RecVox

Voice recording and phonetogram

RecVox is a program for recording voice and phonetograms aka voice range profiles. It is primarily intended as a tool for research and education. It is not validated for clinical use and lacks integration with a journal system, but may serve as an demonstrator for future implementations of phonetography for clinical purposes.

A technical description and validation of RecVox is now available.

RecVox provides SPL calibrated recordings of sound files and on-the-fly analysis och SPL and fo. Files can be saved calibrated in the SMP and Sopran formats, or normalized and uncalibrated in WAV format. While recording, a real time phonetogram is displayed, potentially giving feedback to the subject. RecVox can open phonetograms in the stdpg file format saved by other programs, and open and overlay several phonetograms for comparison. By using two microphones, coaching in the same acoustic space as the subject is possible.

RecVox does now record 24 bit sound.

Important note: SPL calibration in RecVox is important, not only to get accurate SPL readings but also for the operation of the fo extraction. A decent external soundcard is recommended, and an omnidirectional electret/condenser microphone is required for accurate recording. Do not use directional (e.g. cardioid) or dynamic microphones. Also, a quiet environment is required for accurate recording of the lower profile in the phonetogram. You may test the functionality of RecVox using the internal sound card and microphone of a laptop, but it is likely that the recording performance will be poor. Such equipment does not handle the full range of the voice well.

Recommended equipment

RecVox runs on PC and Microsoft Windows.

This hardware has been tested with good results with RecVox:

Sound card: Focusrite 2i2, 1st & 2nd generation. Probably several other models will work well. Look for USB connection, 48V phantom power, 2 channels input, 2 channels output.

Stand mounted microphone: Line audio OM1

Head mounted microphone: Sennheiser MKE2. This mic does not come with a head mount but can be mounted in a custom head mount, preferably at 10-15 cm distance.

Head mounted microphone: DPA 4166. This mic performs well, however the distance is a bit short for accurate SPL measurements.

A loudspeaker for playing the calibration tone and keyboard tone.

Headphones for playing noise to provoke the lombard effect (raising of the voice).

A calibrator that fits the microphone or a sound level meter for accurate calibration.

Serious recordings are preferably done in a sound proof booth, no smaller than ~2 by 2 m. In a sound proof booth, you will want an extra monitor inside the booth for the subject, and some means of audio communication between the coach and the subject.

Finding a good microphone for phonetography is a pain, the Sennheiser mic above is the best I have tested, but it requires some assembly. The main difficulty is that headmounting at a distance of 10-15 cm is optimal for phonetography. Such microphones are not commercially available. It is quite hard to keep a constant distance with a stand mounted microphone, and the typical distance with commercial head mounted mics is a bit too short for accurate SPL measurement.

Note: This software is actual freeware, with no ads, time limits, spyware or anything like that.

Note: Tolvan Data takes no responsibility whatsoever for the quality of the software, by using the software you take full responsibility for validating that the software meets your requirements, and whatever legal requirements that might be applicable in your country. In particular, Tolvan Data does not accept the role as a supplier of medical devices (sw. medicintekniska produkter).