Technology Institute for music educators ti: me course 2a Advanced Sequencing, Second Edition
Download 3.6 Mb.
|
- Navigate this page:
- Appendix K: Bounce to disk
|
Time-Based Signal Processing The “aux” send and return structure described in the previous section are primarily used to integrate time-based effects into a mix, including delay, delay effects (like flanging and chorusing), and reverb. These effects are added via a “send” because this method allows multiple tracks to share the same effect. This is both an effective use of computer processing power (critical with reverb devices that are typically processor intensive) and can help to place the tracks in the mix in the same sonic environment. Delay Effects Delays are devices that delay or hold a copy of a signal for user-defined amount of time. When the delay signal is mixed with the unprocessed signal it adds a sense of depth and dimension. Delays can create several types of effects, primarily based on the amount of delay time. Delay effects with a delay time that ranges from 1 to 50 milliseconds are perceived as part of the original signal and not as a discrete repeat or echo. Representative effects include: Flanging: 1-15 milliseconds of delay time, which results in a type of phase cancellation called comb filtering Phase shifting: 1-15 milliseconds of delay time, which results from running the delayed signal through an all-pass filter (a device that passes all frequencies, but just not at the same rate). When the delayed signal is combined with the unprocessed signal, phase cancellation occurs. Doubling: 15-50 milliseconds of delay time, which results in a sense of fullness similar to a vocal track that is recorded twice and then played back simultaneously. Chorusing: 15-50 milliseconds of delay time with pitch modulation. Using the example of a vocalist, the delay and pitch detuning that are part of a chorusing effect create the perception that the performer is more like a choir than a single vocalist. When the delay time is more than 50 milliseconds, the delayed signal is perceived as a discrete event. Echo is the best known of these delay effects. The parameters commonly found in delay effects include: Delay time: in milliseconds Modulation: delay by itself can create a static, uninteresting effect (especially when the delay times are less than 50ms). Modulating the delay time can change the nature of the effect much like a violinist who adds vibrato to a performance. Modulation is added to a delay effect by using a Low Frequency Oscillator (LFO). The LFO oscillates at a very slow rate and is used to control the rate of change in the delay time or other parameter. Rate: The speed at which the LFO causes the delay time to shift Width: The range of drift in the set delay time Feedback (regeneration): The amount of delayed signal that is rerouted to the input of the delay. Increasing the feedback will increase the intensity of the effect or, in the case of echo-like effects, it will increase the number of repeats. With echo, increasing feedback beyond a certain point may create an infinite loop of repeats, and possible distortion. Wet/dry mix: This determines the ratio of processed to unprocessed signal output. Higher percentages will result in the output of more delayed or “wet” signal. When a delay (or reverb) is setup as a “bus” effect, the wet/dry mix is always set to 100% wet. It is not uncommon to find delay effects directly placed on an audio or instrument track. When that is the case, the wet/dry mix parameter on the delay effect is set by “ear” to achieve the desired effect. Repeating delay: One whose delay time is set to seconds or milliseconds. This time may still be modulated by an LFO. Tempo or Tap delay: A delay effect synchronized to the musical tempo of a song, either by entering a rhythmic value, or by tapping the tempo on the computer keyboard key, foot pedal or MIDI controller. Normally this delay time cannot be modulated further. Reverb Reverb is defined as multiple echoes (closely spaced and random) that reflect within an acoustic space. These echoes are packed together so densely that they are not perceived as discrete events but instead as one decaying signal (ambience). This natural effect can be reproduced by a virtual reverb plug-in to add a sense of dimension and warmth to a recording. Typical reverb parameters include: Pre-delay: the time between the arrival of direct sound (no reflections) and the first reflections at the listener Early reflections: The first reflections that arrive at the listener. They reflect a single time off the primary boundaries of an acoustic space (room) before arriving at the listener. Early reflections give the strongest impression of an acoustics space’s dimensions and construction. Delayed reflections: A multitude of reflections that are so closely spaced that they are perceived as one single decaying signal. Mix: the ratio of unprocessed to processed signal. Because reverb plug-ins are typically used in a send and return setup, the output is set to 100% wet. EQ: The construction materials in a room greatly effect the resulting reverb. Spaces with highly reflective surfaces produce brighter, longer reverbs. Those with absorptive surfaces tend to produce warmer (if not duller) reverb. EQ parameters like high frequency roll-off help to simulate different types of spaces.
When a MIDI project is finished, it’s likely that you’ll want to generate a stereo audio file that can be burned to disc or played on an MP3 player. To accomplish this, most DAWs follow a similar process referred to as a bounce to disk. The next few paragraphs will discuss basic “bounce” procedures and parameters including dither, bit rates, file types and file formats. Dither Quantization error is an undesirable byproduct of the process of digitizing an analog audio signal or changing the bit rate of a digital audio signal. Like tape hiss, quantization error is primarily a concern when the audio signal level is soft enough that it doesn’t mask the noise that results from quantization error. Unlike tape hiss, though, this noise is not perceived as an artifact separate from the audio signal. Instead it is perceived as being correlated to the audio signal, and has the harmonic characteristics of a distorted square wave as the least significant (smallest value) bit fluctuates in an ordered pattern between zero and one. In order to solve this problem the signal is dithered, which is a process that causes the signal level to randomly fluctuate at the least significant bit. The result of dithering the signal is that the noise resulting from quantization error is converted into a more palatable pink noise with more of a sine wave shape, and the noise is perceived as decorrelated from the actual audio signal. A common point at which the bit rate of a digital audio signal might be changed is during the bounce to disk process. Since one the biggest improvements in digital audio quality is to work at the highest bit rate possible, DAW users might choose to create 24-bit sessions, even if they are solely using virtual MIDI instruments. In the end, though, the resulting bounce will need to be 16-bit in order to burn it to an audio CD, etc. How dither is added to a bounce depends upon the DAW program. In Pro Tools and Sonar for example, the bounce is dithered by loading a dither plug-in on one of the master track inserts. As a policy, dither should always be the last plug-in insert and the last process performed. That way if any other plug-ins are used on the master track the dithering won’t be “undone.” In some other programs (Logic or Ableton Live), dither is added and configured in the bounce to disk dialog box. Dither parameters are few and not that complicated. First, remember that dither is not required on a bounce where the bit rate is not being changed. If changing the bit rate, the target bit rate will need to be set—most likely to 16-bit. Next, there may be noise shaping options. The noise shaping process attempts to move noise out of the human hearing range. While noise shaping can be an effective tool, it often creates the impression that a bounce has been equalized. So, our recommendation is to bounce a project multiple times using different bounce options and choose the one that either sounds the best or has the most transparent effect on the original multi-track mix. Example: Pro Tools Dither Plug-In 
Download 3.6 Mb. Share with your friends:
|