DEAR MICROPHONE MANUFACTURERS, SCIENTISTS, AUDIO ENGINEERS and CAMERA GEEKS: In no way is this test completely scientific. The mics are in close proximity of each other, but not so close to be completely scientific. The results show circumstantial results of consumer level microphone performance in a (hopefuly) entertaining and slightly enlightening way. That is all.
One of the questions I am most frequently asked by strangers is not “how are you?” but “how do you record audio?”
Ironically, when I look at the audio quality of videos I’ve made over the past two years, I find listening to some of the older samples fairly painful. However, over time I’ve gotten better, usually learning by trial and error. And those errors have included the purchasing of a lot of DSLR audio equipment.
This post is not designed to review those products or even talk about them in detail. Instead, I’ve made a video which simply stacks up 8 audio tracks in one video next to each other in what is The Coffee Pot Song Audio Test.
ABOUT THE TEST: Last week, photographers Peter Carney and James Wasserman and I rearranged my living room to film Kirk “Magic Fingers” Kenney play The Coffee Pot Song. We setup seven cameras all within one meters of each other, but all roughly the same distance away from subject. The idea was to see how some of the more common DSLR audio options recorded in a quiet room with soft and loud sounds. In post, I then switched the audio from one mic to the next so we could hear in real time what was changing in audio quality.
The floor plan of this test worked out like this:
- Canon 60D with Carl Zeiss Planar 50mm f/1.4 with:
In camera audio, control group (no external microphones)
Mounted with Zoom H1 Ultra-Portable Digital Audio Recorder
- Canon 5D Mark III with Carl Zeiss Distagon 28mm f/2.8 with:
Rode VideoMic Pro Compact Shotgun Microphone
- Canon 5D Mark II Carl Zeiss Sonnar 180mm f/2.8
Rode VideoMic – Camera Mounted Shotgun Microphone
- Canon 60D with Carl Zeiss Sonnar 135mm f/2.8 with:
Rode Stereo VideoMic Pro
- Canon 5D Mark III with Carl Zeiss, 85mm f/1.4 with:
juicedLink RM333 Riggy Micro Low-Noise Preamp
Rode NTG-2 Phantom Powered Condenser Shotgun Microphone
Sony UWP-V1 Wireless Lavalier Microphone
- Canon 5D Mark II with Canon 100mm f/2.8 Macro:
Insert shots, no audio, camera at close proximity
- Canon 5D Mark II with Canon L 24-104mm f/4 with:
Rode VideoMic – Camera Mounted Shotgun Microphone
- Additionally, independent from cameras, we recorded the song with:
Android — HTC incredible S
The top video shows all of these mics in sequence. The mics go through two rounds each followed by a final third quick round. Below is a single track of the audio which I thought was best.
AND THE WINNER IS …
FIRST PLACE: The clear winner in this experiment was the 5D Mark III paired with the juicedlink Riggy preamp, mixed with the Rode NTG-2 and the Sony Lav. However, this audio setup is a significantly bigger investment. For all three pieces you’re looking at $1240 from B&H. While I was able to control the mix of this to the lav near his color and the NTG-2 Shotgun mounted on my cage, where you really hear the difference is in the noise level. Play particular attention to the areas where there is silence, such as at the end during the credit roll. While the Riggy cuts time away from post production by eliminating the syncing process, this example does show the very affordable preamp’s ability to diminish unnecessary noise.
SECOND PLACE: Looking at the three most common Rode options the results are more difficult to examen. Out of these three options I found the less expensive Rode Shotgun Mic ($149 from B&H) to have the lowest amount of noise. However, in scenarios where audio is not coming from a single direction, I’ve encountered great problems with this mic (see good example of this mic here and problematic example here). The Rode Stereo Mic ($299 from B&H) I found does the best job focusing in on what audio is important, minimizing possibly unwanted background sounds. However, this mic captures the most noise out of the three. This mic is also the smallest of the three, making it the easiest to transport safely and not interfere with your rig as much as the Shotgun or Compact Shotgun. But in general, as a scratch audio option or interview mic option, I find this mic to be far too noisy (see bad audio example of a video created solely from this microphone here. The Rode compact shotgun ($229 from B&H) I found to be a mix between the two, offering slightly less noise than the Stereo Mic, but not as deep a sound as the larger form Shotgun.
THIRD PLACE: We also recorded with both the Zoom H1 ($88 from B&H) and Zoom H4N ($245 from B&H) audio recorders, however, in the process of recording 10 other audio sources, lost our H4N footage. The H1 footage is however, fairly remarkable for its price. I’ve had great luck with these small things and for its price — I don’t think much compares to it if you have the mic properly placed (check this link for a great example of its ability).
HONORABLE MENTION:Another big winner in this experiment was the iPhone 4. While the gain wasn’t great, compared to the Android there was no comparison. The iPhone’s footage follows the Zoom H1’s track and the difference is actually fairly small.
CONCLUSION: Like most things in life, and especially in the video world, you get what you pay for. More money means better audio. Saving a few bucks in this case, means more noise.