Bruce Watson
Forum Replies Created
-
[Gary Angle] “…cannot find my way through the maze of recorders and methods to acquire good audio.”
Certainly understandable. Audio is just as complex as video, but everyone wants to concentrate on the video side. What they don’t understand is that the audience will stand up and walk out on bad audio, while that same audience will stay and watch all kinds of crappy video. But I’m probably not telling you anything you don’t already know.
[Gary Angle] “I’d like to ask for help in how to get really good audio that I don’t have to spend lots of time in post fixing!”
I’ve been where you are now. I spent years learning audio for video. Not that it made me any kind of expert. But I can tell you what I found out, which you might find useful. Or not. IDK. YMMV. Use at own risk. All that.
First, mounting any kind of microphone on camera is the road to truly crappy audio. The optimum placement for the camera is hardly ever the optimum placement for a microphone. 99% of the time it’s an extremely bad placement for a microphone. Which is why so much “run and gun” video has such bad audio.
Second, proper mic placement often devolves down to signal-to-noise ratio. The dialog you want to record is your signal, everything else is noise. You control this by how close you get your mic to your signal source. That is, how close you get your mic to the talent’s mouth. Typically, you’d want a hypercardiod no farther out than 60cm. Out in front of talent, above the frame line, pointed down at talent’s mouth at around a 45 degree angle. This is my “sweet spot”, for better or worse.
If you want to think of it another way, getting in close lets you maximize the sound of the voice while you minimize the sound of the room. If you get in close it usually sounds pretty good — if your mic is way back (mounted on camera? Do not do this) you get as much room as you get voice, so you get that boxy “small room sound” because you’re brought up all that noise like the rapid reflections of the voice off the room surfaces. And that drives understandability way down into the “crappy sound” area. Not good. Do not go there.
Third, the phrase “fix it in post” should be banned from the language. It’s nearly impossible to fix bad audio in post. Capture the audio correctly in the first place if you want to avoid the “nightmare post” audio experience. The tool for getting it right in the first place is a good pair of quality headphones worn over the ears every time you record anything. Anything at all. As Richard says, recording audio without monitoring on headphones is like recording video without looking through the viewfinder. But it’s not enough to wear the headphones — you have to learn to listen. It’s way too easy to put those headphones on and then ignore them. Train yourself to listen.
Forth, almost all cameras have cheap and nasty audio sections. That is, really pretty bad microphone preamps. And line level inputs that are really just putting an attenuator (aka “pad”) on the signal then routing the attenuated line level signal through the mic level parts (and those nasty mic preamps) anyway. So… to fix and control this, you need a field mixer, even if you’re still going to record to camera. The one I use for this duty is a Sound Devices MixPre-D. There are bunches of them on the used markets for less than half of retail. They give you three major advantages over your camera audio section. These being excellent mic preamps, excellent meters, and excellent limiters. There are other advantages as well.
Fifth, most people should not own a shotgun mic. Shotguns can be tricky to use (especially indoors, if you don’t know how, leave the shotgun outside). Instead, your first mic purchase should be a hypercardiod, which can be used indoors or out without much problem.
————————————————————————————————-
So… what do you do with a mixer? First you have to find out what the mixer’s meters are actually showing you. The MixPre-D’s meters are showing you dBU, which is not at all the same as the dBFS your camera is probably showing you. So you have to find out what you’re looking at. When in doubt, read the manual. If you’re using a MixPre-D, it can send a calibration tone to your camera. Do that, and set your camera to around -20dBFS. Then, don’t ever touch the camera audio levels again. Do all your changes on the mixer. No, I’m not kidding.
When it comes time to record your video / audio, get your microphone in place (using a boom op, boom pole on a c-stand, mic talent up with lavs, whatever you’re using) then have your talent “rehearse” a bit to give you a sound check. Use this to set the gain on your mixer to the equivalent of around -12dBFS peaks (on a dBU scale that’s around +8dB (and no, that’s not a typo, that’s a plus sign). +8 dBU is the first red LED on the MixPre-D. I recommend leaving the limiter set at factory defaults (for a MixPre-D that’s +18dBU IIRC). You do it like this, your setup is way closer to un-clipable than you might think.
As to mics, the small documentary crowd still seems to like the Audio Technica at4053b, or the Audix SCX1-HC. There are of course dozens of similar mics that will do the job. But these two keep coming up in the forums I watch. Make of that what you will. I’ve actually got a 4053b; it’s a great mic for the price. I imagine the Audix is too, but I’ve never used one.
On the lavalier side, most people go straight to radios (wireless). That’s almost always a mistake. Radios are a last resort, not a first. If you can go wired, you should. There are three reasons: cables sound better than even the most expensive radio, cables are way more reliable, and oh yes, cables are way less expensive. And if you think ahead and buy an XLR-plugin power converter, you can run your lavalier off the phantom power from the mixer. Which is convenient.
The other problem with wireless lavalier kits is usually the mic itself. Sennheiser for instance is known for merely adequate kit mics in their bottom end wireless kits. At least the gen-2/3 kits. I hear they upgraded the old ME-2 kit mic for the gen-4 kits, IDK. Anyway, the biggest bang for the buck as it were still seems to me to be the Oscar Soundtech 801 or 802 lavs. When I bought a pair for my old Sennheiser G3 wireless kits I got large step up in audio quality. About as much as going from camera mic preamps to the MixPre-D.
BTW, never use a directional lavalier. Only use omni lavalier mics. This too will save you from many hours of post-audio hell.
So, there’s years of experience crushed down into a single post. Clearly there’s a lot more, but maybe this will get you pointed in the right direction.
-
For acoustic music in general (that includes singers) the general order of importance of what effects the quality of the recording is often seen as:
1) The space you’re recording in
2) The quality of the artist
3) The quality of the artist’s instrument
4) The quality and experience of the recordist
5) The equipment usedI’m assuming that you aren’t recording in a small highly treated studio, because if you were you probably would have said so. So what I say below assumes you are not in a studio.
Recording singers in a small space is going to result in “small room sound” (due to the early reflections and room modes). You can help the sound some by recording in room that has carpets, full book shelves, thick curtains for windows, upholstered chairs, etc. Basically, you want to break up (diffuse) the sound reflecting off the ceiling, walls, and floor. This only really works for mid to high frequencies; low frequencies generally require bass traps. That said, don’t expect miracles. The only thing that sounds like a performance hall is a performance hall.
As to mics, I recommend the Line Audio CM4 (cardioid) or OM1 (omni) for your purposes. They are very inexpensive as these things go, and sound way better than what they cost. The downside is that they aren’t very sensitive or very quiet. Neither of which will matter much in your situation, and may actually be a benefit.
As to recording mono, I don’t recommend it. If you record stereo, you’ll probably use it, but if you really want mono you can just drop one of the mics out of the mix, or you can sum to mono (depending on the stereo system you used). If you record stereo you get that flexibility. If you record mono and then people want stereo, you’re just out of luck.
-
[Ty Ford] “What you have here is audio that was recorded with the talent too far away from the microphone. The result is this “roomy” sound. It may have also been data compressed, but your main culprit is bad acoustics.”
Yup. I concur. Sounds like audio from a camera. And as we all know, the optimum placement for a microphone is hardly ever the optimum placement for a camera. Yet people persist in placing mics on top of cameras, and end up with sound like this.
There are software packages out there that aim to reduce the reverb (small room sound) from audio like this. Ty has pointed you toward them. If that helps, you can then use some EQ to bring the levels up on the higher frequencies to restore some of the consonants in the speakers voice to make him more understandable, and to bring back some “air” so it doesn’t sound so suffocating. Sometimes with sound like this a little compression helps too; experiment with it if you want.
To prevent this in the future, get the microphone closer to the speaker. The easy way is to clip an omni lavalier on him, as close to the sternum as possible. This is especially effective if the speaker is roaming around while talking, as a teacher might. If you can confine their movements (sit them in a chair, for example, a sit-down interview) then you can use a condenser mic on a boom pole on a c-stand (easy and effective, and which usually sounds better to my ears unless the room is really bad).
That said, it’s never likely going to be good audio. But you can make it better than it currently is.
-
This kind of thing has been researched. You can search around and find the peer reviewed papers in journals. What researchers have found is that viewers don’t like for the sound stage to move. Some even get physically sick (that usually takes tracking the sound with a complex camera move) sorta like sea sickness. It disturbs their equilibrium somehow. There’s a good reason movies use a center channel for dialog and don’t pan the dialog left and right to follow the actors.
Flipping the sound, swapping left and right, is about as disorienting to the viewer as having a camera angle hopping back and forth violating the 180 degree rule. When used for this effect it’s OK. When not, you’re just disorienting the viewer for no good reason. Don’t expect the viewers to appreciate it. Just sayin’.
Said another way entirely: The reason not to do this is that it’s a big and sudden change in the sound. This pulls the viewer out of their “suspension of disbelief” so that they quit listening to the music (which is the point of the video, yes? The music?) and start listening to the sound itself to try to figure out what suddenly changed. So not only are the musicians not on screen while you’re showing the audience, now the music isn’t being heard either. You get a complete break of continuity. I’m thinking this isn’t really what your client wants.
But you never know, and clients pay the bills. If you can’t reason with him, you can always do it both ways and let him pick his favorite. If he picks “wrong” have him sign a waiver absolving you of the responsibility and do it his way.
-
[John Simon] “Suddenly, it stopped getting the audio signal, even tho the RF indicator on the receiver shows a 100% signal. There is no signal indicated on the receiver’s display for RF. Also, the “Mute” indicator on the receiver is always on, even tho I can switch it on and off on the transmitter.”
Could be something like a new cell tower powered up. The 600MHz band isn’t ours anymore, which makes your old G2s illegal at some point, I don’t remember the dates for the change over. Regardless, a cell phone signal might cause the symptoms you are experiencing. I’m quite sure the Sennheiser receiver would not be able to interpret a cell phone signal. If I were writing the code, I would certainly mute the device in the face of an unknown signal.
-
This kind of audio editing is relatively painless with a modern DAW (digital audio workstation), everything from Reaper to Samplitude to Pyramix all make this kind of editing easy. Even the audio packages that come with NLE (non-linear editing) software like DiVinci Resolve and the Adobe suite can do this.
That said, the iZotope RX family is highly regarded and aimed at doing audio repairs. One would think this kind of repair is exactly what they do. So I’m relatively sure your downlevel version should do this also.
As to how it’s done in your software and your version, I have no clue. I’ve never needed what iZotope has so I’ve never used it. But if you google around for a tutorial I’m sure you’ll find step-by-step instructions out there somewhere.
If you can’t find a suitable tutorial, you might want to ask in the Creative Cow Audio Professionals forum.
-
[Sergio Cabrera] “I know audiojungle, but their licenses is not what I’m looking for, I want to pay once per file and use the audio files freely in several renders.”
I believe that’s called “royalty free” music. It’s everywhere once you know what to search for. My old fav was the Music Bakery.
-
[Chris Wright] “I was hoping you could create 3-d sound in realtime like how they used to make audio for cartoons with 5 guys standing around each other making sound effects.”
Well, if you could do that, what you’d record would be the work area behind the camera. Which is full of people and equipment. It’s got the DOP and his assistants, the director and his assistants, lights and light controls, people to set them up and control them, etc. Don’t expect these people and their equipment to be perfectly silent, they will not be.
Typically on a film set, when you see a room on screen it’s only two walls and a floor. So if you record what the room sounds like, it sounds huge (because it’s not really a room at all), which doesn’t sound at all like what’s shown on screen (which is, say, a room in an office).
It’s not like this stuff hasn’t been tried before. It was tried when movies expanded to stereo. A lot of things were tried. And rejected. This is why we have a center channel for dialog, and why dialog does not track the actors. Because dialog tracking the actors does not work for the audience. There’s plenty of research and papers published in peer review journals to prove what they guys discovered back at the time. Google away; most of it is on-line now.
You don’t want to accept the wisdom of those who came before us? Fine. It’s easy enough to learn this yourself. Go try it. Setup a surround sound rig on set and record everything. Try different techniques, from double M/S to IRT cross (two compact systems, there are many more to choose from). If you like what you get, use it. As Duke Ellington used to say “if it sounds good, it is good.”
But please report back what you find and what you decide — you aren’t the first to ask these questions, and you won’t be the last. Let other people learn from your experience.
-
[David C Jones] “some sort of system or interference noise.”
That’s not much to go on. Can you post a sample? Or at least make an attempt to describe what you hear? And how/where are you mounting the mic? This, mounting, is often where the problems come from. The COS-11D is an excellent mic that hears everything, and that can be a problem. For example, clothing rustle can sound like EMF interference at times. If you don’t mount the mic with a proper strain relief loop the handling noise (cable microphonics) can sound like HVAC rumble at times.
Better mics make us raise the level of our games. I’m just sayin’.
[EDIT]
It’s highly unlikely that the COS-11D and the Sennheiser kit mic have the same sensitivity. So make sure you preform proper gain staging by setting your transmitter’s mic gain to match the mic if you haven’t already. And of course set the receiver to handle the signal coming in from the transmitter.
[/EDIT] -
[Chris Wright] “I thought maybe you could get some good foley or dialogue this way.”
First, Foley is a man’s name, so it’s capitalized.
Dialog is nearly always recorded in mono, and recorded close to the speaker’s mouth to maximize the signal-to-noise ratio. IOW, recordists get in close to maximize recording the speaker’s voice and to minimize recording any local noise as much as possible, such as reverb, ambiance, reflections, whatever you want to call the sounds around the speaker. Reverb will be added to the dialog in post, to make the dialog sound like it was spoken in the space depicted in the film, not the space in which it was actually recorded. That’s an important distinction.
As to surround sound… practically the only sound that’s recorded on set is dialog. So there’s no need for stereo recording, and no need for any kind of surround sound recording on set. For example, consider the scenes in bars, restaurants, etc. you’ve seen in movies. All those people in the background of the shot are silent — they are acting, not talking. It can be rather eerie to film scenes like this; silence except for the dialog, but that’s how it’s generally done. The sound of the extras in the background is called “walla” and is recorded separately, typically on a sound stage and not in the bar/restaurant/club set. Because again, it’s about supporting the story being shown on screen; it’s not about sounding like the film set.
Basically everything you hear during a movie, other than the dialog itself, is done in post. All the non-dialog sound, from footsteps to the creak of door hinges, to the sound of that airplane flying overhead or the dog barking out that window, to music, to the added effects and reverb to make the audience believe the sound came from the space shown on-screen, all of that is added in post. So nearly all of the stereo considerations, and all of the surround sound, and Dolby Atmos considerations, are done in post. Almost none of that is being recorded while the scene is being filmed.
With any rule there are always exceptions. The exceptions here include “live” shows (think game/talk shows on TV), sports events (like the just finished French Open tennis tournament which used an impressive array of mics, each in its own wind basket with furry cover), parts of movies that include live events, concert films, especially classical music in historic spaces (often well over 100 mic channels used), that kind of thing. Some documentary work maybe. But narrative film work still tends to be very much dialog only recording on set.