Blog

Microphone Research, Testing & Conclusions

Introduction

The All That Glitters project has a video component in addition to the still photos that I’m capturing for each burlesque performer. The camera I’m using, the Canon 5D Mark II, records high-definition digital video, which is fantastic. However, the audio capabilities of the 5D are greatly lacking. The camera features a built-in microphone that picks up handling noise and any movement of the camera’s internal mechanisms. I needed to find a better method for recording audio along with the video. Fortunately, the 5D has a microphone input.

I spent a few days researching and reading reviews for a variety of microphones. These are my notes and conclusions based on this research.

As I mentioned, the 5D has a microphone input, albeit through a 1/8″ connection. This narrowed my microphone options for capturing audio with the camera to those with 1/8″ connections. I could use an XLR-to-1/8″ adapter, but I was curious to see what my options were with this limit in place.

Long ago I made an assumption that the best audio for video was recorded with a shotgun microphone. But I learned early on in my research that shotgun microphones were best used outdoors, not indoors where I anticipate most of the interviews for this project will occur. The common theme for microphones for indoor video & audio recording was hypercardioid.

On the other hand, I wanted to try to find an all-purpose microphone that I could use with my camera beyond just this project. So, I took the limitations of indoor shotgun recording into account while still including shotgun microphones in my research.

Recording Devices

Before beginning the microphone research I took into consideration the options I have for recording audio. These include:

• MacBook Pro (1/8″ input)
  
• Canon 5D Mark II (1/8″ input)
  
• Sony MiniDisc (1/8″ input)
  
• Tascam DA-P1 DAT Recorder (XLR input)
  

My requirements for a recording device are:

1. Ease of portability
   
2. Ease of use (as few clicks or steps as possible to begin recording)
   
3. High quality, low noise, digital recording
   
4. Manual control of audio input levels
   

I briefly considered the MacBook Pro, but eliminated it early as it fails #2. Sure, Macs are easy to use, but the amount of time and steps required to get set up, boot and configure the software exceeds my comfort levels when I have someone sitting in front of me waiting to be interviewed. I need to be able to hit record and go.

I was also immediately suspicious about whether or not the 5D would be a viable recording option, even with a good microphone. I already own a Sony ECM-MS907 so I plugged that into the 5D for a quick test.

Initial testing revealed that there is no manual control over the input levels into the camera. Input levels are regulated by the camera and it sounds like there is a compressor of some sort in place that controls the audio input. As a result, the noise floor (hiss) in the audio recording is greatly amplified when there is silence between speaking. I want to be in control of my audio input levels. But since I hadn’t tested the 5D with a good microphone I needed to follow through to make a sound conclusion (no pun intended).

The MiniDisc is definitely portable, easy to use, records digitally (but lacks digital output) and will give me all of the recording control I need.

The DAT recorder is also portable, easy to use, and provides top-quality audio recording through XLR inputs. Additionally, it has digital output for ease of transfer to audio editing software.

I devised the hypothesis that, with a decent microphone, the 5D would become mid-quality level audio recording device providing for reference and backup audio and that I would likely need to rely on a second microphone into the DAT recorder for my top quality audio. I would simply mix the two in post-production.

Armed with the information about what I could record to, I launched into microphone research.

Microphone Research

Since I am already a fan and user of Rode microphones, I took a look at their latest offerings and discovered the Rode VideoMic, specifically built for camcorders. I also learned about the Rode NT3 microphone, a hypercardioid condenser microphone with a lot of mentions and supporters in on-line forums.

NPR’s This American Life (TAL) is a significant inspiration to me, and given the quality of their recordings, I figured they’d have some suggestions of microphones to consider. They use the Audio-Technica 835B, which isn’t as widely available as some newer models. They also direct users to Transom.org’s Tools page, which has some helpful information. Of special note is their articled titled What Microphone Do I Get?

Based on the information from TAL about Audio Technica microphones, combined with some articles and forums I read, I added the Audio Technica AT897 to the list.

YouTube was a helpful resource for finding reviews and tests demonstrating the sound quality of each of the mics.

The most valuable reference while making my final decision was based on the video samples provided by dvestore.com. They provide samples of the AT897 as well as the Rode NT3 and VideoMic.

After much deliberation I settled on an order for the Rode NT3 and VideoMic. The VideoMic would provide audio input to the 5D, and the NT3 would connect to the DAT recorder.

Microphone Testing

My first test of the microphones and their respective results were as follows:

1. Sony ECM-MS907 connected to Sony MiniDisc – Decent quality, but not at the level I wanted. Lots of room noise, not surprising considering this isn’t a very directional mic.
   
2. Rode VideoMic connected to Canon 5D Mark II – Pretty good, much better than the Sony, but as I mentioned above, the 5D messes up the consistency in sound with the built in compression. Functional, but not ideal.
   
3. Rode NT3 connected to Tascam DAT recorder – Outstanding. I immediately liked the results of this mic.
   

I was disappointed with the limitation of the 5D audio recording capabilities, but still wanted to give the VideoMic a fair trial, so I switched the Sony mic with the VideoMic on the MiniDisc.

1. Rode VideoMic connected to Sony MiniDisc – Solid quality, and definitely better than the Sony. There was a bit of room noise, but it was tolerable and potentially controllable with some post-production.
   
2. Sony ECM-MS907 connected to Canon 5D Mark II – Same result as when I first tested the Sony to the 5D…not as bad as the internal mic, but still pretty bad.
   

I won’t be loading any samples of the audio or video I recorded, not only because there is a plethora of reference material already available on-line, but mainly because I was in my PJs when I recorded it.

Conclusions

Ideally, the audio and video would be recorded by the same device. However, the current limitations of the audio recording on the Canon 5D Mark II, combined with my budget constraints that prevent me from acquiring a dedicated device, require some creative alternatives.

Thus, I have launched the project by recording video using the 5D, audio into the 5D using the Rode VideoMic, and audio into the DAT recording using the Rode NT3. My initial attempts to combine the audio and video in post have proven successful. I am very pleased with the Rode NT3, and I expect that the Rode VideoMic will provide for reliable general-purpose audio recording on the 5D.

An Open Letter

Dear Canon,

Please modify your firmware to include the functionality for users to disable audio compression and manually control the audio input volumes when recording video on the Canon 5D Mark II. Sure, the manual control over the video settings is fantastic, but now let’s get the audio functionality caught up.

Love,

bcj.