North Suburban HAMMOND ORGAN Society

Digital signal processors

A trio of processors. The top one is a noise gate. It eliminates unwanted background noise from a musical instrument. If, which is a common problem with some older electronic organs, there is some unwanted background noise such as hissing or power humming, these noises could intrude during silent pauses in a recording.

The noise gate allows signals that are above a preset level to pass through but blocks all signals below the preset level. The theory is that during an actual musical performance, the musical signal will drown out the background noise. Thus the noise gate lets the louder musical signals go through, but in the absence of the musical signal, it cuts out completely, thus preventing background hum or tube hiss, for example from intruding.

The other two units are multi-function effects processors, capable of producing reverberation, echo delay, chorus, tremolo, pitch changing and other related effects. As such, they can also transform mono signals to stereo signals and are invaluable for modern music production. DSPs make recordings sound big and full, and can add concert hall acoustics to music that is recorded in acoustically "dead" studios.

The invention of these digital signal processors, or effects processors has, along with MIDI and digital audio, made it possible to set up a recording studio in a den, basement, or even a bedroom or small office and turn out recordings that rival some of the best commercial recordings in sound quality. Of course, there's no substitute for genuine musicianship. DSP cannot make a beginner sound like a pro but it can make a good recording sound great.

One of our primary purposes for DSP is the simulation of Leslie speaker effects. In a real Leslie, with its rotating speakers, horns or baffles, the resulting effect, which they call tremolo, is extremely complex. The motion of the sound source creates doppler pitch changes which results in a complex vibrato. The rotation also develops amplitude-varying tremolo. As the speaker rotates towards the listener, he hears an apparent pitch rise. As the speaker rotates farther and rotates away from the listener, he hears an apparent pitch decrease. At the same time as the speaker rotates away from the listener, it may be rotating towards an opposite wall, thus the sound wave reflected from the wall will have an apparent pitch increase. The effects of Leslie tremolo encompass both vibrato and tremolo and are infinitely complex. Anyone who has ever heard a Leslie speaker knows how big, full and rich sounding the result can be.

Digital signal processing can reproduce these effects quite accurately and develop a complex resulting vibrato/tremolo that very closely approximates the sound of a Leslie speaker. When simulated Leslie tremolo is heard through a pair of stereo speakers, the result is essentially indistinguishable from the real thing. The beauty of this is that we can record this signal directly without the use of microphones which guarantees a really good and accurate recording free from any standing wave patterns in the room or extraneous background noises. Furthermore, we have a great deal of control over the digital Leslie simulation. We can change its complexity, intensity, and speed. We can make it choppy with a strong tremolo component, or we can take the tremolo out entirely and just have a very complex vibrato. A final advantage is that today, a digital effects processor that can do all of this is considerably less expensive than a real Leslie speaker.

Is DSP a complete substitute for a real Leslie? Probably not. In a real room with a real Leslie, the effect is more elaborate than a simulation. But on a stereo recording, the results can be indistinguishable.

In addition to pitch vibrato of the usual type, it's possible to create several other types of vibrato. For example, if we take two signals, and change one of them by arithmetically adding, say, six cycles to every component of the signal and then combine this with the original signal, we'll get a so-called heterodyne vibrato which is quite different from normal vibrato but extremely useful for some effects. If you are simulating a vibraharp either on an electronic organ or by means of a MIDI instrument, a heterodyne vibrato sounds much more like the type of unique tremolo-vibrato effect that you get with a real vibraharp.

Still another effect which I mentioned on a previous page is a stereo tremolo where the signal is split into two components, each of which is given a tremolo, but when the signal in one channel increases, the signal in the other decreases. This creates a complex phase-changing vibrato which is different from standard vibrato and different from a heterodyne vibrato. And of course, we can combine all of these effects.

In a real hall, there are many subtle components to the hall acoustics or reverb. These include early reflections (from nearby surfaces) pre-delay (a few milliseconds of almost total silence until the early reflections are heard). Then there is the subsequent reverb which comes back from the interior surfaces of the concert hall. All of these mix and form a very complex reverberation which is so extremely important for music. Digital processors simulate all of these effects which is why the reverbs that good ones generate sound completely natural. And, best of all, every one of these variables is controllable by the recording engineer.


 Previous Page   Page 8.    Next page