How to use Vegas audio filters to enhance thin/tinny/or echoy/reverb sounding voices?

[Zhong Cheung]

I use the Giant Squid lav mic plugged into a Zoom H1 recorder to record wedding vows. They are placed both on the groom and the officiant. The bride is captured from either one of the other people.

However, most of my sound turns out pretty horrible. People say that it sounds noisy, thin, robotic, unclear. They say that it has some echoes or reverb.

I don't know much about audio editing. Anyone have tips on how to make thin/tinny/robotic/reverb-y voices sound more full, clear, and present without the reverbs or echoes? Thanks!

[Seth Bloombaum]

Have you tried listening to each mic separately in your edit? You would do this using the track mute or track solo controls.

In many circumstances, having the same voice mixed from two different mics will produce various undesirable qualities, that might be described as thin-ness, distance, echoey, etc. The reason for this is that sound takes time to travel from the mouth to the mic, and if two mics are in use, different amounts of time between the two mics. Phasing errors can result from these slight differences.

If this is the problem, the solution is to use only one mic at a time during your edit - whichever sounds best for a particular participant. Likely the officiant's for him/her, the groom's for him, and the groom's for the bride.

There's also the need to get the lav close to the source when micing. A little change in distance can make for a large difference in the amount of indirect sound (room echo, etc.) vs. direct sound (straight from source to mic).

But, thin and robotic does suggest phasing errors. Less commonly, a GS lav into an H1 might be recorded with the two channels out of phase - that would result in terrible sound or none at all, and the need to boost significantly in Vegas. The solution here is to select a single channel of the recording, right-click and select channels - Left Only (or Right Only).

[Mike Hammond]

Maybe you can post a sample (mp3) of what you're talking about?

[Seth Bloombaum]

Seems like the forum was down for a while?

I was trying to add, earlier:

Editing on headphones can be misleading about what viewers hear. It's real hard to fix a problem that you can't see or hear - good monitoring is pretty important.

Some stereo mixes don't work well in mono playback. The solution is to check mono playback before completing the project.

[Zhong Cheung]

There is only one mic per person. They aren't mixed.

The groom's vows I used were a single mono track from his single lav mic. The officiant's speech portion was only using his recorder/lav mic. I didn't try to mix both the groom and officiant's audio tracks together.

The bride's vows were just what I got from the groom's recording.

I will try to upload audio samples in a bit. Thanks!

[Mike Hammond]

Hi Zhong,

It'll be easier to give suggestions once you've uploaded something. But from your last post there might be simple fixes.

Here's what I do as long as I have good audio (meaning "clean audio" and not clashing with other audio streams which can lead to wierd sounding stuff). One thing, duplicate your audio/vocal tracks and then put one on the left channel and one on the right. That open up the audio and makes it sound fuller. Also, for thinny sounding voices, open your EQ and boost the bass and treble (in Pro 10 they are the "1" and "4" labelled dots). See if this helps.

[Zhong Cheung]

Here is the groom's vows with no edits: http://studiodreamscape.com/blog/wp-...s-No-Edits.mp3

Bride's (much worse than groom's) with no edits: http://studiodreamscape.com/blog/wp-...s-No-Edits.mp3

Officiant with no edits: http://studiodreamscape.com/blog/wp-...t-No-Edits.mp3

The couple complained that they sounded robotic, thin, tinny, echo-y, unclear, high-pitched, etc...especially on the bride's vows. They were actually okay with how the officiant sounded because he naturally has a deeper voice.

I tried playing with EQ (1 and 4), but the groom said the only difference he could hear was that it was "louder, but still robotic."

Here is the video that you can see these three tracks used. Officiant is very beginning of video, groom is around 2:08, and bride is very end of video.

Rachel + Dan: Wedding Highlights Film on Vimeo

[Chris Harding]

Hi Zhong

Just for interest where was the officiant's PA speakers and how loud were they??? It sounds very much to me that the PA level was actually higher than the groom and bride's voices. If the lav was getting a better level from the speakers rather than the actual voice then your audio would be awfully tinny!!

If this is the case then both the lavs have essentially taken preference from the PA speakers and they are often loud and tinny and being far away you will get thin audio..almost as if you had the mic way too far from the talent.

However if this is the case it certainly isn't going to help you now but it's useful in the future to make sure any PA speakers are on the guests side and facing towards them too.

In Vegas I would seriously start with a big boost of low frequencies ...keep the frequency below 100 hZ (but you can play with it and then on tab (1) (the low shelf) push the DB slider to probably as high a +10db...and start playing in that area...at this stage forget about boosting anything higher so you can essentially try to get a bit of richness in the audio. However remember if you push the gain 10db at say 20 hz your track levels will severely over modulate and go heavily into the red so you will have to pull your volume slider down at least 10 - 15db negative to keep your track around the -6db mark.

I often use boundary mics for wedding speeches and you lose a huge amount of lows with them and that's the trick I use to give the audio a bit of oomph!!

Chris

[Zhong Cheung]

The PA speakers were to the side of the officiant pointed out towards the audience. Speakers were quite loud, but not ear-piercingly so. Could hear everything clearly if you were there in person, even in the back of the audience.

Of course the PA levels would be louder than the people's voice though. Otherwise what's the point of a PA system if their normal voices were louder and good enough for the audience to hear? Lol.

However, both these lavs were recording their voices rather than the PA speakers (at least that's the goal...how would it be getting the PA systems more than their actual voice?). I think it's just that the pastor has a much louder voice in real life. I might have set his Zoom H1 input recording levels higher too than the groom's.

Thanks!

[Murray Christian]

The bride sounds almost completely off mic. I guess that's because she didn't have one (duhur me). She could use some compression. So could he for that matter.
Some bottom end boosting would help, but you may discover just how windy it was.
Basically excessive er... representation in frequencies between about 200 and 8000 are going to make it sound "tinny" to most people. It's just a matter of things being too much in one place, if you know what I mean. You could probably stand to try cutting your mids a bit in various places and see what else appears.

This kind of salvage isn't a straightforward thing though unfortunately. It's hard to say what's available to you without hearing all the raw.

[Chris Harding]

Hi Zhong

I was just asking as I had a wedding a long time back where the lav mics were 'seeing' the PA speaker as a louder sound source as they were in line with the celebrant (what we call them here!)

Did you try pushing the low frequencies at least 10db???? Also on my browser I could only listen to the clips...on the audio timeline is the waveform normal in size ..with a lav on the groom's lapel you should be almost at 0db (for my groom audio I need to drop the level at least -3db) ...if the waveform itself is thin then either the input levels were wrong or the lavs were too far away!!

Hope you get it acceptable for them!!

Chris

[Murray Christian]

OK so I had a fairly quick muck about with them just to try and see what was there.

http://members.iinet.net.au/~ragtag/...Vows-Edit1.wav

http://members.iinet.net.au/~ragtag/Officiant-edit1.wav

http://members.iinet.net.au/~ragtag/Dans-Vows-Edit2.wav

They're about 2.5 to 5 mb.
I didn't do any proper speaker tests or anything, so I'm not sure if they're that much better. I also can't tell if they'll work better in the mix or worse. What you're hearing, one way or the other, is some compression and some quite specific multi band EQing. I'm not sure Vegas can do the EQ-ing part in real time. 8 Pro can't, that's for sure. You have to make permanent changes to the audio clip. It would be a fiddly business no matter what you were doing it in in any case.

I can say that I don't think you got any significant amount of the bride cleanly. It's all PA and there's nothing to really be done about that.
I'm not really hearing anything 'robotic' as such. It's not great, obviously, but I think a lot of the complaints can be put down to people forgetting what stuff really sounds like. Which happens all the time. In memory they'll supplant the crappy PA sound they heard with the voice of the person they know (plus ears are better at picking out certain things and ignoring others too). That and clients always want movie sound, right?

You do have a bit of a noise problem though. I don't think the noise is what the clients were complaining about. That was probably the sea and the wind. But the recording doesn't seem very clear and the noise is quite high, not giving you a lot of room to maneuver. I've never used an H1 or those mics so maybe this is usual. I'm sure it's fine if you nail it. But for those times when you're outdoors on a windy day and the micing isn't the best you might want to consider something better.

[Seth Bloombaum]

I too think that we're hearing quite a bit of the PA system, and not enough of the bride's voice direct to the mic.

The PA also seems to be bouncing off the stone building that is visible in a couple of the wide shots.

All of this shows up as non-direct sources being slightly delayed and out of phase with her voice, therefore tinny and robotic. Although, this recording is really not very bad as such things go.

As suggested, EQ will help, music mix might help as well to mask these effects. But, when these slightly delayed sources are recorded, there's only a little help that can be done in post.

In this particular case, (I don't know how many of your weddings are like this), I'd really want a recording from the PA mixer. This could be to a camera or separate recorder. Why? Because these speeches are being made with lips nearly on the PA mic, therefore very direct sound. This is a case that proves the point that mic placement is more important than mic quality! Such a recording *should* show a very clean and full recording of each speech.

[Edward Troxel]

Originally Posted by Murray Christian

I'm not sure Vegas can do the EQ-ing part in real time. 8 Pro can't, that's for sure.

Why not? My pro 8 can EQ in real time...

[Murray Christian]

I was talking about a 20 band graphic and more precise stuff, specifically. But I guess they're there too, now I look. I've forgotten what plug ins it comes with and what are sundry bits and pieces I've accumulated from all over the place.