Off camera audio solutions for 5D shooting

[Jon Fairhurst]

I just packaged up the H4n and sent it back to Southern California today during lunch. It's a great little unit!

I haven't had time to listen closely to the results, but the H4n sounded very good. At first blush, I'd say the juicedLink/5D2-ML win when time is precious and you don't want to spend time syncing. The H4n wins when you want to hide it on the set on in the wardrobe, or want your sound person to free-roam without a cable to the camera.

The Microtrack II and BeachTek DXA-5D/5D2 have significantly more noise, but are still usable for smaller screen applications if your signal is hot enough.

All will be revealed after I edit and upload the video...

[Christopher Lovenguth]

I am using the H4N with two Sennheiser G2's in the XLR and it works great! I keep the H4N on or near my camera recording a stereo track from it's on-board mics, record audio on the 5D and then record two mono tracks from each G2. I start my H4N and then I slate my audio by snapping fingers when I start my 5D, which the on-board H4N's and the 5D pick up as a spike but usually the G2's don't pick up and since all 4 tracks of the H4N are in sync when you drop them in your time line, I just match up the 5D footage to the audio tack.

For me I'm having no drift issues but I'm using Premiere pro and hardly ever go over 6-8 mins with a clip before I'm cutting. I'm really thinking the drift issue where you have to change the audio by 99% is a FCP thing. I record on the H4N in 48Hz 24bit just to get everything out of it as possible.

[Jon Goodman]

I'm using a SD 442 mixer with a 744T recorder. Sync with a slate. NO better audio solution. Of course you need an extra pair of hands to run it.
Cheers
Jon

[Jon Fairhurst]

Originally Posted by Jon Goodman

I'm using a SD 442 mixer with a 744T recorder. Sync with a slate. NO better audio solution. Of course you need an extra pair of hands to run it.

And the budget to afford it. :)

[Evan Donn]

Originally Posted by Christopher Lovenguth

For me I'm having no drift issues but I'm using Premiere pro and hardly ever go over 6-8 mins with a clip before I'm cutting. I'm really thinking the drift issue where you have to change the audio by 99% is a FCP thing.

I suspect you're right - here's why: I'm using Pluraleyes to sync multiple video clips to single long audio takes from the H4n but I haven't been retiming the audio first. You would think each clip would start out in sync and then drift over it's duration, but instead each successive clip is starting out slightly more out of sync - and they all fall perfectly in sync once you retime the audio. This suggests to me that when pluraleyes analyzes the audio it's doing so at the correct timing, but final cut is playing it back at the wrong rate...

Ok, just checked something and confirmed this. I have a 16/48 wav from the H4n from a recent interview. Quicktime reports it's duration as 28:27.60. In the FCP bin it's reported with a duration of 28:25:21 (non-drop frame). On a 1080p30 48/16 timeline it has a duration of 28:25:26. Retiming to 99.9% puts that up to 28:27:17, almost exactly the correct duration. Placing it in a 29.97 timeline results in a correct duration of 28:27;18. So I'm not sure exactly what's going on but FCP seems to be misinterpreting the time base of the audio clips on import, and I don't see any preference which can change this behavior.

[Evan Donn]

Just saw this post on the Pluraleyes blog from last month - apparently if you create a proper easy setup in FCP with a 30p timebase it eliminates the need to retime H4n audio in order to sync it with the 5D:

25 Hour Day: DSLR Dual-System Audio: The 99.9% Solution

Sounds like it's a bug in FCP - just changing the sequence settings to 30p isn't enough to prevent it retiming the audio as I described in my last post.

I just set this up and am running a job with the pluraleyes beta on approximately 1.5 hours worth of material - 32 video clips and 6 audio, I'll report back on the results when it's done.