Audio level issues while making firearms videos

[matt8090]

I'm making firearm review videos for my YouTube channel. I've been using a lavalier mic plugged directly into my camera and that works quite well. However, I'd rather not be tethered to the camera and tripod at times. My camera is a Canon Vixia HF R400. I recently got a Rode Videomic Pro to use. I absolutely love the sound quality in most circumstances but in scenes when I'm shooting a gun the sound level will drop waaaay down. It won't come back up to normal levels until after the shooting stops and I've been talking again for about 4-5 seconds. I've got the power on the middle slot, not on the high pass filter, and bottom switch on -10 dB. I hope I've explained the problem well enough. I've included a link to a small sample video that shows what's happening. Any help would be greatly appreciated guys!

https://vimeo.com/150072442

[Brainiac]

I do not have this camera but did a Google search for keyword AGC (Automatic Gain Control) because your problem sounded like it might be related to that. According to an Amazon customer review, the purchaser (who said he confirmed with Canon) states that the AGC cannot be disabled.

Brainiac

[Cornucopia]

Couple of things you should do (beside the obvious of NOT using AGC):

1. Make sure you're using a mic that has a pad (audio equivalent of neutral density filter) feature. Then use that feature. The sound level generated from gunshots is 1000s of times greater than what mos mics are natively able to handle. Just like you wear headphones to protect your ears in the firing range, you want to physically or electronically (in analog realm) protect the mic capsule/pre-amp stage and A->D stage.
2. Best practice is to bank/stagger the input. IOW, split the mono mic signal into two separate channels where one channel's input level (sensitivity) is set much lower (-10 to -25dB) than the other (all other things remaining identical). This allows optimal dynamic range to be covered BOTH during the loud and quiet parts.
3. Record at 24bit, 96kHz, LPCM and keep it at least that good throughout the editing process, until you are ready to do your final compression for distribution.
4. If you're able to, it makes sense to utilize those banked channel segments separately during the corresponding moments (dimmed when loud and undimmed when quiet) and transition between them smoothly. A sort of manual gain control, rather than automatic.
5. Even better would be to combine the 2 banked channels into a single "super"-channel of 32, 48bit integer format or possibly 32-64bit floatingpoint format. Then use tailored dynamic range compression to squeeze your range down from the HUGE range of the superchannel to what one would expect to be a normally perceived range. Much like bracketting exposures to create an HDR image and then using perceptual (not global) TMOs to derive a standard LDR image.
6. If the above is workable, close-mic your gun subject. This will greatly reduce surrounding noise.
7. Perform practice recordings to get the levels right, and mark down the settings so you know how to adjust/combine during editing.
8. If AGC cannot be disabled, use double system audio with a separate dedicated recorder (like a Zoom H4N).
9. For better isolation from noise & better hum rejection, use a better condenser mike with better pickup pattern (supercard, shotgun) and use balanced low-z connections.

Scott