Am I Kidding Myself?

[Charles Canary]

I’m relatively new to audio. I’ve spent years learning the other aspects of filmmaking and am now up to audio. I put it last because it was what I was most afraid of. I’m still afraid of it.

I have a little kit consisting of the Akg se300b w/ck93, Sound Devices mm-1, and the Zoom H4. I use Audition and Izotope RX.

I have spent an enormous amount of time on pre-prod for my next film. I know exactly how my film will look. But I'd also like it to sound a very specific way, and I'll twist myself inside out to achieve this.

The sound of the dialogue in this film is what I’m after:

YouTube - Sister my Sister part 2

There is something very distinct about it. It sounds thick and old, and sort of stuffy, and it seems so inextricably bound with the picture that it is hard to image how to replicate it. I have tried so many combinations of things in Audition and just can’t seem to get anything similar.

What I mean by am I kidding myself is, is this kind of audio impossible with my equipment, or do I just not have the knowledge of post manipulation? I’d appreciate some ideas on the post side of things. I’m not looking for any silver bullet, golden arrow, etc., just some thoughts.

You might say the audio sounds degraded because of youtube, but it is still the same basic sound on the dvd. In fact, everyone’s voice in the film and all the sound effects have the same type of sound, the footsteps on the wooden floors, the thick crunching sound of a pin through fabric, etc.. I just can’t put my finger on it.

If there are books I can read, equipment I can buy, tests I can do, anything, please tell me. I fully realize how difficult and how specialized audio production is, and how hard you people work to get where you are. I want to work hard too.

Sorry for the long post. Thanks in advance.

[Chris Rackauckas]

If you boom it correctly and know how to use eqs and compression you can do it. You may want a short shotgun for outside.

[Steve House]

Layers upon layers. Dialog recorded very close so there's minimal room in the voices. Maybe shot on a set constructed on a quiet soundstage rather than on a "found" location. Lot's of Foley, it's very likely that every single sound you hear, from the footsteps on the wood floor to the beans snapping in the kitchen, was recorded and laid in separately and deliberately to build up the final tapestry.

[Chris Swanberg]

Two good books to have are Jay Rose's book "Producing Great Sound for Film and Video" and S. Dean Mile's book "Location Audio Simplified". The latter book can be purchased at his website : Home.

[Charles Canary]

Thanks for the replies. I am still deciding over a shotgun, the Sanken cs-1 is nice, though I may save for the cs-3e. To make something clear, I live in a relatively remote part of Australia. The nearest rental house is 20+ hours drive. That and the fact that I work on my films evenings and weekends means it is more economical to buy my own equipment.

Chris R: We always have the mic just out of frame, and aim it at the chest area. I understand how an eq works, though I don’t know how to sweep certain frequencies to locate and pull out vocals, etc. What is the best way to learn this stuff? I think I’m starting to get a handle on compression. Does the audio in the clip sound very compressed? Apart from eq and compression, is anything else going on?

Steve: Thanks. On closer inspection I see that the sound in one instance doesn’t match the actions, so I’d imagine all sounds are foley, like you said. Would the AKG suffice for recording foley fx? I’ve read a large diaphragm mic is best.

Chris S: Thanks for the recommendations. I’m fairly sure I have the audio gathering side of things covered for my specific gear. It is more the post manipulation I’m after. Do these books cover post production?

The AKG is ok for this application then? Is there anything I’m missing or could improve in the audio chain?

If I were to record a sample of dialogue say next weekend, and link the file here, would anyone be willing to have a go modifying it for me? Conversely, if no one has time, if I post my best effort, could someone critique it? Maybe give some guidance?

Many, many thanks.

[Mike Demmers]

I disagree with some of the previous posters.

I don't think this is a lot of layered foley. The voices are very unfocused sounding, not the kind of pinpoint dead center mono effect you get from close miced voices (or tight pattern mics). The dialog wanders a bit in the stereo image, unfocused.

The effects also sound pretty unfocused, very ambient - again unlike normal close miced foley.

The ambience on both the voices and effects is very short, and consistent. I can hear the backgound noise pumping up and down a bit.

The voices can be heard at varying distances as the actors move.

This sounds to me, like what you would get by stereo micing a sound stage, assuming you had good, old fashioned actors who know how to project a bit, and then compressing the sound.

It sounds like what I have gotten when I did exactly that, at least. ;-)

Or maybe an MS mic setup with a bit too much of the 'S'. It sounds a bit phasey.

I think this is a very low budget production and maybe they either just stereo miced the whole thing, or saved money on foley by recording a stereo pair at the same time as the boom and just mixed it in.

If not, they sure went to a lot of trouble to sound this unclear.

As to the sound not matching in places, that would be caused by using the sound from one take (a master, perhaps) and cutting in other takes (closeups, etc).

This is not very good audio by modern standards. But you have heard a lot of audio like this in older movies, perhaps we tend to associate it with more 'real' because there was less possibilty of messing with the audio in those days - sort of like how old B&W film is perceived as more 'honest'. We don't think it has been manipulated, effected, 'Photoshopped'.

It sounds thick and old, and sort of stuffy ... You might say the audio sounds degraded because of youtube, but it is still the same basic sound on the dvd. In fact, everyone’s voice in the film and all the sound effects have the same type of sound, the footsteps on the wooden floors, the thick crunching sound of a pin through fabric, etc.. I just can’t put my finger on it.

I think this is because they were recorded in the same room, at the same time. If you want to duplicate this, try booming normally, but also put up a stereo pair (to two separate tracks), and then mix in the stereo tracks in post to get that effect. The effects tracks are also compressed, perhaps fairly heavily.

Part of what you are hearing is the 'stereo-ness' of the 'crunching sound...'. And its bounce off the close walls of the set/stage.

It would be pretty easy to try this out at home before committing to your production. Make sure none of the mics is any closer than 15 feet to another mic.

To do this in post, try feeding all the foley and a little of the voice into a very small room reverb with some fairly heavy first reflections. Unlike normal practice, then compress both the original foley -and- the reverb together.

Another thought - something else you might try in the initial recording would be to have the actors play the scene twice, once boomed normally, the second time with them just mouthing the words (no sound) and record the stereo mics at that time.
Sort of do-it-yourself foley. This would pretty well guarantee some out of sync effects just like the one you like. ;-). But the background could then be heaviy compressed without interfering with the voice. ...probably woud not work, too out of sync...never mind that last idea...

A quick look at the IMDB page...hmmm. Low budget, yes. First time director. Interestingly, the sound department is larger than the camera department! Was this because:

1. Green director just KNEW soud was SO important.

Or

2. First time director screwed up sound so badly many people had to be called in to try to fix it...

Pick a likely one. ;-)



-Mike

[Charles Canary]

Mike, how to say thanks... your post is helpful, insightful, and it made me laugh :) I knew, from listening to the film so many times, that there was something odd about the sound. That’s when I got concerned about my equipment, and decided I needed to seek help.

Indeed the film is low-budget, and directed by a first-timer who came from theatre. It’s fairly apparent she had only her theatre experience to bring to the film, so the idea of micing the “stages” with a stereo pair sounds very plausible. The imdb details do list a boom operator, so maybe your notion of them saving on foley is right.

The audio is poor relative to today’s standards, and that’s exactly what I want. I want the film to look and sound like it was made years ago. I agree with you about the old films being perceived as more real. I’m trying to efface any pretense from the film and create something that feels genuine. For me it has always been about the specifics, not what’s considered the best.

Unfortunately, as to your suggestions, I am limited by my equipment at present--I knew I should have bought the mixpre. I don’t suppose there’s anyway of creating a verisimilitude of “stereo-micness” with just one mic and a one channel pre? Gosh, what a ridiculous scenario this is turning into! Could you recommend a set of stereo mics?

Your post, although making the task now seem harder, made my day. Thanks for that.

[Mike Demmers]

I listened to the audio one more time, and learned some more. The frequency response of the background effects changes depending upon whether there are voices present, and their level. This suggests some single ended noise reduction may have been used. This would make sense if the original audio was recorded a little too low level, or distant, and was, as a result, a bit hissy. Taking myself back to the early 90's, that is part of what -I- used to do, among other things, to clean up crappy audio that desperate green producers brought me.

The net result of this is that, when by themselves, the background noises sound rather dull - no high end. When the voices come in, there are a bit more highs, making the voices a little more present.

Listening closely to the delays on the percussive sounds between the two channels, I would guess them as being more than 15 ms, but less than 40ms. So I would guess the (stereo pair) mics were spaced 20-25 feet apart. Or thereabouts.

This is definitely a spaced stereo pair, not a coincident mic placement.

You really need a stereo pair to get this effect. With only one mic, the best you could do would be to back off the mic, roll off the highs, add about 5% tube distortion, and then compress it heavily. But that would sound more like early 30's mono sound, not really this effect.

They should be condenser mics (dynamics would be noisy), but they do not need to be very good ones - in fact really cheap ones might be just the ticket for this. Not so cheap they are horribly noisy though.

You really need to be able to record three tracks separately. Without that capabiity, you run great risk of losing it all.

With your current gear, plus two more cheap mics, I might try this first:

Two mics at the front of the stage, spaced about 25 feet apart, pointed slightly down (not at the actors heads), and straight back at the back of the stage. The idea being to pick up more of the ambient sounds and not so much of the voices. You may have to point them in more, you have to listen to them.

Feed these to the Zoom.

Normal boom. Use your normal setup, feed to camera.

In post - you will have to manually sync the Zoom and camera tracks. And more than once, as they drift off. Plus, you may need to make minor delays/adjustments in timing to make them truly lock together.

Eq a little of the voice presence range out of the ambient mics. There will be voice bleed on these tracks, this will help the main voice track punch through a bit better.

Roll off some of the highs on the voice track. Notice this track you like is not very sparkly, the highs are more than a 1930s track (almost telephone quality) but less than modern tracks. Listen to the S's to match that.

Roll off even more of the highs on the ambient track.

You will need to roll off lows on both tracks as well- notice there are no real lows on the movie tracks, even the music seems to be missing the lowest octave or two. Use the movie as a reference for this.

Now mix the two together, and compress until the background sounds seem about as present as on the movie you like.

Fine tune eq and compression from there. Have some of the movie sound handy as a reference so you do not get off track.

This may or may not do it, it is just what I would start with, a first try.

Another possible thing would be to use the spaced mics as above, but point them at the actors, and use -mostly- them, and less of the boom. You just have to test...

Make tests before you shoot!!!

-Mike

PS:

Make tests before you shoot!!!

PPS:

Remember to make tests before you shoot!!!

[Christopher Glavan]

Mike-

Some incredibly informative stuff here. Very glad to have you on the boards. I hail from P-town, nice to see familiar faces =) Also, just so you're aware- when I'm ready to tackle sound engineering (I mean really ready to commit some time to it) I'm coming up there to convince you to mentor me. If that doesn't work I'll stalk you on all your projects and learn your best tricks anyway!

[Charles Canary]

Mike:

I think I’d subconsciously factored out the Zoom’s XLR’s because the pre-amps are so lousy. They really are noisy on high, and are fairly ineffective on medium gain setting.

By a “cheap” stereo pair, I take it you don’t mean this cheap:

BEHRINGER - C2 STUDIO CONDENSER MICROPHONES PAIR

But how about these:

http://www.music123.com/MXL-993-Cond...26986.Music123

Or these:

Rode NT5-MP Cardoid Condenser Microphone Pair :: Vocal & Studio Microphones :: Microphones :: Audio :: New Media Sales Pty Ltd

Anything specific you could recommend?

I think we are getting somewhere now. What you’ve been able to intuit seems logical when listening to the audio. I’m definitely ready to give it a go.

And don’t worry about my testing before I shoot. I won’t be shooting anything until I get what I want, regardless of how many tests I must do.

Thank you ever so much for your time and advice. As we say in Australia, you are worth your weight in Uranium ;-)

[Gary Nattrass]

I read a lot of the previous comments and I think a lot of them are totally off the mark.

Most if not all dialogue recorded on set and location here in the UK is done with one mic and without any frills at all. a sennheiser 416 (0r 816) is the defacto for drama recording. Sometimes a smaller mic such as a sanken or neuman km is used on a fisher boom but sound poles tend to have the 416 on it.

It is very rare that a stereo or M/S mic will be used at all and this certainly sounds like a fixed set in a studio sound stage. It could be location but if it is then the same recording set-up will be used, the wide shot certainly sounds correct perspective for a 416 on a panamic boom.

I have had a good listen on my sony 7506 cans and what I hear is a 416 in a set but there is also some sort of problem with the transfer, it may be from an optical track as there is some pumping in the background noise or this may be the you tube transfer. Listen to the bit when Julie Walters is laughing and you can hear her going on and off mic as she moves her head, the 416 will be about 18ins above her head at 45 degrees. There has been very little added in post and apart from the servant bell most of what you are hearing is real, it may also be that there is an expander on the dialogue tracks which will also explain the background disapearing.
As for the tonal quality it is usual for dialogue mixers here in the Uk to roll off all of the top end EQ at about 6k and cut the bass at around 150hz. This is a fall out from the optical transfer days where bandwidth was limited and it is still done on a lot of dramas.

Drama here in the UK tends to be recorded very simply and there is a certain discipline to this technique as everyone knows what the limitations are and a very consistent sound track can be attained. Certainly most of the high end costume dramas such as this are recorded this way and the post follows with very little messing around with the raw dialogue tracks.

Hope this helps I have over 29 years experience in UK drama production and most things I have done at a high end have been done on a 416 at the end of a panamic.
A similar sound can be had with a rode or cheaper mic but th rule of thumb is keep it simple and follow the shot framing to get the correct perspective.

[Gary Nattrass]

Just had a look on IMDB and the dubbing mixer was Dean Humphreys, he is one of the top dubbing guys here in the UK and he works at Twickenham sound studios. Size of sound crew was pretty normal for film drama prod in the 90's Chris Munro was the recordist and he went on to some some other low budget stuff??? http://www.imdb.com/name/nm0613101/ NOT!
Guess in recording was a 416/816 to a stereo nagra IV, may have been just a mono nagra with pilot tone. Anyway he will tell you all here:http://uk.rottentomatoes.com/dor/obj...=e7n3fnhe168f3
Two mics seen on the booms are a 416 and a 816 as stated, the 816 is in the blimp that is 3 ft long.

Dean is one of the old skool film mixers and he is vastly experienced so I would guess that a lot of the nasties are from you tube or dolby miss tracks.

His IMDB is here:Dean Humphreys Claire Manning my neighbour's daughter and a good friend of my own daughter is now one of his sound editing team.

Sister my Sister was done in 1994 so was probably done on magnetic film stock as I dont think Twickenham had digital till the later 1990's. Also the age of is will point to a 416 or 816 mc too.

Mag film stock is basically analogue audio tape with sprocket holes like 35/16mm film, the 35mm stock has three audio tracks and will also have dolby SR or A noise reduction, the quality you are hearing is a lot to do with that too as film mag is limited in its audio frequency range.

[Mike Demmers]

I think I’d subconsciously factored out the Zoom’s XLR’s because the pre-amps are so lousy. They really are noisy on high, and are fairly ineffective on medium gain setting.

That's 'H4' not 'H4n', right? Definitely not with dynamic mics, but with reasonably high output condensors, maybe. You have some mitigating factors in place in this case that will help with the noise.

By a “cheap” stereo pair, I take it you don’t mean this cheap:

BEHRINGER - C2 STUDIO CONDENSER MICROPHONES PAIR



Maybe...

I am not the best person to ask about cheap mics. The last time I had any interest in them was over thirty years ago, so I am not exactly up to speed in that area. Thirty years ago we did not have the kind of competition there is in mics now, not to mention cheap Chinese manufacturing. The only time I can remember looking for cheap condener microphones, I bought a pair of AKG 451s, which I think cost me around $300 each in the late 70's. I don't think there was anything cheaper at the time (for true condensers, not electrets).

So to be able to buy a pair of any kind at $99 seems rather miraculous to me.

And they might be fine for this partiucular application. Remeber we are talikng about rolling off highs to create that old fashioned sound, which will reduce hiss.

On the other hand, a nice pair of condensers is a pretty handy thing to have in your audio kit, and you may not always have such a forgiving task for them. So maybe the very cheapest may not be the best choice. My old AKGs still serve me well, I surely did not waste my money - and you won't either.

I doubt you could go wrong with the Rode, it bumps you up a class.

If you decide to go for a less expensive solution try to try several out first - pick the one with the highest output and lowest noise. Unfortunately, the specs on these tend to have many traps for the unwary, due to, shall we say, extremely 'creative' testing conditions, so believe your ears before the manufacturers numbers.

I think we are getting somewhere now. What you’ve been able to intuit seems logical when listening to the audio. I’m definitely ready to give it a go.

Ok...I must once again mention something though, because I have this nagging felling that some of my audio forbears are spinning in their graves, and I just hate that...

Are you SURE it is *this* particular movie's sound you want to emulate? Because its audio really does have some problems that the fine audio engineers even of the 30's and 40's would have gagged at.

You got me interested in this with your question, so I spent some time listening very closely to movies in the 1930 - about 1955 range, trying to put my finger on just what characteristics the sound had that makes it so instantly identifiable.

I don't think you need to trash the sound quite as much as that movie to get the 'old movie' sound you seem to be looking for.

Here is what I think makes up that sonic signature - in order of importance -

#1. Most important, I think this is roughly half the sound. These movies were all recorded on sound stages - and those sound stages had a particular sound - and this is surprisingly consistent between studios. These are large spaces, and normally would have been very mushy sounding, with a reverb time way too long for voice. But these people were running state of the art facilities, they read the research, read Beranek, and followed the acoustic wisdom of the day, which was that for voice, a reverb time of 1/3 to 3/4 second sounded best (today, we tend to just get things as dry as possible and process it later - partially because we do not have the luxury of always recording in the controlled environment of a sound stage). So they treated these stages to get just that.

When you shorten the natural reverb time of a large space in that way, it creates a very specific kind of ambience: a short delay after the sound, then a very short, but very smooth, reverb.

And that effect is on ALL the voices, everywhere. Even in cars (actually done on sound stages), in street scenes, on top of mountains, in outer space...all done on the same stages, with the same reverb. That part of the sound is practically identical from 1930 to the mid fifties.

You cannot get exactly that sound in real locations. Though you may be able to by recording dry and using modern reverbs effects.

2. Limited frequency response. No highs, no lows. You just can't put much of either on an optical soundtrack without it crapping out, and if you did manage to, the theater systems of the day couldn't have reproduced it anyway. This does change over the years though. Jumping from 1930 to 1940 to 1950 is like someone is slowly turning the bass and treble knobs up. Slowly, the technology was getting better. But the gap between 1955 and 1980 is still easily heard.

3. Slightly more distant micing of the voices, more real ambient sounds, less foley. I lump these together because one leads inevitably to the other. They had foley almost from the beginning. But unlike today, not EVERYTHING was foleyed. Just the stuff that had to be, or which was specially important.

4. Crappy limiters with a certain sound, and more actual gain riding than today to try to avoid the problems. Modern gain cells hadn't been invented; the way the variable-mu (gotta love these old terms, did you ever own a triple conversion super-heterodyne radio set ? ;-) ) tube compressors of the day worked was essentially to just overload on peaks producing - when workng normally - 2 to 5 percent distortion. I never realized it until I really listened, but you hear this practically every time some actor gets loud, or laughs. It's not too offensive, because this is good old tube second harmonic distortion, but it is everywhere.

5. Yes, there was some phasing sometimes when actors got too close to a set. Yes, sometimes they got a little too far from the mic and sound a bit off-mic. But they knew these were problems, and the better productions had much less of this sort of thing than the cheap ones. Your reference movie has a lot of this sort of thing, but while it is there on older movies I don't think it is the biggest key to the old fashioned sound.

The technical people that made the sound of these movies were limited by their technology, much more than their knowlege. Mostly, they were not making the kinds of mistakes that green directors make - that have to be fixed in post.

-Mike

[Mike Demmers]

Gary - There is nothing like direct experience in that market to vastly improve the quality of your guesses! ;-)

Now that you have explained the usual procedures there, I am still curous if you think there might be any possibility that a green director, from a stage background, might have over ridden their sound person (I looked this crew up on IMBD too, before I commented) and had some more mics up to catch ambience. This drama has a lot of long spaces where the ambient sounds are all that is happening. I swear I think I hear some stereo in there on at least some of the ambient sounds.

But I suppose you could be right and that is from some decoding problem too. I only know what I hear, which is pretty unfocused. I actually swapped the phase on my speakers at one point to make sure the problem wasn't on my end. ;-)

dolby miss tracks Unfamiliar term to me, does this mean 'Dolby decoding errors'?

It's interesting how people do things differently in different countries/scenes. You say it is usual for dialogue mixers here in the Uk to roll off all of the top end EQ at about 6k and cut the bass at around 150hz. implying this is still the case - I never would have guessed this was still being done, so assumed it must be to fix something.

Ask any two audio engineers a question, you'll get at least three different answers.

I wrote the previous post before reading Garys, but I think this supports what I was saying in it - whatever is causing the problems in that movie, whether Utube or dolby errors, or cosmic rays from Planet X, trying to duplicate certain of them may not be the way to go.

-Mike

[Gary Nattrass]

Unusual that a new director would overule the sound crew if they are more experienced, the fact that the recordist went onto do five bond movies inc casion royale and quantaof solace show his experience and he is one of the top guys in the UK.

Dolby miss tracks are very common on old material and if a 1db line up problem is copied it then becomes a 2db problem and so on, it will make the audio pump and cause the type of problems that you also hear with bad encoding in the digital world.

The roll off at top and bottom end was very common in the days of MAG and optical sound tracks as the freq response of the systems was very low and in the case of optical it was better to limit the freq content due to the optical printer working on percentages as opposed to db's. That way you got the max amount percentage of signal on to the track without any spurious frequencies affecting the transfer.

Most dialogue in dolby surround tends to be made narrow band so as to isolate it on the centre speaker leaving full bandwidth for the fx and music.

As for the you tube track I have listened to it several times and it pretty much sounds mono to me but with decoding problems due to the you tube crap bandwidth.
Any stereo or mono ambient sounds tend to be added in post, I have even heard noise gates or expanders used for dialogue tracks with a constant buzz track added to give a constant noise floor, this can be done more on period dramas as you need to get rid of distant traffic etc.