Raven Viewer

Introduction

Raven Viewer is a Free Tool for Visualizing Audio

  • Play and visualize our digital audio recordings
  • Watch spectrograms, power spectra, and waveforms. These are graphs of frequency over time. Raven Viewer creates these views in real time and scrolls them as the recording plays
  • Play and visualize our digital audio recordings, including video recordings with associated audio
  • Control zoom, playback speed, resolution, & more
  • Inspect low-frequency sounds that are inaudible to humans
  • Experience sound in a new way

Playing versus visualizing audio and video

To use Raven Viewer, select a sound or video to visualize, and then select Raven Viewer. You only need the Adobe Flash plug-in to play sounds and videos. Flash is probably already installed. To play a sound or video, select the play button.

Download

Raven Viewer uses QuickTime and a custom plugin.

  1. Install the Firefox web browser (suggested, but not required)
  2. Install the Apple QuickTime (done, if you have itunes)
  3. Install Raven Viewer for Windows or Mac OSX
  4. Restart your computer

Flash Alternative

If you prefer not to install QuickTime, we also support a Flash based audio/video player.

  1. Install the flash plugin (probably already installed)

Minimum Requirements

Windows
  • QuickTime 7.0.3 or higher
  • Flash 9 or higher
  • Pentium PC
  • 256 MB RAM
  • Windows 2000 or higher
Apple
  • QuickTime 7.0.3 or higher
  • Flash 9 or higher
  • G3 processor (PowerPC & Intel compatible)
  • 256 MB RAM
  • Mac OS X 10.3.9 or later

About the Raven Viewer Plug-In

The Audio Visualization (QSSP) component for QuickTime on which Raven Viewer is based was developed by Totally Hip Technologies, Inc and EdgeState to allow a multimedia author to visually display audio in their movies.

Totally Hip Technologies Inc's mission is to give content developers and producers complete creative control by producing software that empowers users to create, produce and deliver the most engaging and interactive experiences on the Web. Since it was founded in 1995 Totally Hip Technologies Inc. has been consistently developing and marketing award-winning interactive Web development technologies.

More Sound Visualization and Analysis tools from the Bioacoustics Research Program at the Cornell Lab of Ornithology

More Sound Visualization and Analysis tools from other sources.

Details

What is Sound Visualization?

waveformWaveforms: Sounds can only propagate in a medium such as air, water, or solids. When no sound is present, the molecules in the medium all experience the same average pressure. When a sound is created, successive peaks of higher than normal pressure and valleys of lower than normal pressure radiate away from the sound source as waves. A microphone can pick up these passing waves and convert them into an electrical recording of the sound. A graph of how a passing sound varied the pressure in front of the microphone over time is called the waveform of that sound. To the right is an example of the waveform of a wild Mitred Parakeet's call.

The horizontal dashed line is the average pressure in the medium outside the microphone when no sound is present. The height of the peaks and valleys around this line indicates the amplitude of the sound: higher peaks and lower valleys will be heard as a louder sound.

spectrogramSpectrograms: The waveform of the parakeet's call is not steady: successive peaks and valleys have different amplitudes, and some waves are narrower while others are wider. This means that the parakeet's call consists of many different sound frequencies. We can decompose any sound waveform into its component pure frequency components. A graph of the frequency composition of a sound (vertical axis) over time (horizontal axis) is called a spectrogram: Here we can see three successive calls of the parakeet. Each one consists of a stack of different sound frequencies. The contrast of the color against the background indicates the amplitude of each frequency at each instant. Thus the fifth frequency component in these stacks (counting from the bottom) has a greater amplitude than the first component.
spectrumPower Spectra: A power spectrum is simply a vertical slice through a spectrogram. The vertical white line in Raven Viewer indicates which part of a sound is currently being played. It also determines which vertical slice will be shown in a power spectrum of that sound. To the right is a power spectrum slice through the parakeet's call, where the vertical white line is drawn. The power spectrum slice is rotated clockwise so that the frequency is now on the horizontal axis, and the amplitude of each frequency component in that slice is on the vertical axis. We can see that the fifth component in this slice does in fact have the greatest amplitude.

Download our Sound Analysis Primer to read more about sound visualization.

Adjusting the Settings in Raven Viewer

You can customize each of the three views by changing the Raven Viewer settings. To change the settings for a particular view, select the appropriate tab at the bottom of the Raven Viewer window.

Navigation

  • Playback Controls: The five buttons below the views provide the main controls for playback: Beginning = move play head to beginning of recording; Rewind = move backwards slowly without playing; Play = play the sound; Stop = stop playback at the current point; Fast Forward = move forwards in the clip quickly without playing it.
  • Playhead: The playhead is shown as a vertical line in the waveform and spectrogram views. It indicates the part of the sound clip that is currently being played.
  • Playback Slider: The playback slider (located between the control buttons and the views) can be used to move the playback head over any part of the sound. Subsequent playback will now begin at this selected location.
  • Location Clicking: Clicking the cursor anywhere inside the waveform or spectrogram views will move the playhead to that location. This works whether the sound is being played or is currently stopped. This is very useful for examining a particular part of a sound clip.

Shared Settings (at top right of window)

  • Loop: When set to OFF, the sound will play only once and stop. If set to ON, the entire clip will play and then automatically restart at the beginning.
  • Speed: This is normally set to 1X. Moving the setting to a lower value will play the sound back at a slower speed (e.g. setting it to 0.5X will play the sound back at half speed). Birds can often hear rapid temporal patterns that humans cannot: slowing down the playback of a sound can reveal many subtle details not heard at normal speed. Similarly, very slow sounds can be sped up to get through a playback faster.
  • Timeline Zoom: This setting sets the scale for the time axis in both the waveform and spectrogram views. A low setting scrunches the recording along the time axis, and a high setting spreads it out.
  • Cursor Readings:
    • Time: Moving the cursor along the time axis in the waveform or spectrogram views will show "Mouse Time" and which channel is being examined (if it is a stereo recording) as the time in hours:minutes:seconds for that point relative to the beginning of the recording. You do not need to click to get this reading.
    • Frequency: Moving the cursor anywhere inside a spectrogram view will post the frequency and channel viewed at the current cursor point. Do not click--just move the cursor to the location of interest.

Waveform Settings

  • Amplitude: The amplitude slider allows you to adjust the size of the waveform image within its window. An ideal setting would be just enough so that the entire waveform just fits in the display area. The amplitude slider uses a logarithmic scale.
  • Averaging: When the time scale of the waveform view is compressed, (the most common situation), many successive waves will be represented by a single vertical line in the waveform image. Setting waveform averaging to ON produces waveform images in which a vertical line composed of successive waves at consistently higher amplitude will be shown at greater contrast against the background than one in which the amplitudes of successive waves are not consistently high. The default setting of OFF draws all lines in the waveform at the same contrast against the background.
  • Color: As with the other views, users have a choice of several different waveform and background color combinations. Choose the one that works best for you.
  • Enable: As with the other views, choosing SHOW makes this view visible whereas HIDE does not. If you want a larger image for your waveform view, hide the spectrogram and power spectrum views.

Spectrogram Settings

  • FFT Size: (Fast Fourier Transform) This setting determines the time and frequency resolution of the spectrogram by determining the size of successive segments of the sound that are broken into their frequency components. Longer segments (e.g. larger FFT sizes) improve the accuracy with which the frequency of each component can be determined. However, this reduces the ability to track rapid temporal changes in the sound. Shorter segments (small FFT sizes) provide better temporal tracking, but result in sloppier estimates of the frequency of each component. The optimal setting is usually an intermediate FFT size. It is often useful to stop playback with a typical part of the sound at the playback head. Then try several different FFT sizes. The one that shows the frequency components (as in the Mitred Parakeet example above) as clear thin lines without fuzzing up the separate calls is usually the best setting. This will differ for different species. Experiment!

Download our Sound Analysis Primer to read more about FFT Size.

Power Spectrum Settings

  • Choosing sample site: The power spectrum shows the component frequencies and their relative amplitudes for a small segment of the sound clip that is currently centered around the playhead. You can see how the power spectrum changes in a sound by clicking in the waveform or spectrogram window in small increments to move the playhead in tiny steps.
  • Average: When ON, this control produces a running average of successive power spectra as well as the instantaneous spectra as a sound is played. For static examination of a sound's power spectra, it is better to turn this control OFF.
  • Color: As with other views, users have a number of different color schemes to choose from.
  • Fill: Most users will want this control turned ON. However, the option to see only the graph of the amplitudes vs frequency of components is available.
  • Scale: Most users examining bird or mammal sounds will want to set the frequency scale for the power spectra to LINEAR. This means that the interval between adjacent frequencies is the same along the horizontal axis of the view. However, those studying whale sounds may want to set the scale to LOGARITHM which expands the lower end of the frequency scale but still lets the upper end be present albeit compressed.
  • XScale: This slider "stretches" the frequency (horizontal) axis in the power spectrum view. If the frequencies of interest are not low frequencies, this stretching may move them out of the view. To bring them back into view, use the next control, the XOffset, to change which part of the frequency axis is seen in the view window.
  • XOffset: This control slides the fraction of the frequency scale which is visible in the power spectrum window back and forth. It only changes the view if the XScale control is set to a value other than zero.
  • YScale: This control "stretches" the amplitude (vertical) axis in the power spectrum view. For many bird recordings, the default value of zero is too low. An intermediate value of 20-40 may be necessary to get a good sense of how the components of the bird's song vary in amplitude. Changing the YScale usually requires a concurrent adjustment of the YOffset so that just the amplitude peaks of interest are visible above the bottom of the graph.
  • YOffset: This slider moves the range of amplitude of components visible in the power spectrum view up and down. Since every recording has noise in it, moving this slider too high simply fills the window with the ambient noise variations. Moving it too low hides the peaks of interest. Note that this is the only offset control in Raven Viewer that works even if the associated scale setting (here YScale) is equal to zero.
  • Enable: As with the other views, selecting HIDE removes the power spectrum view from the window, whereas selecting SHOW makes it visible.

Audio Settings

  • Balance: This slider adjusts the relative loudness of the playback's two channels into the user's speakers or earphones. If the recording is stereo, this will alter the relative amplitudes of the two channels. If the recording is mono, this will simply change the relative amplitudes at which the single channel is played into the two earphones/speakers. Adjusting this slider does NOT change the amplitude of the waveform displays for the two channels; it only changes what is heard by the user.
  • Bass: This control can be used to reduce the lower frequency content of what is heard during playback from the default of 100% to a very low value. It does not affect the spectrogram or waveform views.
  • Channels: For a stereo recording, the central position of this slider ensures that both channels will be heard during playback and the selected views for both channels will be visible in the viewer window. Moving it to the left will play back and show the views for the left channel only; moving it to the right will play back and display views for the right channel only. For a mono recording, leaving the slider in the left or central positions will ensure that any selected views are visible. Moving the slider to the right will hide all selected views.
  • Treble: This control can be used to reduce the higher frequency content of what is heard during playback from the default of 100% to a very low value. It does not affect the spectrogram or waveform views.
  • Volume: This slider can be used to reduce the amplitude of the playback in the user's speakers or earphones. It does not affect any of the views of the sound.

Video Settings

  • Enable: Selecting HIDE will remove the video playback from the viewer window. Selecting SHOW will make the video playback visible.

CatalogRecord

  • This tab will list some of the metadata associated with the currently selected sound or video clip.

Media

  • This tab lists the format in which the original recording was made and various parameters about the original recording.

Display

  • Playhead Color: This slider lets the user select the color for the playhead line. By choosing the current background color, this can be used to hide the playhead.

FAQ

What is Raven Viewer?

Raven Viewer lets you visualize animal sounds in the Macaulay Library's online archive. With Raven Viewer, you can display and control the spectrograms, waveforms, and power spectra of audio and video files over the Internet.

Raven Viewer took 2nd place in the prestigious "Science and Engineering Visualization Challenge" - an international competition sponsored by Science magazine and the National Science Foundation.

Raven Viewer was developed by the Cornell Lab of Ornithology and is a FREE interactive QuickTime tool that works on Macintosh and Windows.

Do I need Raven Viewer to listen to recordings?

No, you do not need Raven Viewer to listen to and view animal recordings from the Macaulay Library, but you do need to have the Adobe Flash plug-in installed in your browser.

Why do I need QuickTime?

Raven Viewer was implemented with Apple's QuickTime. The latest version of QuickTime is needed, allowing us to take advantage of the latest video and audio playback capabilities.

Where can I learn more about Raven Viewer's features?

See the details tab. specifically for you.

What are the system requirements for Raven Viewer?

See the download tab.

Why are there so many things to install?

There are two things to install to use Raven Viewer: QuickTime, and a Raven Viewer plug-in (a QuickTime component). You probably already have a version of QuickTime on your computer. Just make sure it is QuickTime 7 or newer.

You will only have to do this once. Once done, you can visualize the world's largest online archive of animal sounds and videos for free.

Why do I need to install other stuff besides QuickTime?

We use a standard set of functionalities in QuickTime for most of our interactivity, but in order to draw the audio visualizations, we are using third party QuickTime components (plug-ins), developed specially for us and you. You can download and install these components from our

I just installed everything but I am still asked to update. What's wrong?

If you just installed the Raven Viewer plug-ins but are still asked to update, it is most likely because you need to restart your web browser and restart QuickTime (in order to initialize and register the components).

I have some feedback, where do I send it?

Feedback, positive or negative, is always a positive thing for us! Our website is always under development, and we aim to improve it. You can send feedback from our contact page.

Can I use Raven Viewer on my own website?

We do not currently license Raven Viewer, but if you are really interested in it, you can contact us and try to convince us.

What if I want to open my own recording?

In the future, you will be able to visualize your own audio files within Raven Viewer. Sound file formats supported are AIFF, WAV, MP3, MP4, AIFC, and any QuickTime format.

That said, if you are serious about audio visualization and analysis, the Cornell Lab of Ornithology has a suite of desktop tools available for you, including Raven Lite, which is a free software program that lets users record, save, and visualize sounds as spectrograms and waveforms. Raven Lite is intended for students, educators, and hobbyists, and can be used for learning about sounds, as an aid in birdsong recognition, and in musical instruction.

Do the audio visualization displays in Raven Viewer render the full resolution of the audio files?

The audio available within our online catalog is compressed using MP3 codec for single audio files and AAC codec for audio associated with video. Due to the compression algorithm used, some of the frequency resolution is lost (mostly in the upper frequencies). This should not impair your ability to listen and to visualize the audio with Raven Viewer, but if you wish to analyze the full-resolution audio file, you need to contact us to order the high resolution file (usually digitized at 96/192 Khz @ 24 bits).

The audio visualizations do not seem to be rendering fast enough for my computer. How can I improve their display?

A simple answer would be to upgrade your computer :-) But more realistically, there are a few steps that you can take within Raven Viewer to improve the rendering of the audio visualization.

  • Reduce the number of views displayed. Simultaneously displaying the left/right channels for the spectrogram, waveform, and power spectrum requires more processing than displaying fewer channels.
  • Zooming out of the timeline while the sound is still playing might prevent the audio visualizations from properly tracking the current time. If you zoom out entirely, Raven Viewer will only draw the visualizations once and then move the playhead instead of rendering the visualizations as they move across the display.
Close Title