Processing Recorded Dialogue (Environment Effects)

Hi, all.

I'm mixing recorded dialogue for a cartoon that I'm going to be working on, and I've come up against a problem.  When the dialogue is just played flat with no EQ or effects it sounds, well, flat.  Like, for example, my characters are outside talking.  The sound of their voices is just clean like they're in a recording booth, which sounds out of place for people who are supposed to be outside.  I've tried adding an ambient noise track, but that further accentuates the clean-ness of the recordings and makes it even more awkward.

Do you guys have any tips for achieving the sound of different environments through EQ and effects?  I need to figure out how to make it sound like my characters are outside/inside and that sort of thing.  All of my dialogue will be recorded in pieces and emailed to me from different locations, so this rules out recording in different environments.

Also, do you have any tips for just EQing (and adding effects to) spoken dialogue?  I'm not quite sure what I should be doing to the audio to get good results.  So far I've been rolling off the super-low end and the super-high end so that any sounds outside the range of the human voice aren't muddying things up, but that's about as far as I tend to go with processing (I'm a n00b).  So, anyways, I could use even the most basic tips.

Thanks in advance for any advice.

Sounds like you're looking for reverb, my friend. For outside simulations, your best bet is probably going to be convolution reverbs. Sonar has one of these. They use things called 'impulses' which are kind of like a snapshot of the acoustics of a space. So if you take an 'impulse' of an outside environment, you can run your audio through this convolution reverb using an impulse of an outside environment, and it makes it sound like you're outside.

There are lots of websites out there with plenty of free impulses of lots of things. If you do a quick google you should find plenty. SONAR might even have some ideal impulses in it's own library. You won't have to make your own outside impulses (though you can if you want).

Fortunately, they are also very easy to use.

Shoot back if you have any further questions.



Furthermore, eq your voices so they sound consistent. You certainly want that. Some light compression could also be very beneficial. I would however suggest using volume automation first to get all the levels fairly consistent (or louder if it is necessary for the part - however you typically wouldn't want to be going significantly louder, you'll scare/deafen the listeners). Then I would apply some light compression to tighten things up a little further PROVIDED YOU KNOW HOW TO USE A COMPRESSOR! Otherwise stick to volume automation, cause you might make it worse if you don't know what you're doing with the compressor.

Rus W


^ Matt? How far off base was, I?

Haha, I think we're on the same page.


I personally wouldn't bother with spectrum analysers though. With something so variable as the voice with no other music, I don't see it giving any useful info. I'd prefer to use my ears.


Delays could work but eh. Difficult. I have also never seen a traditional reverb with an outside setting.


In film I believe it is quite common for the sound guy to take impulses of the rooms for every scene (and potentially even every actor location if they move around a bit and reflections change significantly - ie in the centre of a church vs the corner of the church) in case overdubs are needed.

If you have well recorded voices and you are visually outside then think of what happens to sound when outside. If anything you loose bottom end and top end and some level. I have recorded actors for an animation all in the studio but the final scenes went outside a lot.

I rolled off bottom and top end to make the vocals less toppy and smoother sounding and a little less bassy. (filters mainly) When someone is outside and even further away you need a little more top end off and lower the level.

If I were to use a reverb to simulate an outside acoustic it would have to be not normal or room sounding. Something to just add a little outside effect to the dialogue and it would be different to the reverb used inside a room for example. (Early refs can work well here for outside) But once inside that could be a more natural convolution room sound for example. You have to get fancy with reverb changes and levels when going inside and outside during final post audio mixing. I ended up with a set of inside treated tracks and outside tracks and just move the dialogue around between them. 

Any outside atmos effects are only going to reinforce outside. Just keep them at a refined level compared to the dialogue.  If you have got nice well recorded stereo outside atmos then you can get that sitting at a nice level where the dialogue treatments you end up doing sound more effective. It is the outside atmos effects that are really telling us in some ways we are outside. When that is right the dialogue almost takes care of itself.

Outside there is extranious noises plus there isn't usually a reverberant field but more of a short subtle slap echo that occurs dependent on the environment and proximity to structures. For round numbers sake, if a character is talking and there's a building wall 20' away then applying a 20 ms delay with roll off can project the character into a pseudo space.

Even though I may refer to EQ ing (filtering really) the ends of the spectrum I think it is also important to think about the range of frequencies in between. I was recently teaching post audio for TV using Pro Tools and the students had to do a voice over read of themselves. After editing they were interested with the EQ stages that could be applied and the fact their own voices could be made to sound quite better than the flat original recording. Even though that was done with a quality mike and pre etc in a nice booth too. Inside a booth or tight space I have been known to mic close up as usual but also from a little distance away to get a different perspective on the voice over. This can help in post mixing when moving from space to space.
 
Each voice has its own character in its own way and it is good to be able to lock in on those frequencies. You can try creating a nice peak and sweep it slowly while the voice is playing back. This might help you locate the frequencies then you need to make a smoother boost curve and use less to just enhance etc.. It is hard to just recommend boosting or cutting at certain frequencies because they vary from voice to voice. It is good to home in on the characteristic sound of each voice which can involve a little boost in the right places to really bring these qualities out more.

If you are rolling off tops for outside uses then you can afford a little boost to keep clarity up there between say 4 and 6 K Hz etc.. I just want to convey the idea that after the correct filtering is applied then some EQ ing can take place for the better. Each character outside can have this EQ on its own track or buss making it all easier to control during post mixing.

Using effects is OK with outside visuals bit be mindful of where you might be. If you are standing between two buildings then some form of delay might be in order but if it's outside in a field with no reflective surfaces anywhere that won't work. That might just need the right EQ and perhaps a very tight convolution space to create a very soft sense of ambience even though it might not be there in real life. It's OK to take poetic license here and there.

The main thing is if you are outside have a look around and see where you are and get that nice stereo atmos effect in there to set it up perfectly. Every outside must have some sort of atmos and you will immediately get a sense you are out there. On shoots I have recorded atmos on serious stereo devices as well as the devices Mike suggests. They are very good now and no one needs to know you are necessarily recording either. They can do that now easily. But if you are doing an animation for example you have to source the effects but they are all there and been done well too. You may need several effects to create the perfect atmos you are trying to create. For example a country fair atmos may be hard to source but all the individual elements are easily available so you can build it up. That is fun and you have control over the effects then on the timeline. World war 2 scenes may be harder to find but all the elements are very available making it easy to build an amazing sound atmos scene. Many mono effects too can be panned and treated for stereo effect later. When the atmos is wider the dialogue sits well right in the centre and does not fight the atmos anywhere near as much. Atmos can often benefit from a little EQ itself and compression to keep it under control and then finally automated for best effect.

Nothing says "we're outside" like the red shouldered hawk from this library :-) :



http://www.sound-ideas.com/index.html





Convolution reverb is especially great at helping match dialog and sound cues from different sources to an interior scene.



Now a days, one of the best ways to get ambient beds for spacial cues is to carry one of these around and grab stuff as you find it:




Have fun.

best regards,
mike

The way I see it:
When I worked in the radio theatre the studio for outdoor scenes was always the dead one, reverb was only used when very loud voices were used . Normal conversation was recorded with dead acoustics and necessary background noise added.
That's how it is, when you are outdoors. Only in a silent city, between the houses, or in a stony valley or a cave can you hear any reverberations using normal conversation volumes.
Except for the examples I mentioned, if there's a slightest bit of reverberation in normal volume speech it positions the actors inevitably in a closed/limited space.