“Camcorders”

Back in high school, I decided to get my own video camera. It was my final year of Media, and the Media cameras.. I just didn’t like them. They looked too small. (I had no logical reason at the time, as I had very little camera knowledge). So I shopped around online, looked at all sorts of fancy “camcorders” from Sony, Panasonic, Canon, JVC… There were miniDV cameras, HDD cameras, direct to DVD cameras.. sorting out what I wanted was a great learning experience. I ended up getting a Panasonic NV-GS500 for cheap on eBay.

It was shiny, relatively large, and I’d read it had a big sensor for shooting in low light conditions. Great reasons for buying a camera, I know. But anyway. While shopping around for my camera, I became curious to know.. what was the most expensive video camera out there? You could browse the cameras in certain price brackets, so I went and looked at the $14000+ section, and thought “shit, what an intense camera, and for $14K that must be what professionals use, wow”. I think there was only one camera in the 14,000+ section — a Canon XL-H1, or something like that.

I was incredibly impressed by the number of dedicated buttons, the lens hood that made the lens look huge and intimidating, and the matte black finish.
But now when I think about it, I hate these sorts of cameras. They’re not “point and shoot” cameras by any means, but there’s this horrible emphasis on automated, computerised functions to make things easier.. and the marketing of these cameras can be extremely deceptive. They look like theyr’e on the bleeding edge of camera technology.. space-age submachine gun things.. yet most of this advanced technology simply helps the camera cut corners and interpolate data. If this sounds too general.. I guess I’ll name names. You’ve got the Canon XL-whatevers, and the Sony HVR-Z-whatevers, Panasonic HMC-whatevers… I don’t like them. They occupy this awkward space between consumer camcorders and fully professional cameras, and everyone seems to think that they’ll have “superior resolution” because they look so futuristic and confusing.
What’s not to like? I’ll tell you a few of the things I hate about said cameras.

Sensor size.
They’re all pretty much 1/3″ CCD or CMOS cameras, meaning that the crop factor is huge. You need wide, wide lenses to squeeze a picture onto such a tiny set of sensors. Hence it’s difficult to get a shallow depth of field, hence your video looks even more like ‘video’. What’s worse is that these cameras are usually HD. Squeezing so many pixels onto a small sensor increases noise

Lenses that stop down as you zoom.
Apparently, this has something to do with making zoom lenses more compact. I’m not sure about Canon’s or Panasonic’s lenses, but Sony’s HVR-Z1/Z5 lenses stop down rapidly. The Z5 goes from F1.6 at wide angle to F3.4 at full telephoto… God help you if you’re doing a night shoot with it (I was last week). I guess if you’re going to use the camera with auto-shutter and auto-gain, the massive light loss won’t be as much of an issue, if you’re prepared for your tight shots to be grainy as hell. But for regular, sane people who aren’t going to resort to auto-gain, getting a tight shot could mean completely changing your lighting setup.. massive pain.

Interframe compression
HDV. It’s.. sort’ve HD.. but not. It uses MPEG-2 long-GOP compression, with sprinklings of keyframes and strings of interpolated pictures in between. Now.. I’m currently editing a project in HDV, and it actually looks quite good, but it takes a lot of processor power to work with HDV footage, and your colour correction options are quite limited because you start seeing compression artefacts everywhere. It looks good, providing you don’t mess with it in post, or confuse it by shooting lots of movement or complex detail. HDV chroma subsampling is also terrible.. 4:2:0.. don’t shoot anything red. Red shirts, people under red lights, etc.. it all blows out, looks flat, blocky and extremely saturated.

Auto-anything.
Alright. So it’s not necessarily bad for a camera to have automated exposure software. But they make it so freaking hard to turn off. You switch the camera to “manual”, but that’s not good enough in many cases.. that only toggles autofocus, that sort of thing. To make shutter speed, aperture, and gain fully manual, there’s some other trick you’ve got to pull off.. it’s quite deceptive.

Fake zoom/focus rings.
By “fake”, I mean digital. You can turn the focus ring round and around and around forever, I guess because the ring isn’t mechanically attached to the lens elements.. it controls them indirectly through digital circuits instead. something like that. The point is that these focus rings aren’t designed for focus pulling.. at least, not the Z5. Maybe I was completely tripping, but on the Z5, the sensitivity of the focus ring seemed to change as you moved from wide to telephoto. This meant that accurately marking focus points on the lens never worked.. you crash zoom, focus, mark the focus point, then zoom out and your focus point will be somewhere completely different. wtf.

that’s probably enough complaining for now. my point is.. don’t be fooled by the shininess or the buttons or the price tag.. they’re usable cameras, but there are many many corners cut.

Shocking revelations

I’m ingesting the footage we got today, at home in iMovie, because it’s so much faster than setting in/out points. And I discover something terrible. The freaking Z1p, despite being on manual, was auto-adjusting the gain settings. Gain is, without a doubt, the most dangerous thing to mess with on a small camera like the Z1. And our overlay footage is grainy as hell. Jesus. It’s incredibly annoying, since pushing a single button on the side of the Z1 would have saved us, and I only discovered it half-way through our shoot. If I’d found it earlier and fixed the gain at 0dB, we’d simply need to boost the intensity of our dedos.. so simple. And so much less grain. As it is, the gain’s probably up to +9dB and I’ve underexposed to compensate for this sneaky boosting.. fuck.

They tell you not to boost the gain, they do warn you about this. On the side of the camera is a hard gain switch.. generally, it’s set at 0, and so you think you’re fine. But no.. oh no. In a typical example of camera software trying to be too smart for its own good, the Z1 still boosts gain, irrespective of your “manual” setting, and your hard gain switch being set at “0”. What they don’t tell you is that in order to override this obligatory gain boost, you have to hit another small button on the side of the camera, labelled “gain”. Once you do this, you’ve set gain to manual, and actually have control of it. You’ll see “0dB” or “6dB” or whatever on your LCD display. If this gain value is absent, it means that the Z1 is manipulating gain as you go, and will make the brightness of your shot fluctuate depending on light levels. Bloody hell. Everyone in the world should make note of this. It should be plastered on billboards all over the city, broadcast on all major television stations on a weekly basis. “Remember that if you’re shooting on a Z1, the gain will boost itself and destroy your footage unless you hit the “gain” button and turn it to manual.” It’s just so counterintuitive, because when I have the camera in “manual” mode, I expect it to be MANUAL. I want NO HELP from the camera software and its automated calculations.

Yes. I’m pissed. There’s no one to blame, and I’m sure I knew that you could toggle the “manual-ness” of the Z1 using an array of small, nondescript buttons on the side. I’ve found this out before, quite a while ago. But it’s so amateurish, such consumer-camcorder design, to have auto-gain as a reasonably default feature. I just didn’t expect that from the Z1. Really should have gone with the DSR on this overlay shoot… the shallow depth of field would have been great, and the gain switch at “0” actually means “0 gain”. Brilliant. Pure genius, no?

“Get over it Josh. Your footage is grainy. Your average viewer doesn’t care about grain, and will easily perceive the imagery you’re trying to convey without noticing anything detrimental.”
Yea, but it’s a lot of grain, you have no idea how much grain.

So what can we do about this pseudo-disaster? (I’m the DOP, I’m allowed to call excess grain a “disaster”)
We can try overtly “stylising” the image in a way that disguises the grain.. by using a soft focus filter in FCP, or darkening each image to make it more ‘atmospheric’ and lessening grain… Or we could desaturate the images, since I don’t mind black&white grain as much.. it’s not as ugly as rainbow colour grain.. ugh. There are workarounds, definitely. And luckily, only half of our footage is grained.. I figured out how to turn it off once we started filming the projection sequences, which are the more visually important parts of the film, I’d argue.

Still.. it’s just such a kick in the face to find that something so simple could derail half our day of shooting. I hate camera automation. seriously.

The next wave

So apparently, a group of Media 3rd years shot their final film using a Canon 500D (or something similar) — a still camera (that can record video, obviously).

I’ve been hearing a lot about these new hybrid cameras recently. Lots of high praise from cinematographers, photography teachers, various websites.. but I remained dubious, thinking that the full HD footage would be compressed to hell with some awfully lossy codec. Or that it would look fine, until you tried to edit it, and found it loaded with sneaky artefacts, or immensely slow to work with because of all the interframe compression… etc.
But then last week, Paul mentioned this 3rd year group that actually decided to use a still camera for their final film. And it worked.. they’re editing it, it looks great.. etc. Perhaps my compression-phobia was misplaced. It’s h.264 compression, which is a complex codec.. i’m quite sure it’s got inter-frame GOPs (the exception is AVC-Intra, Panasonic’s new h.264 codec which only records I-frames) and I don’t know how easy it is to work with in post. But, as with HDV, you can convert the footage to Apple ProRes and the lag will be gone (along with all your hard drive space.. still. ProRes is awesome)

Anyway, my point is.. what does this trend towards still cameras shooting video mean? They shoot full HD, (progressive, I’d guess) and you get to use lenses designed for 35mm sensors. That’s brilliant. The depth of field’s going to be much shallower than it is on a Z1 or something with a tiny sensor. In the case of the Canon 500D, you’ve got an APS-C sized sensor.. not quite the full 35mm, but pretty damn big by camcorder standards. And you could use a more expensive still camera with a 35mm sensor for even better results in this respect.

I’m thinking that the downside of all this is physically operating a still camera. It’s so comparatively small, susceptible to being knocked and shaken. That’s a big problem if you’re wanting to pull focus during a take, or zoom or whatever. You’ve got to do it smoothly, without shifting the frame. I’m not sure how viable it is, but in my head it seems quite difficult. Then you’ve got problems with the viewfinder. Shooting in broad daylight, using the glossy LCD screen on a still camera would be difficult. And if you’re doing an extreme low angle shot, there’s no way to tilt the viewfinder.. you’ve got to be lying under the tripod to see what’s going on (obviously not possible most of the time). Of course, you could bring a field monitor and hook that in, but then you’re operating by proxy.. trying to pan the camera while looking somewhere else, etc.

The upside is, of course, the 35mm lenses and the full HD capture. Apart from RED, I don’t know of any digital video cameras with full frame sensors. Even high end HD models tend to have a 2/3″ sensor..
And so rather than hiring a RED for $850 per day from Lemac, plus $650 per day for a good set of 35mm PL mount cine lenses, you can buy a Canon 500D outright for $1500 or something (And then you make friends with photographer and they let you use their awesome L-series Canon lenses). In some ways, that’s quite compelling.

Of course, the other option is to get a 35mm lens adaptor, and plug it into a Z1 or something similar. Then your camera handles like a proper video camera, and gets the shallow depth of field of full 35mm cameras, and captures in HD, and has a better viewfinder. Lens adaptors also need more light for correct exposure and can cost thousands of dollars, depending on the make. hmm. I really should be talking about my TV2 shoot coming up this monday. Tomorrow. half-promise.

Trouble Shoot

Had an equipment test shoot today for our doco. I think it’s absolutely fantastic that we’re organised enough to have test shoots and all the rest of it.. getting as much foresight as possible, and using all our time efficiently. It’s quite different to TV1 where we were lost in a quagmire of pre-production problems and other pitfalls/hurdles. We’re learning fast now, getting it all together.

Today’s test shoot was invaluable. Can’t stress it enough. I’m going to discuss the results (the look of our footage) as soon as I rip some still images and compress them. But for now, just some practical bits and pieces:

Setting up a Kino
I thought this would be reasonably simple.. and for the most part it was. There’s no separate ballast box, when I always thought there would be.. to regulate the phase of the alternating current. (think the ballast is physically attached to the light itself.. probably). But anyway, two problems:
1. the light must be attached to the C-stand, but it’s attached via a metal ball/socket mechanism which you need to tighten.. or else the light will flop everywhere, loose in the socket. It’s difficult to explain, but to tighten the socket you’ve got to pull out a lever, which acts like a wrench (you have to tighten, then pull the lever up and move the lever back to the starting point, then tighten again, etc). I don’t think that that explanation was helpful at all.. but I know how to do it, and that’s the main thing. (scary to think that without this test shoot, our first interview setup could’ve been wrecked just because of a Kino socket thing).
2. The kino does not seem to slot easily into the C-stand nib thing. This actually took a while to work out.. I was fairly puzzled, thinking that we were missing a piece of the Kino. But Tim solved it in the end — just had to untighten the screw on the Kino to an absolutely ridiculous point, and then it will fit. God.. simple. But again, scary that it could have tripped us (or at least me) up.

Setting up a projector
You need a “dongle” to connect it to a computer. DVI to VGA, in our case. And it gets worse. New macbook pros use a different sort of dongle to older macbook pros and macbooks. Unless you’ve got your own, borrowing from techs or Paul becomes difficult because you’ve got to know how old your macbook pro is, and which dongle it needs.

Turning on the zoom ring
I freaking hate it when people disable the zoom ring and go for the “W – T” button on the Z1. wtf? Especially when you’re setting up the camera in the dark, and can’t remember where the damned toggle is to rectify the problem. For future reference, it’s on the side of the lens.. a big switch.. can’t miss it. (I missed it because it was too obvious.. hate it when that happens)

That’s enough for now. I can sleep peacefully tonight, knowing that at least 4 things won’t go wrong on our actual shoot because they went wrong today. And yea. I’m off to rip still images from our test shoot footage.

Disclaimer: Z1

The previous post is not an example of me hating on the media course, or on its equipment or facilities. I am not grumbling and saying that I think RMIT should shoot everything on RED.

I don’t mind working with the Z1p at all. You can get some great images out of a Z1, and documentary’s all about story anyway. All I was saying, in the previous post, was that there are certain things to be mindful of when using a Z1. I like a challenge, I like working with a camera that has “issues”. That’s all fine. I just feel I’ve got to know the problems, and know them well.

Off the top of my head, the Z1 has 3 1/3″ CCDs, the lens has a focal distance of 4.5mm – 54mm, which would be the equivalent of 32mm – 390mm on a full-frame camera because of the sensor-crop, and the iris can open to F1.6 at 4.5mm, F2.8 at 54mm. It can record in DV, DVCAM, and HDV (sucks because it uses inter-frame compression, which is very difficult for a computer to figure out while editing footage).

And yes. I’m not bitching. Just understanding. I sound like a tech-obsessed robot now, but it means that during production I’ll be fluent, ready to concentrate on aesthetic, narrative-related problems rather than technical ones.

Limitations: Z1

It was only week four, and already we were shooting and editing. Quick progress, I was pleased. I tried to simulate the camera setup that I’d wanted to use for interviewing Gitta, to see if all would work.

My resounding feeling is that while the shoot went fine, the Z1p is in some ways a limitation for our doco, something to be careful of. Yes it’s compact, yes it has a sharp lens and millions of options in the menu.. but being compact and versatile comes at a price.

Disadvantage 1: Sensor size.
The Z1p has 3 1/3″ sensors — I won’t bore you with excess detail, but that’s half the size of the DSR’s sensors (1/1.5″).
There are two implications:
1. Noise. The more pixels you pack into the one space, the noisier an image becomes. Camera manufacturers combat this in numerous ways, but ultimately, boosting gain on a Z1 will most likely produce a lot more grain than the same boost on a DSR.
2. Focal lengths. The smaller your sensors, the more the image is cropped from what your lens provides. Because the CCDs are so miniscule in a Z1, you need extremely wide angle lenses to channel light down into such a small sensor area. Wide angle lenses tend to show everything in focus, it’s very difficult to get a shallow depth of field (selective focus). Now, the lens on a Z1 goes from 4.5mm – 54mm. Anyone who’s used a 35mm sensor DSLR knows that 4.5mm is a ridiculously wide angle lens, and 54mm is pretty much a standard lens. Because of the difference in sensor size, however, 4.5mm is not ridiculously wide on a Z1, but it is still ridiculously in focus. This means that to get an aesthetically pleasing, shallow depth of field, you’re going to need to zoom in.. and quite a lot. 54mm is adequate in terms of selective focus but it’s also the equivalent of 390mm on a 35mm sensor. Jesus. So you’ll get decent focus, but you’ll only get a shot of the subject’s eyeball… unless you move the camera back, a long, long way. And there’s another catch, which leads us to the next problem…
Disadvantage 2: The Lens
Yes, its’ a Zeiss lens. Yes, it’s sharp enough for HD imaging and we’re only using it for SD filming. But we’ve got a problem. At 4.5mm, your aperture can open up to F1.6. Nice. Quite a fast lens. But of course, at 54mm (when fully zoomed in), the lens can only open up to F2.8. Now.. this is still reasonably fast, I’ve seen worse. But the change in light sensitivity is still dramatic. If you start at wide angle, with a wide aperture, watch how much darker it becomes while zooming in.

From this

To this

The DSR’s lens did this as well, but the difference didn’t seem as pronounced, and the larger sensors meant you could get away with closing down the iris a bit to start with, to match the minimum aperture on telephoto.
Yes, this seems fussy. Most zoom lenses have to close down somewhat to reach maximum focal length. But with sensors that are half the size of DSR sensors, you’ve got to zoom in to get a selective focus, which means your camera will not be as light sensitive, the image will shake all over the place at the slightest touch, and you’ll need to be a long way from your subject in order to fit them into frame at the equivalent of 390mm. Crazy.

Disadvantage 3: Menu layout
There’s nothing worse than delving through digital menus and pressing tiny, squished-together menu buttons on the most awkward parts of a small camera body. And that pretty much sums up the Z1p’s menu system. There’s no substitute for dedicated, well-spaced, actual buttons.

Summation
Ultimately, the Z1 is a great camera for edgy documentaries where you’re walking down the street hunting your subject (because at 4.5mm so much will be in focus), and when it gets dark you turn up the gain and the noise looks “gritty” and “raw”… etc. but if you’re setting it up on a tripod in a controlled environment, the Z1’s versatility is meaningless.

Of course, one could say something like “Josh, you’re just obsessed with obtaining a shallow depth of field and if only you’d give up your obsession, the Z1p would be a great camera to use on wide-angle with a wide aperture and low gain.” And yea, I’d probably concur.

Just a tech guy

I’ve noticed something about this course, a tension of sorts.

We’re dealing with ‘tech stuff’. We’re forced to blog and compress video and embed things and manipulate cameras and operate sound mixers. But we’re introduced to these things very carefully, very slowly. Think back to WMT, EMT… back in the day. “We write media texts here. We don’t edit them, God no.” —If we let you write and edit at the same time, you’d lose all conception of reality and make amateurish, trippy but crap videos where you played with every single effect contained in the iMovie arsenal.

No… technology would corrupt your minds, make you excited and giddy. They give you technology when you go to TAFE courses, you see. And this is a university course, you see.

Technology is irrelevant, for we are interested in the workings of your mind, the decisions you make. Executing these decisions with equipment.. that’s another matter entirely.

— Now, don’t get me wrong. I agree with the above statement. I understand the point they were trying to make by making us go through WMT and EMT. But I don’t like this duality that develops, this idea that the great artistic masters of film and television hate technology, while mindless tech-heads spout jargon and program things until the work comes together.

I’ve worked with Final Cut. I’ve used the sound mixer before, I’ve used the Z1P before, I’ve used radio mics and the boom. But this does not mean that I’m just “a tech guy”. I didn’t choose this course so that I could operate equipment. I want to make something… something aesthetically pleasing, or meaningful, or emotionally powerful (better yet, all three).

And yet every time I open my mouth in tutes (which is too often), I feel transmogrified into this cold, ineffectual nerd who does not understand the film process holistically… someone who looks at the nuts and bolts, but never grasps the power that film can have. Obviously no one came out and told me that I was like this, but it’s just the ‘vibe’ I get. You’re not meant to know about tech things. You’re meant to be afraid of them, because you’re an artist and a nice person… a creative master who’s above the nitty gritty problems with equipment. Otherwise, you’re probably an unpleasant computer programmer in disguise.

Yes, I’m pretty tech-savvy. But I want to be a filmmaker, not a scientist. What we’re doing, what we’re making, is both a science and an art.. and both aspects of the process interest me. Surely that’s alright. Surely you can do both?