July 19th, 2010 by Joel Hietanen

Time for just a quickie now, and as it seems the first one to become part of our slowly developing roll. Well, my steadily growing experience in making videographic work seems to lend itself to certain abuses. I, and I have to admit the true causes elude me, have become increasingly interested in the technological side of video production also. Me, a nerd, never. IDDQD.

For our ongoing ‘Pushing the Scene’ project we bought another camera to record video with, in order to get more dynamic footage from interviews with two simultaneous angles. And, as seems to be the hip and pop fashion, it was not a videocam, but an DSLR capable of HD video recording. Now, don’t get me wrong (in what follows), the luxury of shooting footage with exchangeable lenses is unparallel. However, DSLRs are certainly not ready to be (at least in terms of my limited experience) the sole recorders of video (and audio) for video projects. For now, they take wonderfully pristine footage with the correct lens choices. However, in terms of reliability and audio recording there seem to still be some unresolvable obstacles in the way. By the way, our DSLR is the Canon EOS 550D with a Sigma fisheye lens (specifically suitable for low-light shooting) and another Sigma lens for more all-around applications.


Our JVC videocam, the GY-HM100U.


The usual fieldwork equipment of a contemporary ethnographer.

First, a gripe about reliability. The 550D overheats if the shot takes over circa 20 minutes (and most interview settings obviously do). Secondly, the audio capturing capability is practically zero (with the internal mic). This naturally did not come as a surprise, as we knew we were going to use our videocam’s (JVC GY-HM100U) audio capturing capabilities for most of our recording. However, as the Finns say, the hunger increases as you munch away, and therefore I have become increasingly interested in building a presentable rig out of our Canon DSLR for video use.

So, (after days of internet scowring) what do I need? Well, a decent microphone for starters. I went through a whole set of directional test mics (Rode Videomic, Azden SGM-X / SMX-10 etc). Directional meaning to have the ability to capture audio from a certain direction, i.e. the voices of interviewees in otherwise loud surroundings. However, as I learned, this was not the only qualm. As our Canon 550D is not primarily intended for video shooting, there is one rather gargantuan concern. This would be the AGC (‘automatic gain control’) ‘feature’ that basically, for the lack of better wording and excuse my French, absolutely f***s up your recorded audio. Thankfully, Canon seems to have no interest in rectifying this problem, and there are open source micro-projects that tackle this issue for the 5D Mark II model, but not, to date, for the 550D. So yes, indeed, thanks a whole bunch Canon. The ‘methodology’ of videography work certainly lunges the researcher in a whole new realm of ‘what you need to know’, and this just after I thought to have acquired a decent baseline knowledge of editing in terms of all the various video formats and codec issues.

How to go about this, then? How to (preferably) brutalize the AGC out of our 550D. As it turns out, this in itself will not be the ultimate salvation. As I know from my dabblings with music production, I need to be able to monitor the sound entering my camera also (as in ‘not too loud, not too soft’), and of course, a DSLR not dedicated to video shooting lacks all such capabilities. Well, I found at least two options. The rather bulky Beachtech DXA-5DA and the Juicedlink DT454 4-CH. These are basically boxes of audio electronics to attach to your camera that give you the capability to monitor the recorded sound that is being fed to the DSLR.

And here’s the twist and the crux and the help-me-I’ve-had-enough part of this post. I tested two Beachtech DXA-5DA today with 3 different mics, and neither Beachtech DXA-5DA device showed any decent graphical representation of the sound going in. This would be the device to use as it provides a way to circumnavigate the whole AGC issue by fooling the cameras electronics (but that, I have to say, is another whole matter in and of itself). This, in practice, rendered the whole monitoring feature unusable without constant headphone monitoring. With our preference for unobstrusiveness and ‘going in light’ in our ethnographic video work, it would seem kind of peculiar for me to rock a camera rig the Ghostbusters would be envious of with the humble addition of having wires hanging out of my ears. Very subtle, indeed.
So, the outcome. I’ll wait for something else. And we’ll continue to use audio by relying on our JVC video-cam only. It may not take as spectacular footage (especially in the dark), but it is a full-fledged video recorder. To date, it seems, a DSLR is not. Finally, a normative recommendation. If this is your game, use both. Always. And simultaneously.