In both cases, it was very easy to bring the computer to a screeching halt (especially when using long reverb times). I tested it in several VST hosts on both a 400- and a 500-MHz G4.
The only other element of the interface is a set of two buttons letting the user choose between "No latency, high processor load" and "High latency, low processor load." Altiverb is highly processor intensive. There are also knobs for the wet and dry components of the mix (which function exactly how one would expect), as well as a pre-delay knob (which allows one to delay either the wet or the dry signal). Changing the settings on this knob require some "thought" on Altiverb's part, and it may take a few seconds to compute the new value. Anything less than 100% puts a decaying envelope on the original impulse response. A setting of 100% lasts the length of the original impulse response. The largest knob on the interface is devoted to the actual reverberation time. The interface displays a picture of the recording location, and also contains information about the gear used in the recording. In addition, there are plenty more impulse responses available at the Audio Ease Web site, including many produced by users. Altiverb ships with a number of excellent impulse responses, sampled from such places as the acoustically outstanding Concertgebouw in Amsterdam. The actual impulse response is selected from a drop-down menu in the middle of the unit. The controls on the interface are few, and are easy to understand. Like most plug-ins these days, the design emulates a rack-mount device. The user interface is quite simple (see Figure 1). It operates on the Macintosh platform only, and functions in MAS, HTDM, RTAS, and VST hosts, supporting mono, stereo, and quadraphonic output. Audio Ease Altiverb is a great sounding, powerful, easy-to-use reverberation plug-in which is based on convolution. In principle, the results should be essentially the same as a recording made in the exact location where the impulse response was recorded. This recording is called an "impulse response." Convolution-based reverberation convolves (which is to say, multiplies in the frequency domain) an impulse response with whatever sound source is to be reverberated. Either a starter pistol is fired, or a sine wave is swept across a broad range of frequencies, and a recording is made of the results. This second approach is called convolution, and is not based on simulation, but rather on a kind of sampling of the acoustical properties of a real place. There is, however, an alternative approach, which produces more "natural" results. There is no shortage of reverberation plug-ins currently available, and most of them are based on algorithmic simulations of naturally occurring reverberation (based on delay lines, all-pass filters, etc.).
0 kommentar(er)
