Filter fatigue in the home studio

...continued from front page

Consider the sonic characteristics of acoustic instruments. All acoustic instruments have a resonant frequency. When played, they exhibit a "formant", which is the frequency range in which their signal increases in volume. It's the reason that brass and reed and string instruments come in different sizes. It's also the reason that sounds captured on digital sampling devices exhibit the "chipmunk" effect when the sound is played at a pitch other than its true recorded pitch. Ideally, the formant shouldn't shift even though the frequency of the note does. Until recently, samplers usually (erroneously) shifted the formant as well as the pitch. Chipmunk city. Nowadays, elaborate software algorithms on many recent samplers aims to correct this artefact.

Consider the sonic characteristics of recording rooms. All rooms have, in essence, a "formant" - a frequency response curve - which is why the equalizer is such an integral part of a PA. The equalizer compensates for the volume of sounds emphasized by resonant frequencies within the room, owing largely to the lack of enough bean-bags to seat enough audio engineers to correct the problem.

Consider the sonic characteristics of synthesizers. Here, the sound source is usually processed through a filter of some sort. The filter's frequency response curve is not unlike the formant on an acoustic instrument. If you record a multi-track recording of a single synthesizer (think of Wendy Carlos' "Switched on Bach" or "Switched on Brandenburgs") each additional track will have a similar formant which is amplified due to the aggregated sound of the same filters emphasizing the same frequencies across all recorded parts. And to the listener this converts into "filter fatigue" - the tendency for humans to dislike over-emphasis or under-emphasis of certain frequencies within their source material for extended periods. It's why lift music drives people nuts (all the great lift operators have disappeared - ever asked yourself "why?").

In late model multi-timbral synthesizers (where each voice has an identical output filter), the performance modes exhibit the same effect. The cut-off frequency on every part is identical. And you get an overall sound which characterizes the overuse of the one box. It's a phenomenon which, as musicians, we encounter with the"tone" of music time and time again.

Even digital gear can exhibit filter characteristics. The ubiquitous Yamaha DX-7 synthesizer was popular for its crystalline electric piano sounds and for its novel "Fourier addition" method of synthesis. It also exhibited its uncanny ability to prevent six sine wave generators obscuring vocal parts. Its formant was audibly appealing, and inherently useful. The fact that it would never cut through a live mix against a real brass section seemed almost irrelevant. At least you'd be able to go home from any Andrew Lloyd Wobbly theatre show humming the lighting.

Filters are present in the most recent digital recording gear. The advent of modeling software to "replicate" the qualities of digital devices effectively deals in frequencies so high that only your canine will really be able to detect the quality. These late model digital devices are simply trying to be analogue - in mathematical terms, by successive approximations. I would suggest that this digital approximation itself has the characteristics of a formant. Further, a good audio engineer can detect this even without their sunglasses on.

All sound when digitally encoded is turned into a stream of numbers which are manipulated in a very complicated way - and I see some of you staring lovingly at your HP-41C calculator with its shining new 19 inch rack mountings...but I digress. A recorded work is always going to be digitally different than the analog source signal it attempts to recreate. But is the difference *really* noticeable? Record one or two tracks using exactly the same device with the same settings, even with different modeled effects programs, and it begins to sound as if it's been processed with the same digital spanner.

So the unfair advantage of analog is that it often demands multiple (and different) units being employed for any recorded material. And this variety inherently prevents "filter fatigue", which will be detectable on your program source. And maybe your shining new Invidium Q236 Cosmic Disintegrator (or software plugin equivalent), with its arsenal of replicated analog effects, will be useful on one or two tracks. But will the average mortal resist the temptation to use it across every individual track? Where's the "kaboom"?

Filter fatigue. There is only one cure. It will mean we need to buy lots of different outboard recording effects devices and lots of instruments to avoid it. I detect that this may come as a nasty shock to some of you. Somehow, I suspect that shopping for new (and old) gear is a problem that won't ever vaporize in the music and audio industry. The vicarious thrill of shopping will never be replaced by a software plug-in. Unless you suffer from "shopping fatigue".

Home

Spiritual intoxication

...continued from front page

In the presence of such a remarkable piece of work one hesitates to criticize but your "Dues" writer does have responsibilities. If a reviewer asserts opinions regarding programming structure it would be at least useful if the basis on which the opinion was arrived at was also presented. The statement that the "sheer volume of the orchestra was impressive" conveys very little (in fact one would find it difficult to imagine a context within which sheer volume was not impressive). The application of the adjective "bravura" to a range of the voice is a misuse of a specific term. Lest the above comments be considered as carping criticism you "Dues" writer feels that the following must be stated. It is with unbounded joy that one realises that sweet, mellifluous, prose springing from the very heart of the impassioned and spiritually intoxicated soul of the sensitive reviewer still has the possibility of moving us to the very core of our being because newspapers will still print it. (The original review can be found at: http://www.theage.com.au
/articles/2004/03/02/1078191306635.html)

Home