Using portable handheld audio recorders for voiceover

SonyRecorderI picked up a Sony digital handheld stereo field recorder to play around with.  This category of portable digital recorders are primarily made for location recording of  music performances and ambient sound (I use it for both of these purposes), by journalists for conducting interviews, etc.

After playing around with this great recorder, I have found that it can function pretty well as a recorder for capturing voiceover.  The Sony (there are a number of comparable brands available as well) is battery operated, has 2 condenser mics, and can record 24-bit/96kHz stereo WAV files on an internal 4 gig flash drive.  The recorders usually interface to a PC via a USB cable allowing fast transfer of files to a PC for editing.

These types of recorders can be an excellent option for use within a corporate production environment, allowing the recording system to be moved to the voiceover narrator (or a more quiet location) as needed.   They set up quickly (get a small photo tripod for this purpose – 8 inch is fine), and have a headphone jack that allows for monitoring.  The Sony has a fur-style windscreen that minimizes breath pop.   And the quality of the microphones is great.

Consider this approach as a low-cost option to building a standard recording setup.

Listen to your room

One of the key elements that will affect the final quality of a narration recording is the sound of the room in which the recording is created. Many program developers record in an office that has heating, air conditioning and ventilation (HVAC) noises, computer fans, street  noises, or excessive reverberation.

While these problems can be masked to some degree using processing tools, the optimal solution to select or create a room that provides a suitable environment for recording audio.  The characteristics to address include:

  • HVAC, computer fan and other noises – these are always present, and can be difficult to address.  If possible, select a room without a heating vent, and move your PC into a separate room (using long cables for the monitor, mouse and keyboard).  If not possible, then find a location in the area that is least affected by these noises.
  • Experiment with mic placement, and some soundproofing elements.  For computers, get extension cables for your monitor, and wireless keyboards and mice so that you can move the CPU fan as far from the mic as possible.
  • Reverberation – listen to your room to determine the level and characteristics of the reverberation.  Clap your hands sharply, speak loudly, and listen to how long the sound remains.  Many room have excessive reflective surfaces that will be noticeable on recordings.  In some cases, this can be addressed by a combination of microphone placement, and use of sound absorbing materials (either commercial sound absorbing surfaces such a acoustic foams, or ad hoc approaches, such as placing pillows or curtains around the recording area ).
  • Manage noise as much as possible – Be sure to close doors and windows to reduce external noises, shut down nearby PCs that are not in use, shut off overhead lighting that may have a noticeable hum (such as some florescent fixtures), ask nearby coworkers to keep their noise levels low, etc.
  • Move around  – in some cases, moving to a different position within the room can eliminate the major issues.  For example, moving from a corner or wall location to the middle of the room may eliminate some of the more problematic reflections.

I love Noise Gates

When you are speaking, the higher sound levels of your voice will mostly cover up the lower level ambient noise present within the room.  You can hear it if you are looking for it, but in many cases, background noises can be effectively masked by your voice. However, the ongoing ambient noise becomes much more apparent during the pauses between words and sentences.

A noise gate can make a big difference when recording within a less than optimal environments.  tools that cut out the audio signal when the sound level drops below a user-defined level. Noise gates are hardware or software-based audio processing applications that function by inserting silence in the spaces between phrases and during pauses when the audio level drops below a user defined threshold level. Narrators can use gates to reduce the impact of background sounds and noise within their recording space.

Hardware based noise gates are a great addition to any recording setup. Most recording software also includes software-based noise gates that you can apply to a noisy file after recording.  These tools are relatively easy to operate, and can really make a difference in the quality of your final voiceover narration file.

You can also use software-based noise reduction functions that analyse the characteristics of the ongoing noises present within the recording, and apply selective processing to diminish these noises. These noise reduction functions are available as plug-ins or as part of the audio recording and editing software package.

Native audio vs. dedicated recording software

Many multimedia development programs have native recording capabilities that allow users to record directly within the software production environment, instead of using a dedicated recording software or device.

I usually prefer using the dedicated recording software (I like Sony SoundForge), as these programs have greater capabilities and options that enable the narrator to produce higher quality results, with greater flexibility in the production and editing.

For example,  I always run a noise gate processor to remove background noise, and tweak the equalization slightly to increase intelligibility.  These capabilities are well worth the small investment in dedicated recording software.

Always monitor your performance using headphones

A good pair of closed back headphones (that have cups that reduce the sound from the headphones from reaching the microphone) are an essential investment when you are putting together your narration setup.

Closed back headsets allow you to monitor your performance at a sound level that will enable you to hear problems or errors that can impact the quality of the record.  For example, you can listen for rustling of papers, squeaky chair noises, and other glitches that would go unnoticed.

If you need to hear your voice directly, then move one of the ear cups of that ear (and if possible, lower the volume on that side using the balance control).   This will end up saving you time in editing and retakes.

Processing narration audio files

Professional sound recording software provides a range of effect and processing options that are particularly useful for improving the overall quality of narration files.  The ones that I typically use are:

  • Noise gates – this is a processing tool that allows you to set a lower level threshold for the sound, and then mute any sound that occurs below that level, in the space between words and sentences.  Used properly, it can help make your recording sound like it was done in a well soundproofed area, because listeners hear silence in the spaces between statements, while the room sound is mostly masked by the (much louder) sound of your voice.  It can take some experimentation to set a gate properly, and it will not cover up undesirable room sounds perfectly, but it can really help clean up a recording.
  • Equalization – EQ can be used to (slightly) reduce a continuous noise issue (such as HVAC hum), or to brighten up a dull sounding recording to increase intelligibility.  It can also be used to (slightly) fix the sound of excessive room reverberation.
  • Reverb – you can enhance a signal that is too dry by (sparingly) using a little bit of reverb (select small bright room, quick decay).
  • Time compression/expansion – though I have rarely needed to use it, preferring to time the performance to the visual, narrators can use this function to adjust the duration of a narration segment to match the duration of a visual event, such as an animation.  Note that this can result in some artifacts that can make the segment less usable, but it is worth a try to attempt to fix a segment that is too long or too short.

Most recording software includes  these types of processing effects and other processing that is intended to address specific needs, but for most cases, they are not that useful for narration, and can distract from the narrative impact if overused.