It╒s difficult to get a good recording of a sound from your Mac╒s ╘Sound╒ control panel. You╒re limited to 10 seconds, and if you don╒t want to cut the beginning or end of the sound off, you have to start recording too early and stop too late. If there are pauses between parts of the sound, you╒ll hear a lot of background noise. And it╒s hard to get the record volume right.
Sound Editor helps with these problems. It also provides the ability to copy and paste portions of a sound, mix sounds together, and create special effects like echoes and backwards sounds.
You can also use it to modify or replace the sounds in most games and applications. Before editing sounds contained in other programs, please make a backup copy!
About Sampled ╘snd ╒ Resources
Please feel free to skip this section if you╒re already familiar with sampled sounds.
There are various formats in which sounds can be stored on the Macintosh, but the most common is the sampled ╘snd ╒ resource. This is the kind of sound that is created when you use the ╘Sound╒ control panel to record a new system beep. Sampled sounds consist of a series of numbers, or samples, that represent the volume, or amplitude, of a sound at a given moment in time. When you record a sound, samples are taken at a regular interval determined by the sampling rate, which is measured in samples per second. When the sound is played back, your Mac╒s speaker moves in a way that corresponds to the individual samples.
When Sound Editor opens a sampled sound, it draws a graph showing time on the horizontal axis and amplitude on the vertical axis. When the display mode is set to 100% (see below), each horizontal pixel represents a single sample. (The
amplitude scale is always 100%, which means that each vertical pixel represents one of the 256 possible sample values.)
Opening Sounds
Sound Editor will open sampled ╘snd ╒ resources contained in the resource fork of any file, including the single ╘snd ╒ resource in double-clickable ╥System Beep╙ sound files (of type ╘sfil╒ or ╘SFIL╒). When you select an application, game, document, or other file in the Open╔ dialog, ╘snd╩╒ Editor looks for sampled sounds in that file. If there is only one, it is automatically opened. If the file contains two or more sounds, you will be presented with a list. Double-clicking on a sound╒s name will cause that sound to be played. Ranges of sounds may be selected by shift-clicking, and individual sounds may be added to the selection by command-clicking. When you click the Open button, each selected sound will be opened in its own window.
Caution: editing sounds that reside in your System suitcase file is not recommended. To edit these sounds, first drag them out of the System suitcase.
View Options
As mentioned above, the amplitude (y-axis) scale of the displayed waveform is always 100%. However, the time scale can be compressed to 50%, 25%, 10%, or to whatever scale will allow the entire sound to be represented in the visible region of its window.
Selecting Portions of a Sound
Selecting portions of a sound for editing should be intuitive ╤ I've tried to follow the word-processing model. The cursor changes to an I-beam when in the display region of the window. Clicking on any part of the sound will set the insertion point, and dragging or shift-clicking in the window will select a range of samples. You can hear the selected portion of the current sound by choosing Play Selection from the Sound menu.
The Edit Menu
The standard Edit menu items work as you would expect: Cut removes the selected range and places it on the clipboard, and Copy places a copy of the selected range on the clipboard. Paste replaces the selected range with the clipboard contents, or inserts the clipboard contents at the insertion point. Clear removes the selected range of samples from the sound. The Delete and Clear keys are equivalent to choosing Clear from the menu, and the first four function keys on an extended keyboard correspond to Undo, Cut, Copy, and Paste, respectively.
Portions of a sound placed on the clipboard are stored in a standard format, and may be pasted into the Scrapbook or other applications╒ documents.
There are a number of non-standard items in the Edit menu. Mix combines the clipboard contents with the sound being edited, in equal proportion, beginning at the insertion point or at the start of the selection. It does not remove or replace the selected range. If the clipboard contents continue past the end of the sound being edited, the sound is lengthened to accomodate them.
To mix two sounds together, open both sounds. Select All of one, Copy it, and then select the window containing the other sound. Set the insertion point to the spot where you want the sound you just copied to start playing, and choose Mix. If you find that one of the sounds is too loud, you can reduce its volume before mixing; see Scale Volume below.
Zero changes the value of all samples in the selection to 128, which is the value at the ╘zero╒ reference line of the y-axis. Reverse causes the selection to play backwards, and Invert changes each sample to its reflection in the x-axis.
The Sound Menu
Record brings up the standard sound-recording dialog. The length of sound that you can record is limited only by available memory. You can increase the amount of memory available to Sound Editor by selecting its icon in the Finder, choosing Get Info from the File menu, and changing the Preferred Size field to a larger number.
If you record a sound (from a microphone or your Mac╒s audio input jack) and click the Save button, a new window is opened for the newly-recorded sound.
Play and Play Selection can be stopped in mid-sound by pressing command-period, as can Play Clipboard in the Edit menu.
Playback Rate allows you to set the rate at which the sampled sound data will be played back, either by choosing from the list of common rates or by entering a rate into the Custom Rate╔ dialog. Choosing a playback rate that is different than the original sampling rate will make the sound play back faster or slower, leading to amusing results.
Resample allows you to expand or compress the sound data by the ratio you select. Expansion is by linear interpolation, and compression is accomplished by a simple averaging of samples. Choosing a ratio of 1:3, for example, will replace each sample in the original sound with three samples: the first will have the same value as the original, and the next two will have values that lie between the values of the first and subsequent samples in the original sound. Choosing a ratio of 2:1 will replace each pair of samples in the original sound with a single sample whose value is their average.
Resampling has a number of uses: compression is valuable in that it allows sounds with little high-frequency content to be represented by fewer samples at a lower sampling rate, resulting in much smaller sounds.
Expansion is necessary when attempting to mix sounds that were recorded at different rates. For example, if you paste part of a sound that was recorded at 11127 Hz into a sound that was recorded at 22255 Hz, the pasted sound will play back at twice its original speed. If this is not the effect you want, resample the 11127 Hz sound at a ratio of 1:2. This will result in a clip that sounds the same at 22255 Hz as the original did at 11127 Hz. (Note that compressing the 22255 Hz sound at 2:1 will also allow the two sounds to be combined properly, but with some loss of high-frequency content.)
Maximize Volume makes a sound as loud as possible without distortion. It does this by finding the sound╒s peak amplitude, and then scaling the entire sound up so that the peak is ▒127.
Scale Volume╔ has at least three uses. By selecting a range of samples at the beginning of a sound and scaling it from 0% to 100%, you can make the sound fade in. Similarly, scaling a selection at the end of a sound from 100% to 0% will cause the sound to fade out. More subtle effects can be achieved by selecting longer ranges and using 10 or 20% as the minimum scale value. Finally, to mix sounds in unequal proportion, scale down the amplitude of the sound that you intend to be quieter before mixing.
Truncate provides an automated method of disposing of the noise that inevitably precedes and follows a sampled sound. The default values will produce good results under most circumstances; for sounds that fade in or out slowly, smaller values may be necessary.
Noise Gate is similar in that it replaces periods of noise between sound events in a sampled sound (between words, for example) with periods of complete silence. The sound is ╘gated╒ off when its amplitude drops below a threshold that you set. The Delay parameter determines how many samples in a row must be below the threshold before they will be considered ╘noise╒.
A Few Techniques
Resampling ratios other that those listed in the Resample menu can be achieved by combining the supported ratios. For example, to compress a sound at 3:2, first expand at 1:2 and then compress at 3:1.
Chorus effects can be achieved in two ways. The most effective technique is to sample the same sound twice, so that you have two very similar but not identical sounds, and then mix them. Where this is not possible, try this: open a sound, Select All, and copy the entire sound to the clipboard. Place the insertion point between 10 and 100 samples from the beginning of the sound and mix it back in with itself. Longer delay times (a few hundred samples at 22255 Hz, fewer at lower rates) start to produce an interesting slap-back echo effect. For really thick results, perform the same operation a few more times.
Strange and wonderful comb filter and static flange effects can be produced by selecting an entire sound, copying it and pasting it into a new window, inverting it, and mixing it back into the original sound delayed by, say, 10 to 100 samples.
Long echo effects are fun, and quite painless. Open a sound, Select All, and copy the entire sound to a new window. Select All again and scale the selection to 30 or 40%, then mix it back in with the original after setting the insertion point to where the first echo should occur. Go back to the scaled sound and scale it again, to 50% or so, and mix it back in with the original at the next echo location. Repeat as often as desired.
If you repeat the above procedure often enough with a short enough delay time, you╒ll end up with a good approximation of a reverb trail. Mix this back in with the original (scaled to 50% or less and delayed by a few 60ths-of-a-second) for a reasonable ambient effect.
Saving Sounds
There are three different ways to save an edited sound. Save updates the resource in the file from which it was opened. If the sound is newly created, or if you choose Save As╔, a standard dialog will appear asking you to name the sound file. It will be saved as a stand-alone, double-clickable sound file. Add To File╔ adds the sound to the resource fork of the file that you select, assigning it a unique ID number to avoid conflicts with existing resources. If the file you select does not have a resource fork, one will be created.
Version History
1.2 ╤ April 7, 1995
Completely rewrote the display and selection routines. Sound Editor now takes advantage of colour displays, and uses the highlighting colour (as set in the Color Control Panel) to represent the selection range. The insertion point now blinks, and some bugs relating to scrolling the display have been fixed.
The editing commands have been sped up significantly, and the menus have been reorganized and some editing commands renamed for greater clarity. Many other miscellaneous bugs have also been corrected.
1.1.x
Unreleased development versions.
1.0 ╤ February 4, 1995
First released as ╘snd ╒ Editor. It turns out that a ResEdit template for editing sampled sounds exists under that name. My apologies to its author.
Future Plans
There are at least two other directions I'd like to explore: automating the generation of delay effects like echo, chorus, flanging and reverb, and offering frequency equalization tools. I╒m going to be working on both problems, and hope to offer regular updates with new and improved tools. (A parametric equalizer with a fully adjustable bandwidth would be a great addition!) Suggestions for additional tools are always welcome.
Acknowledgements and Legal Stuff
Sound Editor is ⌐ 1995 by David Veldhuizen. This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation; either version 2 of the License, or (at your option) any later version.
This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.
You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
Source code is available on request. Please contact the author before including this program on diskette, CD-ROM, or other media compilations or collections. Or at least send me a copy.
Comments, questions, bug reports, and especially compliments are always welcome.
The author may be contacted at:
E-mail:
d.veldhuizen@genie.geis.com
Post:
David Veldhuizen
335 Lonsdale Road, PH3
Toronto, Ontario, Canada
M5P 1R4
The ╘progress thermometer╒ dialogs in this program are a modified version of Joe Zobkiw╒s PD source.
