Internet Audio: Is yours a star, or a dog? Part Two


NOTE: This is the second post in a two-part article. Click here to read part one

Last week, we talked about how the use of audio and video is exploding on the Internet. This week, we have some tips and observations particularly related to online audio content and production.

As we said last week, “content is king,” so the script deserves some attention.

Audio scripting for the Web

Whether your role is only to voice the content, or you are its producer, the script may already be written. In that case you might skip over this next section. But it pays to be aware of any way to make your services more valuable, so please read on….

Where the script is concerned, sometimes a client will invite suggestions, but generally not. As a VO pro, you know that although perceptive clients appreciate your creative contributions, you should follow the script. It might have been written a certain way for reasons you’re not privy to. Even if you have a really helpful insight, a script critique might be unappreciated, embarrassing or even tactless to offer it outright. It depends on who is in the room with you, and your client’s attitude. But sometimes there’s a way to offer a suggestion indirectly – for example, by asking a thoughtful question, or mentioning previous experience in a way that won’t embarrass anyone.

If you’re actually writing the script, here are four important considerations:

    1. Focus your message to your audience. Sometimes the most interested prospects (maybe the company’s most profitable ones) are small in number.
    2. Hone it, hone it, hone it. If you can say it shorter, do. And if there’s video, let the video do the talking ... visually, that is. Try not to slather the screen with bullet points.
    3. Traditional sales-pitch wisdom is to “tell them what you’ll tell them, tell them, and then tell them what you told them.” This remains true in an A/V presentation, but since it’s not a fleeting verbal pitch, keep the first and last parts brief … just enough to set the scene and to make it remembered.
    4. At the end, as in print media and online text, conclude with a Call To Action, or tell the user how they can get more information. Many writers forget to do this. This content or link should remain available on the page (separate from the video) even after the video is finished.

Putting audio content on the page

Two things drive users away from your site before they even start browsing or reading:

    1. Bad audio
    2. Unwanted audio

Don’t have your audio content start automatically unless the audio or video is the key focus of the page. For example, people expect a YouTube video to play automatically. But if there are other elements of focus on the page, the user doesn’t necessarily want the audio to start playing. (An example here would be Facebook.) This includes gratuitous background music that starts automatically. It’s distracting, it’s annoying, it might be too loud, and what if other people are within earshot?

Here’s an all-too-common worst case scenario: when the user has opened a bunch of tabs, it’s totally annoying to start hearing speech or music and not know which tab it’s from. One solution is to close tabs until the audio goes away … that’s hardly the reaction you want.

However, to pay the bills, the page might need to display an ad (e.g., a banner ad) that plays audio. It’s a possible exception, and the site owner will be the judge of that. But the site designer should at least be aware of the issues that unexpected audio will raise. Two things that will aid user satisfaction in this case are (a) keeping video ads “above the fold,” so they can more easily be located and (b) include a Mute or Pause button.

If you do autostart audio, for whatever reason, make it super easy for the user to locate and click that audio mute button.

If there is more than one audio link on the page, make the first one turn off if another one is clicked. (For example, our players do this on the Weekly Script Recording Contest page at EdgeStudio.com.) LINK: http://www.edgestudio.com/script-contests

VO performance for the Web

Observe all the usual VO reading and mic techniques. EdgeStudio.com is an entire website explaining them, so we won’t itemize them here. But here are a few things that sometimes pop up in Internet work:

As always, don’t touch the mic. For some reason people who would never fondle a condenser mic will nervously mess with their headset.

If not in a quiet studio, then reduce, block or avoid background noise, and position yourself and the mic to minimize it.

As always, enunciation is very important (without being unnaturally crisp). Here are a couple examples of actual closed captioned “translations” seen (among countless others) in an actual video. This text was presumably generated by speech recognition software, but it illustrates how easily human listeners can also misunderstand, particularly when not guided by context:

  • She gets her turn at the microphone
  • She gets return at the microphone

  • From the part of the bank where
  • From the parlor bank where

And as with other VO media, speak naturally, being vocally free. For some reason, some people lapse into (or request) a forced “DJ” sound just because it’s the Internet. If you must do that, have a good reason.

Technical issues for Internet sound

As VO talent, you may need to know only what we’ve mentioned above. If your client hasn’t stipulated the few technical parameters to observe (e.g., maximum decibels, the Sample Rate and compression bit rate), you should ask, and configure your recording and other software to match. But some clients, which may include web developers, might look to you for technical knowledge, or maybe even finished audio. The more you know, the wider range of clients you can serve.

These tips don’t cover everything, by a long shot. But they’re a start…

Don’t lower your standards just because it’s the Internet. In fact, online’s lower audio fidelity means that clarity in your original recording is even more important.

A good headset is often fine for podcasting. A headset encourages some people to speak more normally than they might at a microphone, and has the advantage of being a constant distance from your mouth, whichever way you turn. To avoid popping on plosives, position the mouthpiece just above your mouth, or to the side. A noise-cancelling mic will help minimize background noise such as computer hum.

A headset is also an option for on-camera talent when moving around. It avoids the need for a directional mic and a boom operator.

But remember the old computer data principal, “Garbage In, Garbage Out.” Using a mic of reasonable professional quality will give you an edge against Internet audio’s technical limitations.

Considering that Web content is listened to on such a wide range of devices, judge your final mix on all sorts of speakers, ranging from good, flat-response headphones and studio monitors, to cheap speakers, even a smartphone speaker, and (if you can) a monaural speaker that combines both channels into one. Is it clear and intelligible on all? Be sure sound is coming from both channels. And however good your speakers, be aware that their sound will be colored by the characteristics of the room.

If the video or podcast has an intro, make its volume comparable to what comes after. Don’t choose a super-loud rock arrangement with a heavy base line when it will be followed by a thin voice like, say, Georgia Engel’s (Robert’s mother-in-law on Everybody Loves Raymond). And remember that the perception of excessive volume tends to drive some types of audiences away.

Audio for portable and web devices should be consistent. That calls for normalization and compression, in that order. Reduce the volume of overly loud transients (e.g., a sibilant moment or musical hit), because the peak may limit the ability to normalize the rest of the audio to a consistent maximum. Or normalize using RMS mode -- for guidance on normalizing, see George Whittam’s Whittam’s World, “Peak & RMS Normalization Explained” (Episode 5). Incidentally, if using a hard-limiter compressor, 100ms release time is sufficiently long when those reducing short transients.
LINK: https://www.youtube.com/watch?v=wlcFkvuYBJc

For anything but a brief sound effect, a WAV file will be too big for online use. Ultimately, the audio or video file should be compressed. For audio, that generally means MP3.

If you are creating the MP3 file, listen carefully to the result before posting or sending it. What was an impeccable-sounding WAV file might have artifacts or distortion after compression. For example, an edit that sounded like just a split second of virtually unnoticeable hiss can turn into an objectionable click. Or a volume level at exactly 0 VU might exceed that in the compressed version, causing distortion.

Unless requested otherwise, aim for a finished peak volume of -3 to -1 dB before making the MP3.

You should also understand the difference between Sample Rate and Data Rate. The Sample Rate applies when you are recording. The Data Rate applies when you are compressing.

When recording audio, a sample rate of 44.1 Hz may be sufficient. That’s studio/CD quality. For some purposes, 48 Hz is desirable. Going higher may not produce any usable benefit and will increase your WAV file size. It may also require conversion downward later, which can be theoretically detrimental, too. The standard iTunes file is 44.1 Hz, 16-bit. However, we suggest you standardize on 24-bit for better quality and future flexibility.

When compressing to MP3, you may have the option of choosing the Data Rate. If the audio is strictly speech, you can set it as low as 64 kbits per second (kbps), but that loses tones above 7500 cycles. Most of the human voice is below that frequency, but other sounds include higher frequencies – for example, a 7500-cycle cut-off will cause strings, cymbals and human-speech sibilants to be distorted, or muddy, or absent.

If the recording is voice-only, set the data rate to at least 96 kbps. To preserve musical clarity, choose a yet higher data rate. Even at 128 kbps, tones above 16K will be lost, and the remaining high tones may sound “swishy.”

These Data Rate numbers assume you specified a stereo MP3 file. That’s possible even if the left and right channels are identical (e.g., if you recorded only one track, or one microphone onto two tracks with no stereo music or effects). A stereo MP3 file, in effect, contains two mono signals, so if you instead create a monaural MP3 file, you can safely cut these stereo Data Rate numbers in half – the mono fidelity will be comparable to stereo. At the user’s end, some fairly old players might not support mono MP3 files, but in most situations this is becoming a non-issue.

If you want to mix a stereo recording down to mono, do that in your DAW, not in the compression process. That way you’ll have complete control over the resulting mix and how the two (or more) tracks interact. Then export the mono track to mono or stereo MP3.

Incidentally, kids under 10 are especially good at hearing high frequencies. Most young adults top out at 20,000 cycles, losing more and more of that high-end sensitivity as they grow older or are continually exposed to loud environments. The highest frequency sampled at a 44.1 kHz sample rate is 22050 Hz.

Compressed stereo files from iTunes used to be 128 kbps, but now iTunes delivers files at 256 kbps. Setting your stereo compression parameters at 160 or 192 kbps will be sufficient for online audio, whatever the program content and type of device the listener is using. At the upper end, a 128 kbps mono file is great for mono voice, a 256 kbps stereo file equally great for stereo music.

Will that make the file too big for video streaming? Not to worry. Audio files are relatively small to begin with, and the audio component of a compressed video file is also relatively tiny. Increasing the audio fidelity won’t increase an .MP4 file size significantly.

The bottom line

It may seem a lot to deal with, but it isn’t really. A quality job is a quality job, so focus on that to begin with. Many of these same concerns apply to radio commercials, narrations in other media, etc.

Above all, when recording and presenting online audio, there are just three key points:

    1. Have a purpose.
    2. Be easily understood.
    3. Don’t drive people away!

To learn more about internet audio or to schedule with one of our voice over coaches, call our studio at 888-321-3343 or email training@edgestudio.com.

Post new comment

close

Please before adding your comment.

How to Reach Us

Call us 888-321-3343
Email us training@edgestudio.com

Click for Edge location information...

Meet Your Coaches

Edge Alumni Work Everyday

Get free educational
voice over newsletters!

Get free, educational voice over newsletters

Where should we send them?