Audio for Video. What do you do differently?

[NorseHorse]

This topic came up in the "Discovered" at the Dell Lounge... What a Spectacle! thread. At one point, the camera zoomed in on the string section (up on a roof in open air) and I experience cognative dissonance because of the reverb on the strings. It made me aware of the "man behind the curtain" in a way I don't usually enjoy. Not everyone shares this experience, but it's something I certainly consider when preparing audio for someone to sync to video.

What do you do differently when recording or mixing for video?

- I am less liberal in my use of FX or "tricks". (This should please the realists!) I'm not generally concerned about giving an "authentic" representation, but when you incorporate the visual element, there are times when the audio wouldn't compliment the video if you just process to your ear's content. For example: if you record in a small hall and then use Altiverb to make it sound like you recorded in a cathedral, there is cognitive dissonance between your eyes and ears.

Of course, with music videos and other promos, there is a suspension of disbelief and "otherwordliness" so you can be more liberal in your processing, but for documentary type videos, particularly of live events, there is an assumption that you should be hearing what you see.

Another example: like many of you, I recorded a performance of Handel's Messiah this Christmas. The soloists were on the stage right side of the orchestra. Through the use of a spot mic, I was able to move the voices to the center of the audio recording. The performance was not videotaped, but if it had been, the audio-for-video stream I provide would not have included the centered vocals or it would have been more subtle.

I'll be attempting to upload a video sample soon.

[NorseHorse]

<p><object width="576" height="324" ><param name="allowfullscreen" value="true" /><param name="allowscriptaccess" value="always" /><param name="movie" value="http://www.facebook.com/v/545709935827" /><embed src="http://www.facebook.com/v/545709935827" type="application/x-shockwave-flash" allowscriptaccess="always" allowfullscreen="true" width="400" height="250"></embed><br></object> 

Hey Steve, is it possible to embed videos that aren't from YouTube? In the meantime, I'll include links...

To the rest out there: below you will find links to two videos of the William and Mary Wind Symphony. One done by another engineer, and the other by myself. Both sound decent. But one sounds fake with the video. That isn't to say it doesn't sound "good" - I'll let you judge that - but it does sound fake.

Exhibit A: Videos Posted by The William & Mary Wind Symphony: Savannah River Holiday | Facebook
Exhibit B: Videos Posted by The William & Mary Wind Symphony: Rhapsody for Hanukkah | Facebook

However, I would suggest that the one that sounds fake with the video, also sounds great without the video. It's just that the mix is not appropriate for the video.

Perhaps you wouldn't do anything different when mixing for video. Or perhaps you do things that I haven't even considered. Either way, please share your thoughts!

[huub]

My only aim is to make the audio in the given circumstances sound as good as possible..
Only thing I do differently for video, is add more ambience than I would like to..
If the audience is clapping along for instance, and the cameras show it, I kind of have to make that audible.. I would prefer not to, because I generally don't like adding a whole lot of venue reverb to my mix..
I really don't care about making things sound 'real' and music that sounds 'fake' or 'produced' in a video situation does not bother me at all.. in fact I enjoy that..
This is for music though.. For non music, my only aim is to make the director and producers happy..
Wellll.. I still want make things sound as good as possible of course (in the given circumstances)...

This reminds me of something I really really hate: in movies, if a group is singing, and the camera pans across the group, very often they raise the level of the voice of the person that's on camera.. Man, I've always hated that! Imagine raising the level of instruments that are close up on camera during a concert..

[Remoteness]

Well, it really depends on a few things.
Are we talking about a live performance in front of an audience?
Are we talking about a live performance to “tape” in the studio?
Is it a music video?
Or, is it a documentary where you would want to record exactly what is happening at the time of the capture?

----------

When you brought up the discussion about PT Walkley’s string sound, you NEVER really answered my reply.
I'm thinking there is a future in politics for you my man.

I got the feeling that you have SELECTIVE cognitive dissonance.
Yes, the strings are right in front of you and they were performing on a rooftop.
Should it have been a dryer sound and what about the vocals?
I had the same reverb on the vocals and the strings
Should I have left the vocals swimming in wetness? I don't think so, but...

IMHO, it’s about the taste factor; at the time, I liked how it sounded, the client liked it and the artist signed off on it, so I went with it. Listening to it again, perhaps the reverb could have been brought down by a (redhead) hair, but it doesn't destroy the experience for me. AND, if I brought down the verb on the strings it would have left the vocals in complete wetness. The vocals would have also needed to be brought down by that same (redhead) hair.

IMO, you have "selective cognitive dissonance" syndrome.

----------

For me, when it comes to audio for video, especially in a live performance, its more about what you see is what you should hear rather than how much (or how less) reverb is on the mix.

I also, like to place the instruments and vocals as you see them, but I don’t go out of my way to establish this to an extreme unless it’s necessary for the audio/video relationship.

I checked out both Exhibit A & Exhibit B and noticed something different about the two selections.

I noticed that the harp in Exhibit A was missing from the sound capture.
Perhaps a spot mic on the harp (and other soft sounding instruments) would have helped the final audio to video balance.

This would not be so much of an issue without the video, but man, I hate watching a video and not seeing the instruments that are on a close-up or in view.

I also noticed that Exhibit A had a lot more reverb than necessary (YMMV), but I feel the mix had much more definition than Exhibit B.

IMO, Exhibit A had a lot more detail and clarity despite the extra reverb.
The clearness of the mix seems to cut through the addition reverb, but I bet it would excel even more with a lot less reverb. With or without the video, I feel this mix would have sounded much better with a little less reverb. Exhibit B sounds very nice and natural, but the instruments are missing some clarity and definition.

For me, when it comes to reverb balance it’s really about taste.
When it comes to instrument balance that’s entirely another story.
Refinement and judgment is an important part of the process, but our ears should always be the critic when it comes to the end result.

You talk about “cognitive dissonance between your eyes and ears” when an instrument in one space is producing a sound of another (larger) space, but what about when the piece was written that way or when a keyboard or sampler is playing a part that’s bigger than the space it’s performed in?

All in all the audio capture and final mix must sound as good as possible.
The proper balance of audience and room tone (ambiance) is the key to fusing sound to video in a live production.

In the end, it is really all about making the client, director, producer and artist happy.

[Shadow_7]

My main gripe with video is the delay between action and noise. Using the on board mics of a given camera and some distance. Which is fine (I guess) when fully zoomed out. But the second they zoom in for that nostril shot, the delay is out of place. But if they correct and the delay makes the sound happen before the video, that's even worse. Or if they do a lot of edits without regard for A/V sync. Even worse, if they try to mix what they get at their (balcony) location with what you give from your (on stage) location in real time. Not that I've done any of this yet, but I'm at the 11th hour of taking that video plunge.

[Sonic Lush]

this evening I've just discovered Sake cocktails.

so, I'll be brief.

I captured sound of a band for a video shoot and to make things simple, and free, I used a single MBHO stereo PZM Jecklin disc.


It sounds fine, but when you watch the fabulous HD videos,, it would be nice if the soloist on screen's volume would rise when they are pictured soloing.

like left hand sound didn't know what right hand picture was up to.

other than that it sounds fine.

the visual element does often entice one to suspend adherence to visual and sonic reality.

love walt

[audio ergo sum]

Originally Posted by Sonic Lush

this evening I've just discovered Sake cocktails.

so, I'll be brief.

I captured sound of a band for a video shoot and to make things simple, and free, I used a single MBHO stereo PZM Jecklin disc.


It sounds fine, but when you watch the fabulous HD videos,, it would be nice if the soloist on screen's volume would rise when they are pictured soloing.

like left hand sound didn't know what right hand picture was up to.

other than that it sounds fine.

the visual element does often entice one to suspend adherence to visual and sonic reality.

love walt

The ideal is that picture would care for the requirements of sound and vice versa. In an ideal world without power games this would happen. It's funny, since sound is so much more important than visuals for the reception of music.

"I need this microphone here, so you find another way to get this picture..."
"You are getting too close with this camera shot, it does not correspond to the sound"

Now that sounds reasonable to me.

You all have been brainwashed by the narcissistic video people.
Audio for video? Haha!

It's video for audio if we are talking music!

[videoteque]

it would be nice if the soloist on screen's volume would rise when they are pictured soloing.

It's funny, that's one of the things I hate more in a taped concert...

Maybe I don't like it when it's done too obvious and done well it's great...

I think that if all instruments are well mixed, when you see one of them (and soloing it should sound louder by itself) your brain automagically turns up the volume for it, no??

[Jim vanBergen]

Originally Posted by NorseHorse

What do you do differently when recording or mixing for video?

For me, the first consideration is:
is my live broadcast mix the final mix, or will this be post-mixed?

When it's the final mix, I will usually:

1) focus on having a limited dynamic range (especially for classical music!) to stay within the very limited 14dB dynamic range that most sat trucks are expecting ( the -10 to +4 VU range in analog, or the slightly wider -30 to -10 range for digital). This is not something I hear discussed openly, it's what I have experienced back from transmission trucks when I don't consider the limited dynamics and earlier in my career, would listen to my confidence mix and hear them compressing the living SH** out of my mix. "It was too low for our transmission standards, so we put a limiter on it." Now I discuss this IN ADVANCE and mix accordingly.

2) I too, will limit the amount of added FX processing for most elements. But it totally depends on the moment & what the client wants to hear. Gating, compression, reverb, delay, etc- sometimes you really need a very WET approach for instrumentation, sometimes the songs have FX that are integral. Sometimes you have one chance to get it right...I have experience both nailing it and f*&king it up. I tend to have a LOT more verb/fx on studio-based projects or post-mixed events, and the live capture shows tend to have enough room to reduce that need.

3) too much ambiance/audience response. I am VERY envious of the studio engineers who have been able to tweak their audience mics, zone and limit the house PA and offset the response mics, and get just the right amount of applause/laughter. I have one producer who always wants a little more room- if everyone stops talking, he wants to hear the room tone in the background; kinda like the recent Golden Globes, whether it was too low a ceiling, too much booze for the crowd or too much room in the mix, you heard a LOT of background chatter the entire time. To me, that's a bit excessive. But he who pays the bill gets what he wants.

These are elements I will ALWAYS be working on. And in items two and three, these elements interact so much, it's about their balance and relationship. If your audience response is too present, it's too much room. If you take them out during the songs, you miss the 'instant response' factor that makes it live.

JvB

[audio ergo sum]

Originally Posted by Jim vanBergen

For me, the first consideration is:
is my live broadcast mix the final mix, or will this be post-mixed?

When it's the final mix, I will usually:

1) focus on having a limited dynamic range (especially for classical music!) to stay within the very limited 14dB dynamic range that most sat trucks are expecting ( the -10 to +4 VU range in analog, or the slightly wider -30 to -10 range for digital). This is not something I hear discussed openly, it's what I have experienced back from transmission trucks when I don't consider the limited dynamics and earlier in my career, would listen to my confidence mix and hear them compressing the living SH** out of my mix. "It was too low for our transmission standards, so we put a limiter on it." Now I discuss this IN ADVANCE and mix accordingly.

2) I too, will limit the amount of added FX processing for most elements. But it totally depends on the moment & what the client wants to hear. Gating, compression, reverb, delay, etc- sometimes you really need a very WET approach for instrumentation, sometimes the songs have FX that are integral. Sometimes you have one chance to get it right...I have experience both nailing it and f*&king it up. I tend to have a LOT more verb/fx on studio-based projects or post-mixed events, and the live capture shows tend to have enough room to reduce that need.

3) too much ambiance/audience response. I am VERY envious of the studio engineers who have been able to tweak their audience mics, zone and limit the house PA and offset the response mics, and get just the right amount of applause/laughter. I have one producer who always wants a little more room- if everyone stops talking, he wants to hear the room tone in the background; kinda like the recent Golden Globes, whether it was too low a ceiling, too much booze for the crowd or too much room in the mix, you heard a LOT of background chatter the entire time. To me, that's a bit excessive. But he who pays the bill gets what he wants.

These are elements I will ALWAYS be working on. And in items two and three, these elements interact so much, it's about their balance and relationship. If your audience response is too present, it's too much room. If you take them out during the songs, you miss the 'instant response' factor that makes it live.

JvB

And this is different in audio without video HOW?

Those dynamic limitations are big. Never experienced anything like that here in Europe. Possibly because the payed for bandwidth is actually not sufficient for audio?

[Jim vanBergen]

Originally Posted by audio ergo sum

And this is different in audio without video HOW?
?

This is VERY different from how I approach mixing audio without picture.

In video, image is primary. Audio is secondary and you have to follow picture. Audience shots mean MORE AMBIANCE. The music mix can sometimes (sadly) be considered by a producer, less important than the "being there from the camera's eye."

Music only?
Music & lyrics are primary, Audience is an afterthought- sometimes it is limited to "almost a flavoring". (Client's quote)

FX are VERY important, and I'd never put out a completely dry vocal. I don't use the ambiance as room tone.

I don't have a director creating a visual stimulus in addition to a producer as my end client.

I can have a full and complete dynamic range. If it's radio, I have some, but less stringent dynamics restrictions than TV broadcasts. For webcasts & internet content I unfortunately have more.

Cheers!

[Corran]

Regarding Exhibit B, I must say I really like the recording.

Can you give some specifics on the placement of mics and such? It looks like a couple of AT4040's or similar in wide ORTF but surely that's not all that was used.

My recordings of wind ensemble have "left something to be desired" in my opinion (though my fiancée insists they are great) and I feel it's been captured here.

I personally think I've been miking too close, which is why I was wondering if there was some mics off camera in your recording.

Exhibit A sounds like a gym and/or hockey rink I say hockey rink because I've played a concert in one and that's exactly how it sounded

[NorseHorse]

Originally Posted by Corran

Regarding Exhibit B...
Can you give some specifics on the placement of mics and such? It looks like a couple of AT4040's or similar in wide ORTF but surely that's not all that was used.

The difference in audio quality wasn't necessarily my main point, I was trying to demonstrate how one audio fits the video (i.e. it looks/sounds like they are in the same space) and one doesn't. Both recordings were made with different gear and with different restrictions in regards to where microphones could be placed. Regardless of the gear, overprocessing B could have put it in exactly the same boat as A. And A could have been processed differently to fit more with the video. I say: Sheer audio quality is a separate issue from making sure there isn't a dichotomy between audio and video.

However, to answer your question and to satisfy others who might be curious... Exhibit A was made with two Schoeps with the MK21 wide-cardioid capsules. Exhibit B was made with a Crown SASS mic in the hall (you can see its sillhouette), plus two AT4050s in omni just off the front of the stage (mixed in for a little definition in the soft woodwindy passages). They would have been placed on-stage if set changes had allowed for it.

Happy MLK Day!

[Corran]

Yeah I didn't mean to detract from your original topic, I was just wondering (not many wind ensemble clips around here).

As to your original topic, I haven't noticed a dichotomy between my audio and video when doing A/V production. I try to match what was heard in the hall with the video. But I wouldn't change the balance of my recording. I mean, suppose your video changes from a left view to a right view, would you change the position of your mics virtually between cuts? It seems silly to me. But I don't do video enough to have much of an opinion, though I am getting more into it lately.

[BrianHanke]

I think the problem is simply that we have all, at one time or another, been to an orchestral concert in a theater that looks a lot like the one in the video. And we all know that the sound in such a space is never as lush and present as in clip A. I find clip B to be unrealistic as well, although it is of course better.

I think when there's a visual element we're inclined to expect the sound to behave in a very realistic way. In other words, the sound in a video production should tie in with our past experiences of how a performance of similar type sounded in a similar space.

So, when I watch the videos with the sound muted, I think: this is going to sound dry, it's going to sound a little bit distant (because of where the camera is shooting from) and it's going to be very centered and not have much stereo spread.

But, what I get, especially in clip A, is a wet sound that is all-enveloping and has a massive stereo image. So, in effect I'm getting the exact opposite of what I expect!

I don't know if these comments are at all useful, but I wanted to throw them out there regardless.

(By the way, the recordings in both cases are really, really excellent. )

[rubycakes]

I'd like to echo Corran in saying I wished for a few more discussions of recording Wind Bands (Wind Ensembles, Symphonic Band, Wind Symphonies, Concert Bands) around here. Maybe this is due to the continued second class status bands experience.

FWIW, exhibit A did not fit the video at all. In fact, it almost appears as if the groups was "lip synching".

[XLR2XLR]

I find 48kHz does the trick.

[NorseHorse]

I also record at 48kHz for video projects, even if it's going to CD as well.