We live in a wonderful time for music technology and learning. Nowhere is that more apparent than in our ability to control speed and pitch independently in an audio file. One particular benefit of this capability is that it makes the once laborious task of learning music by ear and transcribing tunes and solos much easier than ever before.
AudioStretch was developed with these tasks in mind, and in the few years it has been available the iOS app has attracted quite a following. The app runs on iOS version 9.1 and later and is compatible with the iPhone, iPad, and iPod Touch. I reviewed the latest version, 184.108.40.206, on both my iPad Air and iPhone 5S, running iOS 10.1.1.
AudioStretch looks and operates the same on all devices, with one exception; on the iPad it works in both portrait and landscape modes. In addition to the full version available for $9.99, there is a free “lite” version of the app that offers most of the basic functionality. Although it has limited ranges for song length, speed and transposition (among other things), the free version is a great way to test-drive the app before you buy.
The app supports both audio and video files, which can be loaded in via your music or video library, from iCloud (which requires the use of iCloud Drive), DropBox, or even from video you shoot on the device itself. You can export audio using these same methods, as well as by email and AirDrop.
AN INTUITIVE INTERFACE
Fig. 1. Straightforward and intuitive: the main screen of AudioStretch as displayed on an iPad.
Once a file is opened in AudioStretch, you are presented with a clean and elegant, mostly grayscale screen layout, with a large waveform occupying the center of the screen (see Figure 1). Above the waveform is a time readout showing the current playback location. Below that is a complete file overview showing where you are within the file. There are transport controls for Play, Rewind and Fast Forward. The Rewind and Fast Forward controls interact with the functions just above them, locating to wherever you drop pin locaters (useful for marking sections of a song), and set A and B locations (which can be looped).
Fig. 2. A video file, which can be manipulated with the touch of a finger, displayed on my iPhone.
Below the transport controls are fields for setting the playback speed and pitch of the file. In the full version of the app the speed can be anywhere from 10x backwards or forwards. Pitch can be changed up or down over a 3-octave range. It may seem strange to want to play a file you are trying to learn back-wards, or to transpose it by such a vast range, but these are actually cool sound design tools. And because you can export your audio, the app goes far beyond just the needs of music transcription. If you load in a video, it shows up in its own window (see Figure 2), which can be resized to fill up more of the screen if you don’t need to see the waveform view.
Press Play to start your file playing. Tap on the minus character to the left of the Speed field and the speed drops by 5 percent for each tap. Drag your finger across the field and the speed changes by finer increments, or it can be changed by a much wider value with a longer swipe—very elegant. The same is true for Pitch: Each tap on the sharp or flat icon moves the pitch a semitone, or swipe for finer resolution. The audio engine sounds very good, and the fidelity holds together surprisingly well even at slow speeds.
Hold your finger on the waveform and you can scrub the audio in either direction to move the playback locator. When you let go, playback resumes. You can hear the audio while you scrub, so it is dead-simple to find a given note or location using both your ears and eyes. Similarly, if you slide your finger across the file overview you can quickly scrub/locate to any section of the tune you want, albeit with no audio monitoring.
For transcribing music or learning solos, the ability to slow down the audio and loop sections, as well as use the tuning controls, gives you the tools you need to get the job done. I really enjoyed working with video and being able to slow things down when I wanted to watch the player’s fingers.
But AudioStretch has a few more tricks up its sleeve. By far my favorite is the ability to freeze the audio by simply double-clicking on the speed field. This is like setting a very short loop at the current location, and I often do this in other programs when I am trying to ascertain all the notes in a chord stab. With the sound frozen, I can try different pitches on my instrument to find the exact voicing or chord quality. Double-click again and the file returns to the current playback speed.
KEYBOARD AND SPECTRUM
Tap on the gear icon in the upper right of the screen to access various settings and information for the app. Here you will find an option called Keyboard and Spectrum. When enabled, it displays a traditional piano keyboard above the waveform and denotes the prominent frequencies while the file plays. Of course, in a full-band track there are going to be a lot of frequencies shown. I found it easiest to follow when running the tempo very slow, and it was also helpful to stop playback on a given chord, so I had time to digest what I was seeing and hearing.
Fig. 3. A view of the Keyboard and Spectrum feature displaying a single F# note played on a piano. Note the additional frequencies being displayed.Fig. 4. Another view of the Keyboard and Spectrum, accurately showing a chord voicing.
Part of the difficulty is that the Spectrum function is analyzing and showing all the frequencies it hears. In Figure 3, I looped a single note played on the piano, and yet we see 11 strong frequencies. The note played was an F#, and astute readers will recognize the harmonic series of overtones being displayed: The lowest note is the actual note played, the second frequency is an octave above, the next is a fifth above that, and so on. (The taller the line, the louder the partial or harmonic is perceived by the algorithm.) The harmonic series is a basic rule of physics, so it is understandable that we are seeing this result, and I’m not saying the Spectrum feature doesn’t work: See Figure 4 for another voicing that is clear and accurate. Rather, I hope the developer will refine the algorithm to lessen this effect, perhaps by significantly lowering the non-fundamental frequencies being displayed.
MORE ROOM FOR IMPROVEMENT
As good as AudioStretch is, I have some recommendations. Currently, if you transpose a song (as I often do to bring unclear bass notes up an octave for a quick check) after you have tuned the file to match the concert pitch of your instrument, the tuning is lost, and you have to reset it. Though that’s easy enough to do, I would love to see a dedicated octave button that I could tap to raise or lower the file by one octave.
In addition, while AudioStretch can read BPM info imbedded in a file, there is no way to set a value when BPM data is absent. The user should be able to tap on the field to open an edit dialog, or be able to set the BPM via tap tempo, so that the hash marks above the waveform can be accurately aligned to the audio.
Last, I often use EQ in other transcription programs to help me hear parts more easily, either by rolling off the treble or bass, or by boosting midrange to hear buried string lines and other parts. AudioStretch has no EQ, although it does support the Audiobus standard and you could route audio through another application. But it would be much easier to have EQ in the app itself. None of these are deal-breakers, but they would make an already fine application that much better.
ANALYSIS AND MUCH MORE
Overall, AudioStretch is an excellent tool for learning and transcribing music, offering an elegant and clean interface, superb audio quality, and wonderfully intuitive operation. I quickly became spoiled by the touch interface and the freeze function.
The only competition for it I could find on iOS is AnytunePro+ for $14.99. It already includes EQ, and an innovative feature for isolating or removing content based on frequency (called ReFrame), so you could remove vocals, etc. However, I much prefer the clean interface and operation of AudioStretch, which allowed me to figure out tunes quickly and accurately. Moreover, its ability to play files backwards and forwards, along with its wide-ranging pitch and speed capabilities, make AudioStretch suitable for sound design work.
Elegant interface. Intuitive audio scrubbing and freeze. Superb audio quality. Supports audio and video. Full version offered free to qualified educators.
No integrated EQ.
Lite version: Free
Full version: $9.99
Jerry Kovarsky had a 30+ year career in product development and management. He is the author of Keyboard for Dummies, and is a longtime contributor to Keyboard magazine.