Morning all,
Ever noticed how when you watch a hollywood movie that when there is background music along with dialog that the music is just there, but also not there.

I was becoming frustrated with how simple volume reductions were not giving us the same effect. It was still sometimes difficult to hear the dialog over the music and if you turned the music down too much the impact was lost.

I set out to find a new way to reduce music but still have it present and I believe I have found it in the Graphic EQ. The plan was to reduce the dynamic range of the music so the increased range of the dialog could take over. So we applied a Graphic EQ in Vegas and gave all music we wanted to soften down a dramatic drop off for everything over 200Hz and also all low end frequencies were also dropped. This left just the middle range.

The following clip was created using this method and we would love to hear what you think;

Carly & Simon | Highlights at Great Fosters on Vimeo

Hey Danny,

I have always been in a search for a way to do this myself but could never really get one solid solution. I will definitely give this a try on my end. Although I edit w/ FCP, I know there are a few filters that are similar.

Great clip by the way!

Thanks for sharing this tip (and wonderful clip). It's another useful tool in the arsenal towards that perfect blend of music and voice. In an acoustic song such as the one you've chosen, there probably isn't a whole lot of strong signal sub 200Hz and above 6KHz. I'd like to see how well it affects the overly-produced/compressed and bass heavy tracks of some other songs. I try to avoid those anyway. Piano, voice, and guitar seem to make a good bed for moving vows, toasts and sound bytes.

It works really well on all tracks I have tried so far. I found by dropping off the bass end it meant the transition from normal to treated is almost seamless. Essentially all it's doing is flattening the track. I believe most nles have a graphic eq so you should be ok in fcp. Ill try and grab a screenshot of my particular graph. I'm sure over time I will refine it.

I did experiment with a compressor/limiter. This too had some good results but the effect was more noticeable, especially on the stronger tracks.

Will check this out during my next edit, i use Vegas and know exactly what you're talking about.

Thanks for this and what an amazing clip, they looked like two hollywood stars to me.

Danny,

You're absolutely right about using EQ techniques to combine music and speech. Several years ago when I was working in TV, there were a certain set of frequencies we took out to "cut a hole" for speech. There's also EQ you apply to the speech that works in conjunction. I'll have to look around for those again. I had them as presets, and forget what they are now.

screenshots of these presets would be great :)

Damn! The secrets out! Ha ha!
The use of EQ is an essential part of any audio mix. Whether simple voice and music or a full audio post or an album. Truth is that most see EQ as a fix for issues and while thats fine, the art of EQ must not be forgotten. Much like painting a scene with lights, EQ can give audio tracks their own place in the mix. It's just as important a tool as panning and volume.

Mix away friends! And use your EQ!!!

Your friendly neighborhood audio engineer turned video producer!

I ain't no expert.. but isn't the 1K range is where the vocal is? Should it be lowering the 1K range in EQ so the vocal of a song won't be that noticeable? I might have to give it a try to find out. Anyway your clip is stunning. Great sound too!

Yes great pix especially the brides tears but any vocals under any *amateur* spoken voices won't do it for me. All audio speakers and rooms have different acoustic characteristics and I'd never take the chance they could conflict, maybe for older folk who have hearing difficulties.

Yes .. in a Hollywood movie with a vocal title track that features, is mixed by pros in a post studio and played in commercial movie houses.

And yes .. music under amateur voices but instrumental, I've even doubled the instrumental break in a vocal track so I could start to fade it up near the end and have the last chorus up clear of the speech.

If wedding clients have a favourite vocal song they insist on including, I'd suggest hiring a pro pianist to record the song in the same tempo as the vocal version and cross fade appropriately. He could even score it to picture, matching the various scenes. Just maybe the wedding musicians could do it.

Cheers.

We did play with using a compressor but didnt like how the end result came out. It just sounded muffled or wrong some how.

No doubt over time we will come up with something better and there is still tweaking to be done but for now, this works so much better than what we used previously.

We also reduce the gain slightly as needed. On some tracks it doesnt need it but for others it does.

Taky, Most of what I read said it was the 200Hz for vocal but with the internet being what it is this varied from post to post. So I played and I was happier doing this around the 200Hz mark. No doubt this will vary though from track to track.

http://farm7.static.flickr.com/6081/6081783355_0dc0d06fd7.jpg

If wedding clients have a favourite vocal song they insist on including, I'd suggest hiring a pro pianist to record the song in the same tempo as the vocal version and cross fade appropriately. He could even score it to picture, matching the various scenes. Just maybe the wedding musicians could do it.


I think this would actually give the best results! To my ears, the song's vocals are still fighting with the bride and groom's vocals. There are some words in there that I can't understand because it's clashing with the song's vocals.

I'm finding the EQ adjustment an interesint methodology to use... I'll have to give it a whirl some time.

To share my bit of experience.... on a recent project, i was also not happy that the vocals in a music track were clashing with the spoken dialogue that I wantd to highlight in the video. So what I did was found a karaoke version of the song that matched the origianl and layered the karaoke and original version on the timeline. I could then adjust the volume of each track accordingly to adjust the entire music+lyrics volume or fade down the original and effectively fade down the lyrics to whatever level I wanted them at and still keep a good presence with the music track behind the spoken audio in the video clip. It worked wonders for me and really polished up my finished product quite a bit.

maybe it is my hearing, but it would be nice if you have a comparison of "with or without" EQ. Id love to see how it really makes a difference. specially when you say

"was still sometimes difficult to hear the dialog over the music and if you turned the music down too much the impact was lost."

The reason I asked is because it could also take time on your workflow. I tried sweetening audios using fft, eq,dynamic processor but it just takes a little time for me to do and found that for the end client, it is non noticiable.

this is just my opinion

The worst is always when the music bed has vocals in it, and you're mixing speech with it. It's very difficult to get them to play nicely -- but not impossible. Nowadays, my patience isn't what it used to be, and I just use instrumentals only :)

Hi sigmund.

I'll try and get a comparison up but the best way is to give it a go yourself.

As for time I've actually found it quicker. In Vegas I used to have to add an audio envelope, add a point, reduce volume.... Now I split the audio, extend it to crossfade and then apply my preset (as seen in the picture as 'reduce vocals'). Then, for all future drops I can just paste the attributes on.

As for if the client will notice. Probably not, they've always been happy with the volume drop method but put side by side, if you ask them which ones better I would like to think they prefer this one. If nothing else, it's sped up my workflow slightly.

Thanks for sharing! Gonna give it a try!