 and it published it at this page on the Koretz Music website. I have reprinted the original article here, but added some extra information at the bottom of the post after the copyright 2004 text.

Compression Basics
In this article I want to present a broad but brief overview on the use of compressors in the recording of sound, be it music, the human voice or general sound effects. The compressor is a great tool to have in your arsenal as a sound engineer, but it is one of those that can either make or break your audio recordings. Used right, it can greatly improve your tracks, used wrong it can ruin the audio and make it sound squashed and muffled. The human ear can detect sounds in a wide dynamic range - from a quiet whisper for example to a huge explosion. However, our recording and playback equipment has physical limitations which mean we have to squash or compress these sounds into a smaller dynamic range in order to reproduce them. So here I will just give you some tips to help you get started in the use of compressors and limiters in controlling your sound's dynamics.

Description
Okay, so here we go. First up I will give you a quick outline on what a compressor or limiter does. Right oh, a Limiter abruptly cuts off the dynamics of the signal above a predefined threshold, while by contrast, a compressor gradually attenuates the signal above the predefined threshold.
Often times the same device can be used as a compressor or a limiter, or it may have two sections, a compressor section and a limiter section. A compressor is generally used to control the dynamics within the normal program level of the material, while a limiter usually has it's threshold set above the average signal level, so that it just catches peaks that rise above those levels. With digital equipment this is often necessary to prevent "overs" or digital clipping if the level exceeds 0 db. While there are a number of different types of compressors and limiters available , either as hardware or software, most have some or all of the following adjustable features available for user control, though sometimes they are given different names.

Threshold
The threshold control enables you to set a point above which the program material begins to have it's level reduced in amplitude. Therefore, any signal which exceeds the threshold point in it's volume will begin to be attenuated by an amount determined by the ratio control.

Ratio
Once a signal exceeds the threshold ,the way the ratio control is set will determine how "Hard" the signal above the threshold is attenuated. For instance, a ratio of 1:1 would mean that no attenuation is occurring at all, at 2:1 the signal above the threshold will have it's gain reduced.
At this ratio, for every decibel of signal gain that happens, only half a decibel of increased output will actually occur. Two decibels signal gain would result in one decibel of increased output, ten db gain would result in a 5 db increase etc.
The higher the ratio , the harder the gain reduction. In practice, any ratio above 8:1 is in effect pretty close to being labelled as a limiter in my book, as not much extra signal gain will be output beyond the threshold at that setting.

Attack Control
The attack control determines how quickly the compressor will react to signals that exceed the threshold. You might think that instant response would be ideal, but in practice a slight delay in time between the signal increase and the onset of compression often results in a more musical and less muffled sound. An example here would be say a kick drum, where you want to here the quick attack transient of the beater to give it the punch, before the resonance of the drum occurs, and the signal is squashed. Generally attack times are set somewhere between 0. 1 to 100milliseconds or so. With a limiter though, you generally want a very quick attack to catch the peaks before they slam into, and over your precious 0db digital recording levels and cause distortion, and maybe equipment damage.

Release Control
The release control determines the amount of time it takes for the signal to return to it's normal gain once the material is safely below the threshold level again. Release can usually be set from times that are near instantaneous, to times of 5 secs or more.

Output or Gain Makeup Control
When we compress or limit a signal, the result is an overall level output that is lower than the original program material, and we normally use this control to bring the level back up to, or beyond the original level, so long as we don't take it above the level where clipping occurs.

Setting the Controls
So, having given a brief outline of what the controls do, I will now attempt to give you a few guidelines as to how to use them. There are many ways that compressors can be used for interesting and creative effects, but that is beyond the scope of this article. Remember it is titled "Compression Basics" so I will try to keep it that way here!

Compression
Okay, begin with a ratio of between 3:1 and 6:1 for most natural sounding results, and gradually lower the threshold till you acheive the desired amount of gain reduction. Remember, the lower the threshold, the more the signal will be compressed. For voices and non percussive instruments try setting the attack time moderately fast, let's say at around 10 milliseconds, and the release at between 0. 5- 1 sec. If the attack time is too fast your, S's and T's will disappear, and dynamic distortion is also more likely to occur at fast attack times, particularly if coupled with a low threshold. If the attack is too slow, the S's and T's may actually stand out too much. If the release is too short the level will appear to fluctuate and pump, as the compression goes in and out, but if too long, pumping side affects will be noticeable when loud passages are followed by quieter ones.
Percussive instruments can be treated by using a faster attack time, to catch the initial transients, or alternatively use of a slower attack time will let the transients through while catching the body of the sound.

Levelling
When set up for levelling, a compressor will keep the overall sound at a more uniform level, while not affecting the short term peaks in the material. In order to acheive this, you would set the threshold relatively low (affecting more of the sound) , the attack time relatively slow, and the release time slow as well.

Peak Limiting
To set up as a peak limiter use a high ratio (8:1 or more), a high threshold, and a fast attack and fast release time.

Programme Limiting
If you desire only to limit the average signal level of the audio and not to catch the peaks then set the attack time at 20 ms or slower.

Makeup Gain
Once you have done your compressing, use the output control to make the gain up so that you raise the signal level back up to, or beyond the unprocessed sound, just be careful not to push it beyond the clip level though.

Using a compressor and limiter together
A very common thing to do is to place a compressor first in a signal chain to raise the overall level of the programme material, and follow it with a limiter set only to catch the peaks that the compressor misses, and thereby keep the level close to the zero decibel mark However, another way is to actually reverse this, and place a limiter first in the signal chain, and then compress the remaining material. This can often avoid holes being punched in the sound by the interaction of these two things, but hey, discussing that is again getting beyond the scope of this article.

Avoiding Side Effects
Like taking medicine, there can be side affects to compression. To avoid them. . . . . . don't overdo it. Too much of a good thing can ruin your health ! Sometimes less is more, so be judicious and don't slam everything to bits with "over compression". Things can sound really lifeless if you do, or can get fatiguing to listen to if there are no dynamics present. If you get distortion in the audio. . . back off on the attack time a little, or set the threshold higher, and the ratio lower. If your audio starts sounding muffled, then slow down the attack time and lower the ratio to let some of the transients pass through before the "squashing" occurs.
Well that's about enough I think . . . this "basics" article is looking pretty darn long to me. AND I'm getting tired hands from typing! Have fun squashing your sound!!!!!!!!! © copyright 2004

Extra information (© copyright 2011)

Should I compress during recording or during mixdown?

With the prevalence of high sample rates and 24 bit recording, is it really necessary to compress during the recording process,or should you just wait till mixdown? Afterall you can now get very good results with much quieter recording due to low noise floor with and not need to compress to get a good signal-to-noise ratio using 24 bit. Well, no it's not generally a necessity to compress during recording, and you can do the compressing later at mixdown, but don't forget that compressors can be a very good tone shaping tool during the recording process, and individual hardware compressors all add their own character to the sound, even if not actually doing much compressing. My view is this: don't be afraid to use compressors during recording, but remember that you are printing the compressed sound if you do it this way, so err on the side of caution. Compress lightly. You can always add more compression later, but you can't take away the compression once it is there.

What does Royalty-free music mean?

Royalty Free Music. What does royalty free mean?
　Many people who see the proliferation of royalty-free sites are confused by what this actually means. I will attempt here to to give a brief , but hopefully easy to understand explanation as to what the term royalty-free acutually means, and how to apply it's use.
"Royalty free" as it applies to music, is the right to use music without having to pay a royalty fee. The use of certain intellectual properties require users to pay a royalty fee to the owner, creator or copyright holder of that property.
Royalty-free does not mean the product is free or public domain, it just means you don't have to pay royalties.
Royalty free or buyout music, and it's use is governed by the the licensing agreement that you agree to when you obtain the music. The license type will determine how you may use the music, whether you can use it in a commercial production or in a personal or non-commercial production, and what the limitations are. License agreements vary considerably. It always pays to read the fine print.When you purchase a piece of music, you buy the rights to synchronize the song with your productions. The music and it's copyright will however, always belong to the copyright holder of the product. Generally you pay a one-off fee to purchase the use of the music, and then no more fees are required. AS long as you abide by the conditions of the license you should have no trouble.
So what are royalties?

Royalties are fees that are paid to rights owners (such as record labels, publishers, composers and artists) for the use of their work. The money paid to composers by publishers for the right to use their songs, is usually determined by an agreed percentage of the price per copy sold. Often only a small percentage of the continuing sales of the music goes to the composer or artist. Payments received for use of or the right to use intellectual rights are called royalties.

What about cue sheets and broadcast royalties?
One thing that can be a confusing issue is royalties for the broadcast of shows on television that use music ( that may or may not be licensed as "royalty-free" music). Television stations pay royalties to the Performing Rights Societies for the right to broadcast music in their shows. Music usage in Television and radio show broadcasts are tracked by documents known as cue sheets. It is important for cue sheets to be submitted so that the composer of the songs receives the royalties from the broadcasting of their music. The broadcasters still pay their royalties to the performance rights societies regardless of whether or not a Cue Sheet is submitted. If no cue sheet has been submitted, the royalty money paid does not go to the person who wrote the music. Instead, it goes into a general fund, which is then distributed to whomever is topping the charts at the time. The composer of the music is entitled to the broadcast royalties accumulated by the use of their music in the broadcast, but this cannot be tracked unless cue sheets have been submitted. This is why most sellers of royalty-free music still require cue sheets be submitted when the music is to be used in a broadcast show.

What is a Cue Sheet?

A cue sheet is a document that lists all the music used in a project, allowing Performance Rights organizations like APRA, BMI, ASCAP, and SESAC (and others worldwide) to properly compensate composers for performances of their work. Cue sheets detail the length of the music used, the track's title, the composer and publisher of the music, and where it was used in the film. It also has some general information about the project. Broadcasters such as TV stations, Radio stations, etc are responsible for paying performance fees to the performing rights societies (PRS). Then the PRS pay the composers. Broadcasters always pay a regular fee to the PRS. As a purchaser of royalty-free music you only need to report what music was used, via the cue sheet system. The composer is then able get their portion of that fee. As the music license purchaser (you) are not required to make any payments to PRS. Webcasts do not require the submission of cue sheets.

To see an example of a Royalty-free music license　visit the Rocksure Soundz.com licensing page.