Your browser is out-of-date!

Update your browser to view this website correctly. Update my browser now


How to sync in a post-apocalyptic desert, with Mad Max: Fury Road’s Osmo and Machin

The production sound mixer and sound dept coordinator pushed their Ambient Lockits to the limit on the set of the action flick

Ambient Recording‘s Timo Klinge interviews Ben Osmo, Mad Max: Fury Road production sound mixer, and Oliver Machin, sound department coordinator and vehicle FX recordist (pictured, L–R), about putting Ambient Lockits to the test during filming in the desert of Namibia

Mad Max: Fury Road is one of the most successful features this year. Without saying too much about the story, it is a hell of a ride, with a lot of fantastic war cars chasing each other through a post-apocalyptic desert.

Timo Klinge: Hi, Oliver, hi, Ben. Thank you for taking the time.

Ben Osmo, Oliver Machin: No problem.

I watched Mad Max last week, and I must say I love it. I love a lot of explosions, and in the future it seems like everything tends to explode: spears… human beings when they are hit by a sandstorm…

Oliver: Yeah, the future doesn’t look very bright, does it?

How much fun was it to deal with all those special effects?

Ben: Most of it was purely visual. There were some real TNT explosions with rocks, which was a huge day for the action unit, but a lot of the action scene explosions were mainly just pops, except for the real crashes. So it was really more trying to get the dialogue while establishing and maintaining the communication, which was our mantra. At the end of the filming in Namibia, Oliver went on with a crew on his own to do dedicated car sound recordings.

For all this work the Ambient gear was invaluable, because often times cameras could’ve been 3–4km away from where we were. Particularly when I was travelling in a small van, the Lockit boxes placed on each camera were irreplaceable. Oliver set them up in the morning, and it was set and forget.

All units departed and were sometimes kilometres away from each other during the day, so there was no chance to react if something went wrong. We placed Lockit boxes on each camera so they were constantly synced from external. This way camera department could change batteries and even shut off the cameras for several hours.

As soon as they repowered the cam, it was back in sync without having them to do anything. Even if they were been closer, once they were on the camera we couldn’t get access to them. Most of the cameras were either on the edge arm vehicles, which is like a techno crane on top of a four-wheel drive with a remote head, so you put the Lockit on that camera and away they go. They’ll just go anywhere directed via radio communication. So they could get tracking shots behind the armada a few centimetres above the ground, travelling at 60–70km/h, then ramping up to 80–90km/h, and then going around to the side of the vehicle to get a close up on the driver while capturing the dialogue.

So often we were so separated that once we’d set up, we were hoping to God that the gear was still going. We were also a bit over enthusiastic about battery changes in the morning, but I can’t even remember if we ever had to change batteries on the Lockits, did we, Ollie?

Oliver: No, we were using Energizer lithium batteries and were initially a bit overly conservative with them. So I made tests and found out that Lockits work at least three full shooting days on one set of AAs. The old ACL203 Lockits were a bit thirstier, so I put them on the slates so that if the lithium Energizer batteries unexpectedly ran out during the second day of use it was not a disaster. As it was, they seemed to do two full days without issue providing I remembered to change the batteries or they didn’t get left on over night!

You really had Lockit boxes on the slates?

Ben: Yeah, we had Lockit boxes on the slates as well. I had Denecke slates as it was requested by the camera department for some strange reason [laughs]. I think, they had already made the templates, but I already owned three Denecke slates anyway, so we just had to buy one more for the production. Knowing that we won’t be able to resync them during the day, we placed a red Lockit box (ACL203) on each slate so they would constantly be in sync with the rest of the system.

Oliver: Additionally, some camera units sometimes didn’t shoot for hours, so with a Lockit box attached to it, they could simply switch off the slate for that time. Also, if they run flat on batteries, they didn’t have to come to us to re-jam it.

Ben: Of course, we had Mark Wasiutak, our boom operator, travelling with the war cars. He was our long arm to the set. So if anything was with the wireless, lavs or slate, he often was able to solve it while we were kilometres away. In fact, I often didn’t even have a picture, and if I did, it was about two seconds delayed. So we all had to rely on wireless audio. So having a setup not needing permanent monitoring was crucial.

I saw the picture of a huge amount of Lockit boxes. Can you recall how many you used in total?

Oliver: I was trying to work that out, because the picture only shows half of them.

Ben: We had four on the slates, one on each of the 11 Alexa cameras, some on the additional crash cams. So I reckon that we probably had about 18 Lockits and three ACC501 controllers each equipped with a GPS antenna [laughs].

Wow! Together with the four Sound Devices 788T and one 744T, that makes a total of 26 synced Lockit circuitries.

Ben: We set up a certain workflow which started with Oliver setting up the three controllers to sync to GMT time via GPS during our 30min pre-call. In the meantime we had breakfast. Afterwards, he passed one to me, and one to the action unit. Everybody jumped into their drive and off everybody went in all directions. I then used mine as master clock in the sound vehicle to jam-sync all recorders. Ollie synced up all the Lockits and placed them on the cameras and stayed with the main unit.

The VFX unit, who often were with the action unit, were developing a GPS positioning logging device that was mounted on each camera. For this reason the VFX dpt wanted the camera to use the UTC time information from the GPS satellites to be used as the time reference for the timecode on the pictures so that the GPS positioning data could be cross referenced with the pictures. Thus, we needed to establish this three-controller workflow, as we had the technical base camp 2–3km away from the main unit. The action unit left every morning often to different locations than main unit. Ollie stayed with his controller with the main unit while I was working with the technical unit.

At the end of the day Oliver collected all Lockits again and double checked with the controller if the timecode was still good, just in case that somebody switched it off or something else happened. This way, we would have known exactly which camera on which shooting day had what issue or drift. Luckily that never happened.

Once jam synced, all units were dead perfect.

So, Ben, you were not daisy chaining all the 788Ts, but simply jam syncing them all in the morning, right?

Ben: Yes, and they were dead perfect – never had a problem. It was just fantastic. We had a great crew of post people with us, who were constantly checking and giving us feedback. The worst they ever reported was a quarter of a frame difference between our up to 26 individual running units after a full shooting day, which still was incredible. They found out afterwards that it was not because of us but because of some cameras working off speed. Anyway, we double checked the tuning and never heard another word from them about sync, and they were happy as Larry [laughs] – even though I don’t know who Larry is…

Maybe Larry Fisher, who recently retired [laughs].So, with the nasty environment and weather conditions, did you ever face any issues with the system?

Ben: No, never. They never skipped a beat; they were fantastic. Of course, we were pretty meticulous and gave them a dust off and cleaning every evening. The only thing we had to exchange afterwards were the leather pouches, as the plastic look through front part cracked thanks to the dry desert heat.
To give you an example of how demanding the environment is, I can tell you that Mercedes send their cars there for rust stress tests.

Did you do any particular preparation on sync before you started the production?

Oliver: Yes. The benefit of using an Ambient Lockit system is that you can recalibrate or tune all devices to one speed. As we used Lockits from different generations and Sound Devices recorders, it was a must for what we wanted to achieve in accuracy to tune them before we started. This was even more important as all of these circuitries were tuned in different temperature zones before, so it was pretty realistic that the different units would have drifted against each other.

To avoid that, I tuned the ACC501 controller to the atomic clock of the GPS satellites. Now I used this as the time reference, and tuned all other Lockit circuitries on set to the same time reference. I know it sounds complicated, but the truth is it’s an easy five-minute job to tune all 28 units and it can prevent hours of work if something goes wrong.

So you tuned all units only once before the job?

Oliver: As far as I can remember I tuned them again once we were half way through, but that was more to see how far they drifted in the meantime. I know that once tuned, they normally don’t need to be retuned for years.

And did they drift?

Oliver: No – actually no unit needed to be recalibrated.

Did you get any feedback from post after you finished on Mad Max?

Ben: I know from the editor, Margie [Sixel], that with sometimes 12 cameras for one shot, they ended up with 250 hours of footage. So it would’ve been a major problem if only some of the material would not have been in sync. But, as said before, they were all as happy as Larry with what we delivered.

Mad Max: Fury Road, written, produced and directed by George Miller, opened on 14 May. As of 23 June 2015, the film has grossed $144.7 million in North America and $202.7 million outside of North America for a worldwide total of $347.4 million (against a budget of $150 million).