Compression - The sound engineer’s view

Tuesday, February 26th, 2008

In our daily work as professional sound engineers at Hoff mastering, the number of audio files we receive from our clients that have already been ‘compressed’ ( ahem… inappropriately …if we may say so) is substantial.We realize that it is often difficult for non-professionals to realize that good pieces can sound real flat if it is not applied properly. This article presents a view of compression starting from a few basic principles, which reflects the way we view the answer to the question : how much compression you can apply yourself, and what you should leave for the boys in our lab ?

Compression, or reduction of the audio dynamic range, is essentially the art of making louder parts of a composition appear softer, and conversely, the softer parts appear louder. On one hand it maintains maximum loudness (would anyone buy an uncompressed piece by Wagner ?) and maintains higher average loudness, but what if we over-compress an average voice recoding with a dynamic range of only 10dB ? The voice sounds abnormally dull, and flat. Our experience over the years show that highly compressed styles (like the disco) soon become monotonous, and fade out of style. Human ears have their own inbuilt compressor systems inside our inner ears (cochleas), and anything that sounds ‘unphysiologically’ compressed appears unappealing. Good compressors have designer circuits to emulate the human ear – by varying the amount of compression according to the average signal level. Excellent compressors have a second layer of circuits : to clamp those peaks in loudness which go beyond a defined level of the average signal level. With human beings living longer, and the higher incidence of hearing impairment seen in all populations, compressing is becoming a fine art – too much and your CD is exceedingly dull, too little and it can get uncomfortable very fast. In the coming years, a progressively bigger market in music for the senior generations would demand very careful compression skills.

Compelling reasons to avoid ‘overall’ compression without professional help

Overall compression can be easy to apply to your mixes by the flick of a mouse button, but the results can often be unsatisfactory ! Consider this :

Professional mastering services have much more sophisticated compressors with defined parameters about attack, ratio, and release times exactly right for your type of music. Experience teaches a lot about the compromises, advantages and disadvantages of overall compression.

Each tune needs a slightly different approach to compression within the same album, unless for some reason you want all of them to sound uncannily ’similar’ !
Listeners playing CDs intuitively expect all commercial CD albums to play at roughly the same sound level on their music system. Professional mastering services strictly adhere to this ‘industry level’ of compression.

Professional mastering engineers apply mastering on a non-destructive basis – after preserving a ‘master copy’ as a reference that can be returned back to if you are not happy with EQ, compression or anything really. Digital software-based EQ or compression at home is however destructive - resulting in an irreversible loss of digital information with each successive step, called dithering, rounding or truncation. The end product might be quite different from what you expected.

Damage to your sound file by non-professional compression is most often irreversible. Loss of life, loss of transience are familiar to most mastering engineers as examples of "enthusiastic compression". Having said that, folks at Hoff Mastering have few ‘tricks of the trade’ to undo some of the damage.

Compression can be a tool to reach your intended market

The subtleties of compression that we advice our customers on, is based on years of experience on the music industry. For instance, in the case of many thrillers, as the movie progresses, so does the dynamic range of the background music ! Towards the climax, as you might have noticed in Hitchcock’s movies, the sound ranges from a barely audible whisper in one frame to a massive polyphonic crescendo in the next.

Coming to music CDs, good music that people buy to listen to over and over again is often skillfully composed- fast paced and loud, bold and dynamic. ‘Over compression’ can kill the flavor that wide dynamic range imparts – Mozart is still the champ when it comes to playing with dynamic ranges. As we have mentioned before, we believe that disco was killed primarily because of its ‘over compressed’ monotony. We would never let that happen to you !

On the other hand, for furniture music, music for relaxation and Gregorian chants the pattern of compression to achieve a subtle, flowing, rising and falling melody can be a different prospect all together. As experienced mastering engineers, we at aim to bring you the wealth of our experience on compression : by producing music intended to reach its desired audience.

Why reverb is a tool commonly overused

Friday, February 15th, 2008

The reason that you can tell by listening to a recording whether it was sung in the bathroom or a spacious hall is due to the property called reverb. Although during the process of recording there is a certain amount of reverb (that can be minimized through the judicious placement of recording equipment), the skillful mixing of reverb most often adds greatly to the aesthetic quality of music.

Digitally speaking, despite the many types of reverb - hall, room, plate, ambience, gated and reverse, there are only a few parameters which are tweaked when adding reverb to a file to add a multitude of effects :

Pre-Delay Time

Setting the time between the impulse and the early first reflection. This creates an illusion of room size and helps to keep the original sound distinct from the reverb.

High Frequency Damping

It simulates the absorptive properties (both air and surface) of a real space by making the high frequency decay time shorter than the overall decay time. Some reverbs also allow low frequency damping to simulate spaces which reflect mainly high frequency sounds. These effects can also be combined to produce other subtle effects.

Overall Decay Time

Sets the value of RT60 (the time the reverb takes to die away by 60dB after the end of the initial sound is produced). The longer the decay time the larger and more reflective the artificial space seems to be.

Early Reflection Pattern and Level

Most reverb settings give the user a choice of preset patterns (rooms, halls, plates, chambers, ambiences etc.) as mentioned earlier. The greater the spacing in between these individual reflections the larger the room ‘feels’. Some reverbs allow the user to alter the level and spacing of the early reflections.

We are accustomed to listening to music played indoors in buildings with its distinct ‘natural’ acoustic properties. The natural reverberation affects all the sounds in a building. Every sound produced in the building has the same reverb. Pop music producers tend to use a variety of reverbs within the same mix; even each instrument can have a different amount of reverb while multiple reverb units can have different type of reverb. However, like most effects, reverb works best when it is used sparingly. It is easy to get carried away and drown everything in multiple spasms of reverb, but this obscures the song’s natural feel, and prevents it from ‘breathing’. If you listen to some well produced music you’ll find that the use of reverb is light and restrained. When long reverb tails are used, the music will have been specially arranged to provide space for them.

Reverberation can be applied to any sound, but in modern pop production it is most often used on percussion and vocals. Remember, the chief rule of using reverb is…don’t overuse it!
As a producer, it is vital to put a special emphasis on the skillful use of reverb to vocal performances. This is one area we believe, that reverb can achieve a virtually complete makeover. This is the very reason why so many of us find ourselves so compellingly different when we sing in our bathrooms. However there are quite a few elements to consider :

• Long reverbs reduce intelligibility, and it is vitally important that the words are clear and distinct
• Bright reverbs can sound exciting, but tend to over-emphasize sibilance (the use of s, sh, ch, z, j sounds in our speech).
• Too little reverb in a vocal composition can make the vocal seem ‘plastered on to’ the backing track, rather than sounding comfortable in the background mix
• Long reverb can sound quite attractive, but tends to fill the spaces in the music; the spaces are just as important as in music as in relationships
• Adding reverb creates the illusion of distance, and normally we want the singer to be at the foreground, (unless we are deliberately mixing backing vocals…)
• For intimate vocals, short settings (less that 1 sec) with a fairly bright reverb are better. Also a high level of early reflections are important.
• A ‘drawn out’, more prolonged effect can be produced by introducing more pre-delay to separate the original sound and the reverb

At times, we at Hoff Mastering applies carefully selected, high quality mastering reverb to submitted mixes. This can create a totality of illusion of space around the music, and combat that dry, ‘glued to the speakers’ sound.

Reverb is very easy to overuse, and although mixing professionals are used to the technique of introducing it even in multiple complicated segments, its best to keep a master copy in the event you want the changes reversed. Once already mixed before submission for professional mastering its practically impossible undo it, even by us !