http:^^www.cs.cornell.edu^info^people^szuwen^cs631^project.html
来自「This data set contains WWW-pages collect」· HTML 代码 · 共 207 行
HTML
207 行
MIME-Version: 1.0
Server: CERN/3.0
Date: Sunday, 01-Dec-96 19:07:57 GMT
Content-Type: text/html
Content-Length: 7129
Last-Modified: Friday, 08-Dec-95 01:33:07 GMT
<html><title>Sound Editor</title><center><h1>Sound Editor</h1><!WA0><!WA0><!WA0><!WA0><a href="http://www.cs.cornell.edu/Info/People/szuwen/szuwen.html">Szu-Wen (Steven) Huang</a>, Jing (Vince) Li, Ting-Chun Janet Liu<hr></center><body><h2>Introduction</h2>This project aims to implement a sampling resolution-independent digitalaudio engineering tool that can be used to create moderately complexspecial effects. The project should sport an easy-to-use and fairly portable interface across various Unix/X platforms. The program shouldwork reasonably well on all kinds of audio samples, from low-fidelityvoice to CD quality music. Ideally, this program should be able toutilize commonly-available hardware features such as mono/stereo audioplayback and recording.<p><hr><h2>Deliverables</h2><ol><li>Working software<li><!WA1><!WA1><!WA1><!WA1><a href="http://www.cs.cornell.edu/Info/People/szuwen/cs631/final.html">Technical documentation</a></ol><hr><h2>Milestones</h2><ol><li>Algorithms research<li>Hardware support and standards research<li>User interface in Tcl/Tk and C<li>Time domain functions<li>Frequency domain related mathematics research<li>Frequency domain functions</ol><hr><h2>General Operations</h2><strong>Playback</strong><br>The program should be able to playback the samples being edited forpreviewing purposes. This function does not aim to interface with adevice capable of production-quality audio synthesis.<p><strong>Pause</strong><br>The program should be able to stop in the middle of a playback.<p><strong>Rewind</strong><br>The program should be able to return to a certain sample earlier intime and resume playback from there.<p><strong>Fast Forward</strong><br>The program should be able to skip ahead to a certain sample later intime and resume playback from there.<p><strong>Seek</strong><br>The program should be able to start playback at any certain sample inthe entire sound file, indexed by either playing time (seconds) orsample number.<p><strong>Record</strong> (Optional)<br>If hardware support permits, this program should be able to digitizesamples from an external audio source. In general, however, thisprogram should assume that samples are digitized by another dedicatedplatform because of quality considerations.<p><strong>Volume Control</strong><br>This program should have the ability to adjust the output volume ofthe playback function appropriately. If possible, a decibel notationshould be employed instead of arbitrary volume levels such as 1 to 10.<p><strong>Balance</strong> (Optional)<br>This program should be able to adjust the ratio between playback outputof the left and right channels on stereo hardware.<p><strong>File Operations</strong> (Optional)<br>Currently, the Sun Audio Format (.au) is deemed most appropriate forreasons of portability and ease of programming. If time permits, supportshould be provided to output resulting samples in the Microsoft WindowsWave (.WAV) format, Creative Labs Voice (.VOC) format, and others.<p><hr><h2>Monitors</h2>The following functions do not alter the input samples in any way, butdisplay them in a manner that might be visually useful (and pretty) tothe audio engineer.<p><strong>Raw Waveform</strong><br>The program should be able to display the pulse-code modulation (PCM)format digitized samples as a waveform in time domain. Multiple filescan be open at the same time to allow simple edit operations such as<em>Cut</em>, <em>Copy</em>, <em>Paste</em>, and some others that usersof commercial software would come to expect.<p><strong>Spectrum</strong><br>The program should display a frequency spectrum synchronized with theplayback to show the distribution of frequencies. Right now, we aresplit between the conventional boombox style of discrete frequencyranges and quantized intensity levels (dancing LED bars), or a more"scientific" high-resolution spectrum display possible only on videomonitors. Suggestions are welcome.<p><strong>Time/Frequency/Amplitude Histogram</strong><br>The program should display a time domain frequency/amplitude graph suchthat time progresses along the x-axis (towards the right side), frequencyranges are plotted along the z-axis ("in" and "out"), and amplitudeplotted along the y-axis (going up). This display would be useful invisualizing amplitude distributions over time.<p><hr><h2>Time Domain Operations</h2>These functions modify the input waveform in the time domain, and do notrequire a <em>Fast Fourier Transform</em> (FFT) or its inverse to operate.All time domain operations can be performed on the entire file or someselected subset. These operations should be resolution-independent.<p><strong>Echo</strong><br>This function digitally simulates an echo. It will be implemented by mixingseveral copies of the sample in decreasing strength with a correspondingshift in time<p><strong>Fade</strong><br>These two function incrementally increase or decrease volume over a rangein time.<p><strong>Mix</strong><br>This function is the digital simulation of mixing two input sources together,and will be implemented with some form of weighted averaging of the samples.<p><strong>Backmasking</strong><br>This function reverses the waveform in time, such that the last input willbecome the first input, and vice versa. This function is useful in includingSatanic messages into rock songs advocating suicide.<p><strong>Silence</strong><br>This function creates a range of silence.<p><strong>Amplitude Scaling</strong><br>This function is used to modify the amplitude of a range of samples toeffect an increase or decrease in volume.<p><strong>Resampling</strong><br>This function allows an input waveform to be <em>upsampled</em> to a highersampling frequency or <em>downsampled</em> to a lower. Upsampling is probablyaccomplished by interpolation, and downsampling by tossing away samples.<p><hr><h2>Frequency Domain Operations</h2>The following operations require processing in the frequency domain.<p><strong>Pitch Scaling</strong><br>This function can be used to effect a pitch shifting without affectingplayback speed, useful in making people talk like chipmunks.<p><strong>Graphic Equalizer</strong><br>This function can be used to enhance or inhibit certain frequency ranges,for example, to reduce noise. Two proposals are being reviewed here, thefirst being the traditional boombox graphic equalizer with sliders, ora free-form tool that will automatically curve-fit certain selected pointsand modify the spectrum accordingly. Suggestions welcome.<p><strong>Addition and Subtraction</strong><br>This function allows the addition or subtraction of a spectrum withanother. This is possibly useful in noise elimination.<p><strong>Morphing</strong><br>This function can be used to transform one spectrum into anotherincrementally over time.<p><hr><address>Last updated: December 7, 1995<br>Maintained by: <!WA2><!WA2><!WA2><!WA2><a href="mailto:szuwen@cs.cornell.edu">Steven</a>,<!WA3><!WA3><!WA3><!WA3><a href="mailto:vince@cs.cornell.edu>Vince</a>, and<a href="mailto:janet@cs.cornell.edu>Janet</a>.</address>⌨️ 快捷键说明
复制代码Ctrl + C
搜索代码Ctrl + F
全屏模式F11
增大字号Ctrl + =
减小字号Ctrl + -
显示快捷键?