elearninglive.com » voiceover

Better Voiceover Audio Using Adobe Soundbooth CS4

Mark — Wed, 28 Jul 2010 18:38:03 +0000

I inherited an Articulate Presenter-based piece in which the audio was recorded at different times, using different microphones, etc. The voiceover talent was no longer available so I re-recorded using new voices. I always record using Audacity, but I find it’s built-in effects are limited and weaker than the ones in Soundbooth. After much trial and error, I landed on the following settings for cleaning up and improving the sound of the audio wav files, in Soundbooth CS4:

1. Apply Vocal Enhancer: Default effect.

2. Normalize process

3. Apply Reduce Noise process:
– Reduction: 25%
– Reduce by: 15%

4. Apply Compressor: For Voice: Subtle
5. Apply EQ Graphic: Bass Management:
– Low: 0
– Low-Mid: -3
– High-Mid: 0
– High: 8
6. Apply *another* EQ Graphic effect, just like above, only this time with the High setting back at 0.

What I found was that I had to always click ‘apply’ for effects before moving on to the next one. In other words, don’t get cute and just load up the ‘effect stack’ and then apply them all – in my case it didn’t sound good if I did it that way.

Unfortunately, Soundbooth does not offer a ‘batch processing’ feature for these effects, so for the 27 or so wav files I had I had to do each individually, which took some time.

Lastly, once I did the above for each of the wav files, I then used Soundbooth’s “Volume Match” batch feature to match the volume of all of the tracks of the first wav file in the group.

The result is not a perfectly clean set of audio and it’s definitely not ‘pro’ level but it’s a heck of a lot better than the source wav files were.

Keep in mind also that the EQ and Compression settings will probably need to be tweaked depending on the type of voice(s) audio you have. In this case, the majority of the voiceover audio was a female with a medium to upper-medium, sorta ‘nasally’ vocal sound. These settings took the ‘nasaly’ out and definitely helped.

Do you have a preferred set of audio processing steps? Also – I’d prefer to just do everything in Audacity. Any plugins that you have found that do a good job cleaning up voiceover audio? What I’ve found is that ‘stock’ effects intended to do this just don’t help much or make it worse. Like Soundbooth’s ‘Vocal Enhancer’ – it’s not very good by itself, in my opinion. I’m amazed that there isn’t a simple tool built to do this. But I guess with everyone’s voices sounding different it’s impossible to get right.

Captivate Audio Output Settings – Part Three – Encoding Speed

Mark — Sat, 26 Jul 2008 15:59:29 +0000

This is part three of the testing, where I’ll focus on the encoding speed setting and its effect on filesize and audio quality, in reference to voiceover audio.

Published SWF Files

Filesize Comparison

Summary

To my ears, the encoding speed set to 0 sounds the best. And amazingly, it yields the lowest filesize. So far, it seems that a setting of 48-44-0 is the best for voiceover-based Captivate movies. I’ll do some further testing this time will large, real movies that contain slide data and a lot of audio (including some music), for the next post.

Captivate Audio Output Settings Comparison – Part Two – Encoding Frequency

Mark — Sat, 26 Jul 2008 15:33:48 +0000

This is part two of a series of articles written with the express purpose of finding the best filesize-to-quality ratio for publishing voiceover audio out of Captivate. Part One can be found here. At the end of the series I’ll provide my thoughts/opinions, followed up by a tabular summary of the facts. And, of course, a recommended setting.

More information about the testing methodology can be found in part one of the series.

Encoding Frequency

There are three encoding frequency settings available in Captivate:

: Captivate Encoding Frequency Options

For this test, I settled on the following settings for each and simply changed the encoding frequency.

Encoding Bitrate: 48kbps
Encoding Speed: 5 (the default)

As with the last test, each swf was published with the three settings shown on the slide:

Published SWF Files

Published File Size Comparison

As you can see, no matter which encoding frequency is selected the published swf is roughly the same file size. However, if you listen closely you’ll hear that the clarity is slightly better when using the 44.10khz setting. This is not so evident in comparison to the 22khz sample, but definitely if you first listen to the 11khz and then the 44khz. So, it seems that it makes no sense, based on my little test here, to ever drop the encoding frequency to below the default setting of 44khz. I guess I’ll have to try with a larger audio file to see if that makes a difference? Or should I redo these tests with a source audio file that has a short musical intro followed by voiceover audio? Perhaps that would yield more definitive results? If anyone with in-depth knowledge of the meanings behind these settings reads this please comment and/or email me if you can explain some of this stuff better so I can include it here.

Summary

For years I’ve always seemed to settle on using an encoding bitrate of 48kbps. Not due to any real testing other than some quick, under-tight-deadline-pressure, testing. To my ears, this comparison seems to confirm that 48kbps is the proper setting for voiceover audio as it yields good clarity as well as good file size. Listen to the 128kbps sample, followed immediately by the 48kbps sample, and then the 32kbps sample. You’ll hear a definite difference between the three, but the 48kbps is still very good. The 32kbps is somewhat muffled. I don’t think any students/users would know the difference between the upper two, but the 32kbps is muffled enough that it would bother me over time.

In the next part of this series I’ll demonstrate how varying the encoding speed setting effects things.