Easiest workflow to get better audio when using a DSLR

Question

I'm looking for some information (note: not a product recommendation) for getting better audio when using a DSLR. Generally, the built-in microphones on DSLRs leave a lot to be desired, so professionals will have a separate sound guy who holds a boom mic or sets up subjects with lavalier microphones or something like that. Let's assume that neither of those are options - I'm just a guy with a camera that likes shooting videos, but I don't like the way the audio sounds.

I'm trying to enumerate my possibilities in terms of getting better audio. I've only come across two so far, with some of the pros and cons I've determined, purely from research. Please let me know what you think, if I've misunderstood or overlooked something.

• Improved mic on the camera, recording into camera
  • Shotgun mic can be placed on top of the camera shoe, with a 3.5mm jack recording directly in to the video file on the camera's memory card
  • Pros: Super easy, no constant set up, possibly adjusting levels on the mic if that's an option it offers. Audio is already synced up.
  • Cons: Only one audio source (it replaces the camera's source), mic requires its own batteries (won't pull from the camera's battery)
• Improved mic on camera, recording to a separate source
  • Same idea, put a microphone on the camera's shoe, but either record to an SD card inside the microphone, or buy a separate audio recorder
  • Pros: Multiple audio sources, more control over audio with the second recorder
  • Cons: Lots of gear to manage, lots of batteries to manage, have to hit record in two places as opposed to one and then sync up the audio in post

I'm leaning towards the first option purely because it's easier. What do you think? Is there another method to getting better audio?

Answer 1

Depending on the level of quality you want to operate at, these two recording options you mentioned are pretty much it: record externally or internally.

The reason to record externally are more than you mention, I.e if shooting an interview, you can leave the sound running but turn off the camera for the moments that are not so interesting.

Also, the "cons" are not so big: syncing can be done by audio fingerprinting automatically in premiere or using an external software like pluraleyes. You can even record smpte timecode with a device like tentaclesync and synchronize audio and video automatically by timecode.

Additionally you will often encounter situations where it is desirable that the microphone is not tethered to the camera, I.e. When shooting a distant wide shot but wanting the mic to be closer.

Also, dlsr cameras internal recording quality is not great in comparison to a proper sound recorder.

While the other cons you mentioned of "dual sound" recordings are valid, I think it is something you will get used to pretty quickly. So, waging your options I think you have to decide between slightly more work but better sound and less work but worse sound.

Answer 2

Imagine using a point-and-shoot camera where you point the camera at something but instead of actually looking through the viewfinder, you only pay attention to whether or not you see an exposure warning light. You don't validate focus. You don't validate composition. You only validate that the sensor is receiving signal not blocked by the lens cap. That is kinda how people treat audio, whether using internal microphones or shotguns strapped on to their cameras. When they listen, rather than just looking at the meters, they are still mostly validating that signal is passing through the microphone.

What you can learn by listening to the microphone as carefully as you might look at framing, composition, balance, focus, perspective, etc., is that many audio environments are a mess. Many audio environments mess up sounds that can only be recognized/understood because we have a really good stereo imaging system (our ears) and a very powerful noise rejection / signal enhancement system (our brain). Randomly locating a microphone in such an environment can sufficiently compromise the signal that our ears and our brain cannot properly (or will not happily) decode it. The solution to that problem is to place the microphone where it is not overwhelmed with flutter echo, modal resonances and cancellations, etc. Usually, that means placing the microphone 3x closer to the source than to any source of confusion, which means keeping it within 6"-36" of the source source (depending on microphone type, source type, room behavior, etc). The reason that pro sound guys boom their mics is because there is absolutely no substitute for putting the microphone in the right place. The reason they use $1000 microphones is because if you are going to go to the trouble of getting good sound, you might as well get excellent sound.

There is essentially no microphone you can buy and attach to your camera that will give you good sound in a variety of circumstances. There are many mics you can buy and attach to your camera that will give you good sound in one particular circumstance: your talent is approximately 2' in front of the camera, and there's no terrible acoustically reflective surface too close to either the talent or the camera.

DSLRs often add insult to injury by also having poor quality preamps that compromise both good and poor microphone outputs. But the most important factor in audio quality is location, then an appropriate microphone type (omni, cardiod, super-cardiod, figure-8, etc) for the location, then the quality of implementation of that microphone, then the quality of implementation of the preamp and finally the digital recording format.