SpeechRecorder

SpeechRecorder is a platform independent audio recording software customized to the requirements of speech recordings. Its distinguishing features are:

This page was last updated Nov 15 2020

• Plain or HTML or SVG or RTF formatted text, image, audio and video (Windows only) prompts
  to elicit speech in many different ways,
• Unicode text
  to display most of the world's languages,
• XML-formatted recording scripts
  for a flexible organization of recording sessions, including options for automatic recordings and randomized prompt selection,
• Configurable speaker and experimenter screens
  for an uncluttered and appealing graphical interface,
• Platform independence
  to run the software on any modern computer, from laptop to desktop.

SpeechRecorder was developed at the Bavarian Archive for Speech Signals (BAS) of the Institute of Phonetics and Speech Processing at Ludwig-Maximlians-Universität München.

This program comes with ABSOLUTELY NO WARRANTY.

Contact: Christoph Draxler, Klaus Jänsch

Screenshot:

Requirements:

PC: Windows XP - Windows 10 or Ubuntu

Mac: Mac OS X 10.7 or later, only Macs with Intel CPU

Audio hardware: microphone and headphone or speakers

Important Note

We decided to remove the audio interface adapter with multi-channel support for Windows because we do not want to renew a license agreement that would be necessary for this purpose. Unfortunately the only way to record multi-channel using SpeechRecorder now is to use an Apple Mac computer running macOS and and a multi-channel capable external audio interface.

We are sorry if this causes inconvenience in your recording projects!

Installation packages:

Version 6.2.6/6.3.4

Notes:

• Windows 10 S is NOT supported.
• Devices with ARM CPUs (e.g. Macs with M1 processor or Microsoft Surface Pro X) are NOT supported (yet).

Documentation

Manual: PDF or HTML

Release notes

Known bugs and problems:

• Title texts of selected tabs are barely legible on macOS 11 (Big Sur). A issue in Java OpenJDK.
• Animated GIF prompts are not animated. They are shown as plain images. Fixed: 6.2.2(Bug ID 0113)
• HTML prompts do not work on macOS. Prompt display appears empty. Fixed: 4.4.52 (Bug ID 0112)
• The TextGrid generated by SpeechRecorder lacks a space between `intervals` and `[0]`. It should be `intervals [0]:` not `intervals[0]:`. Fixed: 4.4.50 (Bug ID 0111)
• 2016-09-30: Recording with limited duration does not stop in capture scope session mode when media prompt (audio,video) is shown. (Bug ID 0055)
  Workaround: Set Project...->Recording->Options->Audio Capture Scope to ITEM
• 2016-10-13: Video prompts (Windows only): Video screen overlaps menu. (Cannot be fixed)
• SVG prompts with embedded images do not work with OpenJDK (on Linux) (Apache Batik bug)
• Wrong display of some mark space combining characters (e.g. with Khmer) on Mac OS X. (Bug ID 0020) (Java bug OpenJDK 8065625)
• RTF (Rich Text Format) prompts do not work on Linux ( Sun/Oracle Bug in Java ).
• (And many others... Please report.)

Release Notes

Planned features:

• Formatted text prompts (without requirement for external SVG,RTF or HTML prompt media files) (available since 6.2.2)
• Open project from command line. (available since 6.2.2)
• Keep playback lines open to avoid laggy playback start on some audio interfaces (8.x.x)
• Use captured audio stream length instead of timers for pre- and postrecording delay (8.0.x)
• Splitting session, speaker and script information to cover more project scenarios (8.0.x)
• Automatic recording level/quality control (10.0.x)
• Video prompts (all platforms)(10.0.x)
• Allow arbitrary path for workspace (10.x.x)
• Replace Java Swing with JavaFX (20.0.x)

Troubleshoot

• Error messages 'Buffer overrun':

  Please read the notes about reliable recordings.

Older versions and other downloads

List of available versions and downloads.

Development

The source and binary Maven packages are now available in our Maven repository.

Archived old source packages

Privacy Policy