Something that intrigues me about DSLR footage...
 

Most of the video I've seen that's been shot with these cameras is of the "pictures + added music" variety - which is usually lovely but gets a bit samey. Is this because these cameras don't really 'do' audio well or is it because people have come from a stills background rather than a video background and therefore don't know about audio mixing and post-pro? Just wondering as it's something I've noticed looking at most of the D90 / 5D II footage out there.

Hope this doesn't offend anyone, just that coming as I do from an audio background it's got me intrigued. Are we still in a phase of people trying to pursue the best images and 'testing' out the cameras (and I realise that's hugely important of course), or is this perhaps a new style of video we're starting to see, driven by these particular cameras?

Please check out Tramm Hudson's work on Magic Lantern firmware. You can disable the Audio Gain Control in the 5D2 for better audio recording with an external preamp.

I've had a look at the magic lantern and it does look very useful. But it doesn't really answer my point - and I guess if more people from the stills world migrate into video (and more video is always good, in my book) we may see more of this particular style of 'beautiful still images which happen to have some movement in them' with a music soundtrack.

I think the 5D Mk II images are really seductive and I love the control of depth-of-field which these cameras give you but I'm fascinated that this seems to make people shoot in a particular style which is something that seems to have emerged with these cameras - a sort of static, reflective quality.

Hi Mark,
I think it's not really to do with the camera because a LOT of stuff you see on Vimeo for instance is the same. If people have projects they want to complete that involve sound, they will use whatever they need, whether it me a DSLR or a video or even film cam.

So if I have a project that involves using sound, I do the sound. If I'm just out and about doing nature stuff I just ignore sound and fit it to music like most other people.

I expect it depends on the person and their background. They may fancy doing a narrative piece but not have connections with any writers or actors. They may fancy doing a documentary but not have the right contacts to get into it.

I certainly wouldn't read too much into the prevalent DSLR output being because of lack of pro audio support.

Just my two penneth.

Cheers

Avey

For myself... I think audio is much harder to pull off on any video footage... every time I
try to use site audio there is always some noisy stuff going on in the background of the
audio... so site audio is not high on my list and I just put the footage to pre recorded
music... just easier that way...

Mark,

Though many do come from a stills background, it's not all the reason why we don't get "good" sound from these vidslrs. Another reason is the camera itself. It isn't that great in capturing sound. Some don't even have mic input. And those that have it, are not xlrs, and their amplifiers are suspect. And this is not mentioning the fact that those with mic input, you can't control gain or have meters to see the settings. You don't even have an audio out with them!

Add that to the 3rd reason that the subject/topic of the project may be live recording, that adds to the issues why most are dubbed with music. Some are outdoors and you do need a 2nd hand or 3rd hand to help you out on sound.

Of course, there is also the fact that if indeed you are serious with sound, even with a non-vidslr, you'd still use another way to record sound. That means, even if you have a RED camera, you'd probably record your sound separately anyway.

Regardless, add to that a better mic system. The cost piles up, not to mention the lenses. Sure they are cheaper than other solutions, but then again, sound will be the last in the priority to upgrade. Mind you, the 5d mk2, lenses, adapter are already expensive, without a good supercardiod mic, a sound recorder, an xlr-converter, etc.

As for the style of shoot, well, with most vidslrs, there isn't really much to do in terms of the way you shoot when you have to fiddle with controls trying to focus and get in frame and other things in your list since most vidslrs are severely crippled in other areas w/c a regular video camera has. As an event shooter, for example, I can't even imagine the work and jumping hoops one has to go through shooting events with vidslrs. If I were shooting other things, they may work because I have the time to set things up and adjust. For events, it's just a pain. :(

In time things will improve. When the GH1 gets produced in numbers, we'll get to know more about what can be done when the camera isn't so much crippled. Of course, the GH1 is not perfect either. Still, it does many things better and right if only as a video camera.

Jon Fairhurst has done alot of tests using external preamps like the Beachtek and Juicedlink via ML and 5D2. The result out of the Juicedlink and ML combo is amazing (to me at least). Hardly any hiss. Can't wait for a fully functional ML.

Thanks all - these are really good insights. And of course it's slightly tinged with envy in that I don't have a 5D MkII and the ability to produce those amazing and beautifully detailed, shallow DOF images. However I can imagine that for event video, DSLRs probably aren't the right beast, as they stand anyway. I was judging the audio from some of the videos posted on the Magic Lantern wiki, and wasn't so impressed, but I've just looked at Jon's comprehensive audio tests on Vimeo and they're very impressive, especially with the Juicedlink / ML combo (as Alex points out above). Of course as suggested, you'd probably want to use another device for audio capture if you were doing anything complex - or more people, or any combination of above!

One thing that does intrigue me is seeing some wonderful DSLR images of people, usually looking pretty un-self-conscious. Which makes me think that the shape of those cameras doesn't provoke the reaction you sometimes get with video cameras, as people may not realise they're being recorded.

Mark, you raise an interesting point that I've also wondered about for the last year or so. It may be that the initial infatuation at being able to shoot video in a new way just left audio and narrative on the sidelines for the time being. Simply finding a compelling short story with real people speaking or good synced nat sound is tough, and it complicates the edit exponentially. The big difference with these cameras is in the pictures, and putting a series of images together with minimal regard to the soundtrack is satisfying, at least for the time being. At least that is my theory of the "testing out phase".
On the other hand this may become a new style of video (that may get old very quickly). Music and voice over stills looked especially effective displayed on the internet a few years ago - better than a picture in the newspaper, more lasting than an image on television. Every still photographer became a potential "multimedia" producer, and now they are all adding video to that mix. Near instantaneous delivery doesn't help either - getting good audio then editing good audio takes time, and time is always in short supply.
I'd be willing to bet Canon never imagined the 5D MkII taking off as a preferred video camera the way it has. In this early generation getting good sync sound (and forget about timecode) is tough work - more art than science. Maybe sound will come back when the camera catches up to our expectations.
For years I've been fortunate to work with a soundman who is a good friend and great asset on any job - another set of eyes and hands as well as ears. Working by myself my prejudice to the visual art shows, that is where I invest my effort even though I know from years of experience people watch television with their ears - silent movies were good but without the advent of sound I can't imagine them taking over the world.
One byproduct of this will be the elevation to royalty for the last remaining soundmen and women. Fifteen years ago they were identified as a dying breed but those that are good are still around and have all the work they care to take on.

From my point of view, because I fall in to this category you're describing, I'm currently concentrating on imagery to get a sense of feeling of how I want to shoot something more serious that will be dialog and capturing audio. So right now I'm out shooting and editing and then maybe throwing a song on top to get a sense of time and space. It's an easy cheap way out while testing and prepping for a more serious project, for me at least. So I think you're seeing what you tend to see in school when it comes to video work. A lot of music videos to keep a mood and impact easily without having to think about audio at the moment while concentrating on getting a skill-set for imagery. Plus you're correct in assuming all us photographers out there converting in to "filmmakers" might not really think about sound (not me personally since I also come from music and amateur sound recording background).

As far as audio, I have the H4N and two Sennheiser G2 wireless mics and I couldn't be more happy with the sound this is creating. I use the H4N mics as backup and ambient recording, my 5D mic for track sync and the G2's for capturing fine audio details. It's a great system for what I need. I have no problem importing in to Premiere pro unlike some people having sync issues in FCP.

Are we returning to the world of silent film?

Imagine watching Reverie in black and white while playing Scott Joplin piano music. All it needs are the black and white title slides for setup and dialog. Add some sprocket jitter and we're there!

Now imagine Reverie with top quality live sound from each scene. We'd hear city sounds, a car, a helicopter and a bunch of random noises. Frankly, you'd get a better result mixing sounds from a sound effects library, or by recording very specific sounds separate from the live filming.

*** The only thing that needs to be captured live is dialog *** (And even that gets replaced in many cases.)

Back in December, the first thing we shot with the camera was narrative:
The Last Outpost on Vimeo

The only live sound (on camera mic!) was the zombie scream. I recorded and processed all of the other foley, sound effects, and music myself using studio gear. This is clearly different than the "video vignettes with music" genre, but we still didn't rely on live sound. (We fully expected to ADR the scream, but didn't need to.)

The one place where live sound is needed beyond dialog is news/documentary. Dan Chung's work is a great example.

I think the main reason that we see so many vignettes is that they're easy. You simply shoot what's available and use the best shots. As soon as you write a story, you need actors, specific locations/sets, costumes, props, etc.

If the camera came with great sound, the vignettes would be more likely to include natural sounds like the ocean, dripping water, traffic, jet takeoffs, etc. But often, natural sounds suck - there's HVAC, refrigeration, planes, autos, dogs, birds, people talking and bumbling off camera. Natural sounds often don't blend across cuts. It's still easier to just add music.

Anyway, with a juicedLink and Magic Lantern firmware, we can now get great, sync'd sound. Filmmakers need to spend the money for the preamp, cable, microphone, and mount/stand/boom - and they have to be willing to load 3rd party firmware in the camera.

It will be interesting to see if we get more stuff with live sound. My guess is that we will mostly get more of the same...

And I shot a narrative over two months ago, with the camera. Then, we didn't have sound solutions that are the promise of Magic Lantern. See it here:

Dead of Winter- Choices Made on Vimeo

But remember, this first 9 months of the this camera- and the learning curve about what can and can't be done is just now being overcome.

In a couple of days I will be posting another narrative that was shot with earlier version ol Magic Lantern. Still fought sound a bit still, because of a choice of mics, but the manual controls of 1.10 and Magic Lantern are clearly the answers we have been waiting for.

Personally I know that for testing the camera out it's much simpler just to grab it and walk out somewhere to shoot without actors, crew, etc - so for most people that's likely to be the first video they put together with this camera. That's what I did for the first couple months - I just didn't bother posting my results because everyone was posting stuff like that.

I've done two sync-sound projects so far - one true documentary and one 'mockumentary':

Sur Mesure: Journeys in Transpersonal Haberdashery on Vimeo

a place like this... on Vimeo

It's taken a little while but I've got a sound workflow that I'm really happy with that is just about as easy as on-camera audio. I'm using an H4n to record and Pluraleyes to sync in FCP - finally found out how to eliminate any sync issues and I really like working this way, enough so that I haven't even tried out the magic lantern firmware with the juicedLink yet. After successfully completing the above projects I finally started using the 5D in place of my XHA1 for paid interview work as well. I just shot 10 short interviews with this combo for a client and everything has worked smoothly so far - I'm starting to think the XHA1 may be permanently retired. I have two more short narrative projects planned for August which will get away from the interview-heavy stuff I've been doing so far and if nothing comes up there the 5D will probably become my only video camera.

Mark I agree with both of your posts.

About people getting filmed and not looking in camera it must be because most US citizens want to be future Hollywood stars, do they stop at tape on the floor spots?? ;-DD

Really, here in Italy/Spain I can't think you can film people from 2 meters and not have them look at you with their worse face...

I don't think that any of the short pieces that my wife Alice & I have been doing would categorise us as Philip Bloom/Vincent Laforet wannabees as we generally avoid the still photo & cool tableaux with jazzy music style. However apart from one interview we generally avoid live sound for all the reasons that John Fairhurst details & not just because of the problems of recording sound with this camera. Even for that interview the intention was to use off camera sound but the Zoom H2 recorder died at the crucial moment (the on/off switch snapped off) & we had to revert to plan B which explains the poor sound even after lots of tweaking in post-production.

Stephen Clarke: Dial M for Merde on Vimeo
Tour de France 2009 on Vimeo
Llorca on the rocks 2009 on Vimeo
Old Nice on Vimeo