Jump to content
Romer

Performance Capture Audio

Recommended Posts

Hi all, first time posting so sorry if the images don't load on the first go. I've been doing audio for performance capture recently, and I'd like to share the rig that myself and the HMC specialist came up with. I am hoping to get some criticism and feedback about our method, and discuss my thoughts/observations after using this setup for a shoot.

 

As it stands, we have been using Joe's Sticky Stuff to adhere the lavalier to the helmet. This is not a permanent solution (obviously), but I have been brainstorming different options. Eventually I would like to purchase Hide-A-Mic concealers and get them permanently mounted to the frame, or fashion some form of mount with plastic that will enable efficient attaching and removal of the microphone. Are there any specific mounts that would fit this well? 

 

The studio I contract at uses SMQV transmitters. The weight of the transmitter isn't an issue for actor comfort; the helmets are counterweighted, the pack is very comfortably situated (I've worn the helmet along with the actors/stunt team from our last shoot.) Weight becomes an issue when a scene is played out with lots of fast head movements and stunts.Ideally, this packs can be substituted for smaller Lectrosonics models, but for now they work fairly well.I don't have as much experience with the SSM transmitters, but from a size standpoint they look perfect. I'd love to hear about your experiences with these units! Also, if you're curious as to why the transmitter is attached to the head in the first place, it's for actor efficiency. Many times, I am the only audio person contracted for a shoot. Often times when actors aren't shooting, they want to remove their HMC for comfort reasons. By having everything self-contained in the HMC frame, actors can simply take off the helmet and relax, rather than yell "Sound! Get this thing off me!"

 

Majority of the time, we are recording cinematics for videogames. I often run into dynamics issues when working with lavaliers in this environment. I've noticed performances are incredibly dynamic in videogame scripts... The last shoot I was on, characters would go from a soft whisper to lung-breaking shouts within half a second. I often find myself having to make the choice of gainstaging the transmitter to get clean recordings of one or the other... We try to avoid compression/limiting as much as possible from the transmitter side. 

 

Lastly, we have been running Sanken COS-11d lavaliers. Obviously an industry standard, but I'd like to hear some different perspectives. To my ears, COS-11s have always sounded bright to compensate for hiding under clothing. I've used DPA 4060s, to my experience they have a less exaggerated high end. I'd love to get your thoughts on different lavs!

 

Hope that's not too much info for one post... If anyone has questions about the entire audio rig for the studio let me know, I'm sure I have photos kicking around. Performance Capture is a strange beast for audio! 

 

 

 

 

 

Share this post


Link to post
Share on other sites

that forehead position is pretty standard for live theatre reinforcement, and if actors are shouting at all, most people go for the DPA 4061 rather than the 4060 as it has a higher max SPL rating by 10dB. I've directly compared the COS-11 and a DPA 4060/4061 in that position and your thoughts are correct re the top end boost of the sanken being problematic unless the performer is a real mumbler, in which case it might be useful.

Share this post


Link to post
Share on other sites

On the mocap shoots I've done put the lav (dpa 4063) on the side of the camera boom arm with a decent fluff on it, it just off to the side to avoid direct breath hits and it works well, especially for low level stuff. I also use zax wireless which you can adjust preamp on the fly and their neverclip system deals with unexpected level changes pretty well. That said high dynamic performance with fast changes is still tricky, especially on multi talent scenes. But that's the best thing about mocap is that it is a dream set up for the boom and that is always the primary capture mic. With a good boom op and remote tx ability you can pretty much deal with anything. I always look forward to mocap shoots because you have the consistent ability to turn in very good tracks. Lav mic on forehead when not using facial cam rigs which still sounds good but I prefer the arm rig though its a bit more fiddly to rig. We also put the Tx on a pouch on the back along side all the other boxes they wear, the removal of them is fast, we have them disconnected faster than the camera team disconnect the facial cam umbilical so all good so never an issue.

Share this post


Link to post
Share on other sites

I also agree that a DPA 4061 would probably serve you best in terms of frequency response and dynamics. I would look at the Core versions.

 

Regarding TX, I think your intuition about the SSM is right. They are far smaller and lighter than the SMQV transmitters (about the size of a Zippo lighter), but do have their own considerations. They use a 3-pin lemo connector instead of a TA5 connector, which means all your current lavs would have to be reterminated, though not an issue if you get new ones like the DPAs (you could get them terminated to Microdot, and use the adapters to go to both TA5 and 3-pin lemo). They also use a completely different battery (np50), so you have to deal with the results of that (lower battery life, battery charging, etc).

 

Outside of that, I think your approach is pretty good. I love motion capture shoots myself. As cjh said booming is amazing.

Share this post


Link to post
Share on other sites

Interesting topic!

I'd also recommend a 4060 or a 4061. The first time I heard a 4060 with a flat cap on a forehead placement changed how I mic people. I go for the forehead whenever possible, haven't come across a better sounding lav solution. To my ear, it can be as nice as a boom. The SSMs are great, I've used them a couple of times and I've really liked them. Very impressive, though they do warm up quite a bit.

Personally, I'd try to get the transmitter off the head. You could still use the SMQVs, maybe give them to talent in a waist strap that's easy to take off after they remove the helmet rig?

Share this post


Link to post
Share on other sites

Are these your main concerns?

 

1. A transmitter smaller and lighter than what you currently have

2. A simple solution for when you're the only audio engineer on set

3. Control of on-board dynamics and that is sure to never clip

 

I don't like to encourage people to choose one brand over another, but it sounds here like the exact product you are asking for is a Zaxcom ZMT.

Share this post


Link to post
Share on other sites

I will look into that Zaxcom model, thanks! Yes, those are my main concerns. Majority of the time clients will opt out of hiring a boom operator, so I am by myself working solely with lavaliers.

Share this post


Link to post
Share on other sites

With a smaller transmitter like the Zaxcom ZMT you could even rig up two lavs on the setup. That would provide safety and redundancy and you could also rig one at a lower gain level and one at higher gain to catch the full dynamic range of the performance. 

 

-Mike

Share this post


Link to post
Share on other sites

Hey all, sorry to bring back an old discussion, but I had some questions with motion capture workflow.

 

I understand the wiring of helmets and camera arms and all that, but I am little unsure of the sync process and the different cameras being used. I see timecode slates being used and all cameras timecode synced, but not sure if it's different to regular production sound jobs. And is there normally a different frame rate for motion capture?

 

If anyone can give me an insight to their audio workflow and gear used for motion capture that'd be awesome. Thanks!

Share this post


Link to post
Share on other sites

From my (limited) point of view:
(Disclaimer, although I did work on 'Avatar' and 'the Junglebook', and have sold TC related stuff to the major game developers, I never ever have been on set, and never saw the actual footage as it was Hollywood 'need to know' base, and I never left my (Dutch) office.)
Cameras are only used to check the integrity of the MoCap data. So, video was there as 'video assist', not shooting.

Now, a MoCap system is a (bunch of) computers that gather data, nothing more.
The video is used as bg, to overlay a quick and dirty rendered animation from the gathered data as soon as possible after a take, to see if the system has worked flawless, and / or a new take is needed. (Perhaps nowadays it can be done in RT, so just a bit of video delay is needed.)


If it's not RT, the Mocap system and the cams need common TC to avoid manual syncing.
On 'the Jungle book, 3 cams were around the set, and were rendered into a HD quad, I forgot what was in the fourth box...

They had an Avid on set for reviewing, and a rough first cut.

I am totally clueless what audio does on a Mocap set, other than catching directors shouting and/or scratch dialogue. I very much doubt any sound will make it into the final mix.
The set looked like a sports hall, a tremendous amount of lines on the floor representing the virtual world, so the talent (dancers / athletes / acrobats / animals, everything that can do pretty moves) would know where to walk, duck, jump etc. Nothing looked as if sound was important.

 

Problems that might arise (one of the reasons I was hired), video might go trough lots of processing and get a delay, while the MoCap system does not, so you might need to compensate for those offsets. So if you want to be prepared, bring some audio delay boxes that won't destroy a TC signal. (I would think any half decent piece of crap will do.)

 

my 2 cents,

hth,

Bouke

Share this post


Link to post
Share on other sites

Thanks Bouke & Philip.

 

The audio I will be recording is for a video game so from what they've told me my audio will definitely be used. A more low budget production on this one. So am I correct by thinking the "reference" or "witness" cameras will just accept regular LTC 23.976 timecode from a regular betso sync box jammed from my sound devices 6 series mixer?

 

So an Avid system or something similar is on set for the regular cameras? Are there common MoCap brand cameras used for these reference cameras? Or are they your normal Arri or Red cameras that we see on film production sets?

Share this post


Link to post
Share on other sites
17 hours ago, darrenaug said:

So am I correct by thinking the "reference" or "witness" cameras will just accept regular LTC 23.976 timecode

 

I know for a fact that 'some' game developers use the cheapest pieces of crap cams, so you might end up putting it onto an audio channel.
 

17 hours ago, darrenaug said:

regular LTC 23.976

 

This is a misconception. 23.976 is a bastard format, nothing 'regular', it was never intended to be used. (Of course that has changed.)
It was intended to speed up to 24 for film, to 25 for Pal, or kept the same speed but with pulldown go to NTSC, thus making this the 'Universal' recording rate if one needed to have deliverables around the world.
Games don't have to deal with this, and games like high frame rates. So it's probably going to be 29.97 or plain 30 FPS.
It might even be 60, but then syncing to LTC will be a bit more problematic, since a normal generator can't do this. (In theory it's no problem at all to generate 60 fps LTC though.)
ASK YOUR CLIENT! Not sure what generators can do 30, but I can :-) (As well as 59.94 / 60)

 

17 hours ago, darrenaug said:

So an Avid system or something similar is on set for the regular cameras? Are there common MoCap brand cameras used for these reference cameras? Or are they your normal Arri or Red cameras that we see on film production sets?

 

I have no clue what a 'MoCap' cam would be.
On a 'virtual set' (Chroma key with moving cams), a decent cam will be used.
But for MoCap, anything goes, as it's just reference, and it could be a 200 buck consumer cam.

So, again, ask your client what will be there. (And if it makes no sense, then report back.)

Bouke

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


×
×
  • Create New...