I think DVD Architect is pulling a bait and “switch” with me

[John Rofrano]

[Don Kimball] ” For some reason some of the sound tracks will intermittenly not allow the change. It says I have selected Normal in the Switches but no increase in volume and no change. That may seem puzzling and its even more puzzling that I can go into the project at another time and try the same audio track and it works! So what is with the finickey switches select/control feature?”

The feature you are using is called “Normalize”. What it does is take the loudest peak in your audio and increases that to 0dB. Then it increases all of the other audio by the same amount. If all of your audio is low, everything gets louder. If, however, there is one loud peak like a hand clap or some other loud sound, it will only take a small amount to increase that to 0dB and then the rest of the audio will only get increased by an equally small amount. Since it does this at the EVENT level, all you need to do is find the loudest part of your event and SPLIT that out into it’s own event and Normalize only the lower events separately.

~jr

http://www.johnrofrano.com
http://www.vasst.com

[Stephen Mann]

Another trick I use is to export the audio to a WAV file and run it through Levelator. Then bring the levelatored .wav file back into Vegas.

https://www.conversationsnetwork.org/levelator

Steve Mann
MannMade Digital Video
http://www.mmdv.com

[Graham Bernard]

Thanks Steve! I keep forgetting to investigate Levelator. Today’s the Day!

Cheers

Grazie

[Graham Bernard]

Guys? Are there any drawbacks to using Levelator? Clipping? Can this be done in Sound Forge?

Cheers

Grazie

[Stephen Mann]

Since Levelator does not alter your input file (it creates a new output file), and it’s free – what’s to loose by trying it?

I recall a review that started with: “Do you believe in magic? You will.”

Steve Mann
MannMade Digital Video
http://www.mmdv.com

[Jeff Schroeder]

Levelator is like automatic gain control. It brings up the low spots and ‘levels’ out the audio. Sometimes it’s great other times it’s not, especially if you have a sensitive mic and talent with an empty stomach.

Jeff

Windows 7 64-bit, ASUS P6X58D, i7 960 3.20GHz, 24.0GB DDR3, 12TB connected storage

[Jim Greene]

In my experience, it’s better to do what John suggested first, then use levelator. You will then probably need to use noise-reduction.

-Jim.

[Dave Haynie]

Yeah…. I guess “levelator” does what I was about to suggest, or something like it.

The Vegas “normalize” setting is almost never what you really want. It does peak normalization.. that means, it finds the loudest part of your audio, and makes that 0VU (0dBfs) (aka, full volume).

The problem you’re going to get right off noise. If someone coughs, if someone bumps something, etc. you may have a momentary but meaningless peak. And while your average audio levels may well be the same from clip to clip, this noise is not going to be .. and you’ll be selecting that as the basis for normalization.

What you really want is some kind of RMS normalization, and across the whole project as one thing. “RMS” means “Root Mean Square” — mathematically, it’s the square root of the average of the squares of each sample. A basic RMS calculation will deliver what’s in essence the average perceived volume of the that audio sample. Once you know the average, you can boost the level to put that average where it ought to be (there’s a whole different set of ideas about what the average level should be for speech, music, etc), probably running that as well through a compressor/limiter to ensure that the occasional peak doesn’t clip — clipping is baaaaad.

There are more sophisticated calculations, which may be able to isolate the greatest sustained power in the audio clip. If you can get such an “average sustained peak”, you can set that closer to 0VU… I’d still use the limiter.

You can probably find plug-ins in Vegas for this… I usually do my normalizing in Forge, and usually split up the tracks. I’ve shot many dozens of stage productions. First thing I do is split music and voice/narration, if I wasn’t lucky enough to have separately recorded tracks for each of these to begin with. You want to separately normalize each, even if you’re going to ultimately mix to 2-channel. You can, and often should, apply heavier compression to speech than music… compressing music too much crapifies it. For speed, it depends a great deal on the content. A typical narration is something you can compress quite a bit; some actor’s speech may well have the same kind of dynamics as music, and shouldn’t be messed with much.

And of course, most of the time, compression is added in this scenario, as your normalizing is, to make up for flaws. If a speaker is properly miced and doesn’t somehow wriggle out of that, you probably don’t have much to do. But if the production counts on stage mics, or if you just get a bad lapel placement. I saw a show a week ago where one of the leads seems to have his mic clipped to the outside of a large collar, so every time he turned to the left, his voice dropped WAY off… fortunately, I wasn’t the sound guy that day :-)…

-Dave

[Don Kimball]

Guys thank you so much for the tips. I shot an info mercial at an auto body paint shop, with no clip on mikes and limited resources. You can imagine how crazy the sound was there for sure. If you are interested here is my first rough draft of the production. This is before employing any of your helpful suggestions.

https://www.youtube.com/watch?v=qKsgo7v38sk

Cheers!

Don

http://www.polytelismedia.wordpress.com

[Jason O’connell]

I dream of the day Sony Vegas offers multiple event normalization (analyzing multiple clips to find the overall highest peak and applying the same gain to all selected events). For purpose of editing and mixing classical music concerts this type of normalization would be ideal. Would anyone else see the benefit of having this feature in Vegas?

– jason