Category Archives: Location sound technique

664 Media Management

Just in case anyone was wondering, I thought I’d outline my typical 664 workflow as it relates to media management.

The Sound Devices 664 is kinda particular about the media cards that it likes, and  while SD and CF cards are quite common, there are only a handful that work in the 664. For this reason, I never release my cards at the end of a shoot… I wait while the DIT copies them to a hard drive. If there’s no DIT working on the set, I’ll copy the files myself to whatever is available. It’s fairly rare for a photographer or producer to show up without a computer, but I always try to bring mine along, just in case.

My 664 media department... cards, case, and reader

My 664 media department… cards, case, and reader

When I first bought my 664, I went out and bought three of each… three 16GB Delkin CF700x UDMA6 CompactFlash, 105MB/s read, 67MB/s write, and three 16GB Delkin SD163x Class 10 Secure Digital, 24MB/s read, 17MB/s write. I also picked up an inexpensive card reader and a case for the cards. One thing you DON’T want to do is use cards that are not on the list of approved media for use by Sound Devices. Go to their website, the list is updated every so often.

These are the cards that I use in my 664. They work fine and are approved by Sound Devices.

These are the cards that I use in my 664. They work fine and are approved by Sound Devices.

I have my 664 set to record the day’s audio files mirrored, so that each card has identical audio files. I use the SD card to transfer the data. It’s difficult to get my fat fingers around the edges of the CF card, plus the CF slots are a little more delicate… I’ve already bent a pin on my card reader, and if that happens on the 664 then it’s back to the factory for a very expensive repair. So the CF card stays in the machine, and I treat it like an internal drive most of the time unless I get some sort of data error on the other card (hasn’t happened yet, knock on wood!)

Transcription recordings are always the fly in the ointment. These are often requested as MP3 recordings with linear timecode on one channel, and audio on the other. It’s possible to do this internally with the 664, but you’ll need a special cable and an open channel. If you patch the timecode out to, say, ch6 input, then you can assign that signal to whatever card will record the MP3. Because, I’m such an old fart, I’ve always been a bit nervous about doing it this way. Back in the day, we had lots of trouble with linear timecode signals bleeding onto other tracks, as it’s generally very hot. If this happens to the main audio tracks, then you’re screwed.

Alternatively, you can use an entirely separate recorder, and that’s how I did it on a recent shoot. I used my Sony PCM-10, and built a special cable for it. The cable has a 3.5mm stereo plug on one end, and the other has a fanout with a single 3.5mm plug and a BNC connector. The BNC gets the timecode and the 3.5mm goes to the 664 tape out. If the cable is built normally, then the timecode signal will be really hot while the audio signal will be really low. I added a teeny resistor inside the connector shell to drop the level of the timecode signal, and get the levels on each channel to match a little more closely. It worked like a charm.

The Chew at the Bluebird Cafe

Here’s a quick couple of photos from a recent interview for ABC’s The Chew with Carla Hall (ex Top Chef runner up from season 5). Carla interviewed two of the principal cast members  from the upcoming TV series Nashville, Clare Bowen and Sam Palladio.

Carla Hall interviews Sam Palladino and Clare Bowen for ABC’s The Chew at the Bluebird Cafe in Nashville, TN

While the interview itself was fairly straightforward, it was a bit whacky in terms of the interconnection. Three wireless and an overhead shotgun going to two cameras… simple, right? Except the producer wanted separate iso tracks on everything and I don’t have a multitrack for location work (yet… it’s on order this week).

The 442 setup with direct outs to the camera. The bag construction doesn’t allow access to the direct out connections, so I had to disconnect everything, take it out of the bag, and reconnect on the table… not something you want to do in a hurry.

Once I determined that we weren’t going to be moving around, I was able to feed two wireless signals to one camera, then one wireless and the shotgun signals to the other using my Sound Devices 442’s direct outs The downside was that the 442’s direct outs are pre-fader, so the only way to adjust levels was at the camera. Not ideal, but it did work and the client was happy… but it illustrated the real need for a good multitrack location recorder. Right now I’m going against conventional wisdom and considering Tascam’s HS-P82 over the Sound Devices 788T or the Zaxcom Nomad. While I generally prefer Sound Devices gear as a matter of course, I like the Tascam’s feature set a little better (8 XLR inputs rather than six, for example) and their pricing is a lot easier to swallow at $3,500 (vs. $6,000 for the Sound Devices). Since the 788 has been out for awhile now, I have a feeling that they’ll be announcing a “new and improved” 788 sometime soon, and the Tascam might fill in the gap until then.

A DSLR Timecode Workflow

Most of the DSLR shoots I’ve worked on have been fairly simple affairs, at least in terms of synchronizing the sound… mostly because there weren’t any other options. These shoots are done old-school “double-system,” where the camera records the visuals and the audio is recorded with a separate recorder. We used to use a Nagra for this (which is still a desirable piece of gear… one day I’ll find one that I can afford.) A standard slate is used at the head of each take to provide a marker for the editors. They line up the little bar of the slate with the crack of the slate’s clapper on the waveform, sliding the audio around until they match, and then the audio is locked to picture. Many folks are using PluralEyes to speed up this process. But if you’re doing a long-form production with a lot of takes, manual sync can become a fairly large chore.

The Ambient Lockit Buddy in the hotshoe mount

I did a shoot recently where the production house supplied an Lockit Buddy (built by Freelance Audio Visual Services in the UK) along with an Ambient Clockit Box (we used an ACL 203) timecode generator. Ambient timecode generators have been around for a long time, and are proven pieces of kit… though they are expensive little morsels. What’s new is the Lockit Buddy. This is  a small passive device designed to drop linear timecode down to mic level, allowing it to be recorded on the DSLR’s right channel. You can also feed a reference audio track to go on the left channel, if you have an extra wireless or line level out. The Lockit Buddy doesn’t generate its own timecode, and requires some sort of timecode signal for it to work… either generated as it was on this shoot, or a feed from a timecode out from a timecode-capable audio recorder (My Edirol R4 Prohas a BNC timecode out that would be perfect for this application, but the DP didn’t want to be tethered to my recorder) or a video camera.

A typical Lockit Buddy setup. Note that there’s no provision to mount either the wireless receiver or the timecode generator… not ideal for handheld shooting or tight production schedules

In use, the Lockit Buddy worked pretty well. We would jam sync the recorder at the beginning of the day and then disconnect the two (with both the recorder and the master generator running in “free run,” of course). The generator and lockit buddy combination was a bit cumbersome on the camera. The Lockit Buddy comes with a hotshoe mount, but not the lockbox, so we had to tape everything on top of the camera. The assembly was  constantly slipping out of the hotshoe mount. Because of the mounting difficulty, we didn’t use the wireless receiver on the camera. But it did work. I’ve even got my own Lockit Buddy on order, as they’re not too expensive (around $140).

setting the sound recording level manually is required for the timecode signal to work, though the constant level of LTC would likely override the audio gain setting of the camera. Sudden loud bursts of noise may cause the level to fluctuate in auto mode, however, causing potential sync problems.

Unfortunately, the post work is being done in the UK, so I can’t report how seamless the system is on the backside. I understand they “had some trouble syncing up the takes,” but they didn’t elaborate. Still, if a scratch audio track is fed to one side, one could still use PluralEyes if the LTC signal somehow failed to lock up. Using a slate is, needless to say, mandatory. And there is more info about the post process on the Ambient website.

While I like the Lockit Buddy, there are some improvements that I’d like to see… maybe in version 2.0. A full-sized XLR connector with a mic/line selector switch would be a big improvement, and wouldn’t require any active circuitry.  This way a DP could plug in a basic mic for an audio scratch track, without requiring a tether to the soundperson. The TA3 connector means that special cables must be built to get audio into the camera, and an additional 3.5mm female jack would be highly convenient. I may do that modification myself, once mine arrives.

How To Make Turtle Clips

I can hear you. You’re saying, “What the #%*& is a turtle clip?”… a reasonable response to the above headline, actually. I didn’t know about them myself until recently. A “turtle clip” is a mic clip with a little wire bail over it. The wire is there to keep fabric away from the mic, preventing clothing rustle.

A commercially available turtle clip for a Tram lav

If you’re flush with cash and short on time, you can buy these from B&H for $20 apiece. Or you could make them yourself. All you need are some bits of copper wire, something round (like a dowel or broom handle) to use as a former, a pair of needlenose pliers, and a pair of cutters. Optionally, you can use a soldering iron if you want to get fancy, but it isn’t required. The clips that I made require a bit of Topstik or Moleskin to secure the mic, but they work as well as the ones from B&H.

Any round object will do as a former. The size determines the diameter of the finished mic clip… about a half-inch or three-quarter-inch rod will do. Wrap the wire around the former as if you were winding a spring one and a half turns. It’s easiest to wind a little extra and then clip off the excess wire later.

Starting a turtle clip. Copper wire is wound around a round former, 1 1/2 turns

Now, using the needlenose pliers, bend a half turn of the wire up 90 degrees. Clip off any excess wire and you’re done, unless you want to solder the ends. Soldering only takes a second and does make the finished clips a tad stronger and smoother, but it isn’t required. If you haven’t done much soldering before, this step could be

Bend one of the loops 90 degrees. I used a vise, but pliers can work fine for this.

a little tricky, but it’s very easy with a little practice. (HINT: Heat the wires first, then apply the solder to the wires. Don’t apply solder to the soldering iron, except maybe to help transfer the heat to the wires.)

Once you’ve made a few of these, other variations and improvements will surely come to you…

Snipping off the excess wire. Having a little extra wire makes handling these a touch easier while you make them.

making clips with two bails, for example, or rectangular shapes rather than round. They will get rather bent out of shape with regular use, but they’re easily replaced and cost only a little time. And if you’re too busy to take the ten minutes it takes to make these, I’ll be happy to sell you some that I’ve made… only twenty dollars each!

My collection of homemade turtle clips. They didn't take long to make, and they get easier with a little practice.

My Big Fat Greek 4-Track Setup

OK, so I’m sitting in a hotel room in Nashville, waiting for the client to arrive. It’s 8:51 PM local time. Their plane lands at 9:30, and they possibly want to shoot some tonight… depending on the party that the band has scheduled.

The client is Record Collection Productions from California, and the subject is a band called Jeff The Brotherhood for Red Bull’s Sound And Vision.

It started out simply enough… they requested two lav mics, a boom, and a 4-channel mixer. A straightforward, normal production setup that I’ve done many times before. But yesterday I was told that the post house HAD to have iso tracks, and they HAD to be timecoded. No problem, I’ve got an Edirol R4Pro timecode-capable 4-track. I’ll just bring my cart, and… no, you’ve gotta be very portable and run everything from a bag.

Naturally I thought I could rent my way out of this problem with help from my pals at Trew Audio. But no, their 788 was already out, and their Zaxcom Deva was parked in Glen’s rack for a movie he’s working on. So it’s back to the drawing board. And by the way, they need a wireless timecode slate, too.

The R4 is a great recorder, but it doesn’t handle mixing duties well at all. In fact, about all you can do to monitor is listen to tracks 1 and 3 in the left side of your headphones, and tracks 2 and 4 on the right… no mix outs, no solo monitoring… it’s a recorder, not a mixer.

The solution that I came up with was rather simple. Since the primary deliverables are the iso tracks, I’m running two Lectro receivers into tracks 3 and 4. The boom mic is hardwired to a splitter, with one output going to track 1 of the recorder, and the other output going to a Lectro transmitter. The receiver goes to the camera, so they’ll have a wireless link with the shotgun on one track.

The 4-track capture setup. Sennheiser ME66 shotgun, wireless timecode slate, R4Pro in a Portabrace with a second bag attached to the top containing 2 transmitters and 2 receivers, a Lectro 195 receiver for the camera, and a run bag with batteries, grip tape, extra cables, etc

In a perfect world, I wouldn’t use a setup like this, since a mono mix to camera goes against my grain a little, but this is about the best solution that can be had with the equipment that is available. We’ll see how it actually shakes out.

More Fun With 5.1… The Double Mid-Side Technique

After my last post on an inexpensive 5.1 workflow, a reader (Peter Tooke… thanks, Peter) advised me to look into the double mid-side technique. This is a method used for 5.1 surround sound recording that I’ve seen used on feature films, and while it can be a little complex to wrap your head around, it’s pretty neat.

The “mid-side” mic technique is a method of recording stereo using a cardioid mic facing forward (the “mid” mic) and a bidirectional, or “figure-8″ mic facing left and right (the”side” mic). These mics require a little processing at the mixer, though, before we get a stereo signal.

Schoeps has several hardware options to support a double mid-side setup

The mid mic is panned to the center, but the side mic is split into two equal signals. You can do this analog using a transformer-balanced splitter, or digitally by copying the input to duplicate channel. Pan one of these mid signals hard left, and pan the other hard right, but… and this is the key… invert the phase on the right channel.

The beauty of the mid-side technique is that if the left and right channels get summed together for a mono signal, you get no phase cancellation between the two signals. We used this technique in the early days of television, when some of the sets were mono and some were stereo. Using mid-side stereo, both stereo and mono televisions sounded fine. (More details about the mid-side technique are here.)

The double mid-side technique is an expansion of the mid-side technique. By simply adding a rear-facing mic we now have a system that can generate five channels of audio using only three tracks. Use a shotgun for the main mid mic, add a bidirectional mic for the side signal, and a cardioid rear-facing mic. The side signal is shared between the front and the rear mics. Each mic is recorded on its own track.

Decoding the three signals into a 5.1 mix is straightforward. The shotgun becomes our center channel, shotgun plus the mid (split, panned left and right, and phase-reversed for the right side)  becomes our left and right front signal,  rear cardioid plus the mid (split, panned left and right, and phase-reversed for the right side)  becomes our left and right rear signal… pretty neat, eh? And like the standard M-S stereo technique, the signals can be summed to mono without phase error.

The Schoeps double mid-side recording solution has the advantage of the entire rig being able to fit inside a zeppelin.

Shoeps has a good bit of information about double mid-side technique on their website, including some elegant hardware solutions for implementing this system in the field. Hang on to your wallet, though, cause this ain’t a Chinese knockoff… it’s some serious hardware that is priced accordingly.

Schoeps even has a free double mid-side plugin that decodes the three signals for a double mid-side system. Order it here.

A screenshot from the free Schoeps double mid-side plugin

A Simple 5.1 Surround Workflow?

I received a call awhile back from my Director, Tim Coghill. Now, I haven’t worked for Tim in over 25 years, but he was, without reservation, the best director I’ve ever known. Period. I’d trust him to talk me through diffusing a roadside bomb.

Tim is the production manager at a SC ETV network station in Spartanburg, SC, and had just finished up a documentary for them. Until recently, PBS National had a requirement that all programs submitted to them must be mixed in 5.1 surround sound (they’ve backed off that requirement as of late).

What tim was asking me about was a simple way to capture 5.1 audio in the field. I know of lots of approaches to 5.1, but none of them could be called simple… other than mixing in plain old stereo and letting a 5.1 synthesizer do all the heavy lifting for you. I understand the newer ones do a pretty good job, too. But Tim was looking for another way, and even after 25 years, when Tim gives me a direction, that’s where I’m going.

I started ruminating on his idea of a simple, inexpensive method of capturing surround sound. I’ve come up with almost a solution. It’s far from perfect, but I figure it’s better than pocket lint.

At the core is the Zoom H4n. I’ve got one & use it a lot. I don’t love it, as it’s consumer-grade gear, and is loaded with compromises. But they’re small (it’s a handheld) battery powered (though it eats batteries like candy), and relatively cheap, about $300 street price. So far so good.

The Zoom H4n- it's ability to act as a pseudo-4-track makes it the key to this workflow

This little piggy has a pair of XLR inputs in addition to two onboard mics, and it has a 4-ch mode that allows you to record from both mics AND the XLRs. (Almost none of the commonly-available digital recorders can record more than two tracks at a time.) So you have a fairly cheap pseudo 4-track to work with. Six channels would be better, but let’s start with this.

Let’s imagine a single-person interview in a space with lots of background action… a deli, let’s say. You set your talent down and point the Zoom AWAY from your talent. The XLRs can be used to record a lav on one channel and a shotgun on the second channel.

The two tracks from the onboard mics are sent to the L/R rear channels. You’ll want to do some field testing to figure out which is the best way to orient the H4n… the soundfield might sound too extreme with the mics pointing away, and it may sound better if the mics face the talent, only placed some distance back.

The shotgun will likely pick up more “room” than the lav. Run the shotgun through a stereo generator and send that signal to the front left and right speakers and put the lav on the center channel. You’ll want to play around with this… the center channel might need a little shotgun, and the L/R fronts might need a little lav. But keep checking for phasing errors by summing everything to stereo and mono once and awhile to make sure you don’t get any really whacko-sounding phase cancellations. Some phasing will be unavoidable, but keep monitoring it just the same.

As you already know, the “.1” part of the 5.1 mix is the subwoofer. Pull these low frequencies wherever they sound best, perhaps the lav if you want to accentuate the speaker’s voice. Other sources might sound better if you want to accentuate the low frequency component of the environment… if you’re shooting on a battlefield, you’ll want those deep, rumbling explosions.

If you’re shooting two people on camera, then you’ll have to decide between two lavs, or a shotgun and a mixed lav signal, or just a shotgun and a single mic. Perhaps the single mic could get the rear signal, and use the zoom mics for the L/R fronts. You’ll need to experiment, and– here’s a crazy idea– use your ears and judgement.

The H4n is small enough that it can fit on a pole or a C-stand. Sometimes placing near a boundary surface, like a table, can give you a little boost as long as no one uses the table to shuffle scripts while you’re shooting. Camera noise will be an issue if you’re shooting a Red or an F900.

It’d be best if you did a dedicated stereo mix on a separate tape, and a 5.1 mix on another. This way you could avoid the whacko phase cancellations on your stereo mix, But rest assured, somebody somewhere will be listening to your 5.1 mix on only two speaker, so keep checking your mix folded down to stereo frequently.

This is about as close to a simple solution as I can come up with. One big drawback is the H4n records at 44.1/16 bit when it’s in 4ch mode. (I think it does, anyway… don’t have the manual in front of me at the moment.) This isn’t a broadcast standard, which unofficially has been decided should be 48K/24bit.

There are some more expensive and complex solutions that may be better… but I’m not sure if the degree of “better” would be worth the hassle. Zoom makes an R16 which I also have. It’s about the cheapest recorder that can record 8 tracks at the same time. And it can be run on battery power. But it too is limited to 44.1/16 when recording 8 channels at once (though I believe it has better specs when doing 4 channels), and the build quality leaves a lot to be desired. The benefit here is that you can use real mics for all 5 channels, and not depend on the recorder’s mics. (The recorder’s mics are OK… not bad, really, and very convenient… but I like having choices). Drawbacks are size, as the R16 is too big to fit in a bag, a more complex setup with five individual mics, and you may want a mixer for more control of the mics (but you’ll need one with direct outs, which is rare in small mixers… even good ones.  The PSC Alphamix (recently discontinued, unfortunately, and hard to fund used), Sound Devices 442 (also recently discontinued, but widely available on the used market), and the current Sound Devices 552 are about the only small ones that I know for certain have direct outs. Audio Developments might. My Eela 191 doesn’t.)

Direct audio outputs on the Sound Devices 552 mixer... a key feature.

My “real” 4-track would also be a better option for a recorder… an Edirol R4Pro, which has timecode and can record up to 96k, but it cost nearly $2k… not an inexpensive solution. Their R44 is a cheaper option at about a grand and is slightly smaller, but no timecode. They’re both too big for a bag, really.

Tascam has some new recorders out. The DR680 can supposedly record eight channels at once, but it has only 4 XLRs. The specifications are a little cryptic as well, as it has

Tascam's DR-680 might be a relatively inexpensive solution for 5.1 field capture, but it's specifications are rather hard to decipher.

The Tascam 680

six mic preamps and claims “up to 96kHz” BWAV. Does that mean it can record 96K with all eight channels going at once? Perhaps…but as I’ve discovered with my Zoom, perhaps not. But the price is right, $800 from B&H. Their HSP82 can do a full 8 tracks, but it’s a full $5k. For that much, though, a Sound Devices 788T is $6-7k and would probably be a better investment.

The A-list flagship recorders are Zaxcom’s Fusion at $8K, or their Deva 5.8Deva 16. Don’t get me started on the Fostex PD606, Sonosax SX-R4, or my favorite-though-I’ve-never-even-seen-a-real-one Nagra VI. (If any of these manufacturers would like me to

The Nagra VI. The gear junkie part of me wants one, but my rational brain won't let me do it

express a more authoritative opinion, they can send me one and I’ll be glad to handle it for them;-)

So there it is… at least a partial solution to the question of simple & inexpensive 5.1 capture. It’s isn’t perfect, but it is something to try, at least.