I considered this option for myself last year and borrowed a DSLR / housing to try it. I chose to stick with a video camera, at least for now. Here are my reasons, mainly about the housings than the camera:
- Stability and Panning: The shape and balance of a good video housing seems to be more conducive to smoother handling, especially when shooting a pan or trucking shot. DLSR housings being shorter front-to-back tend to pitch up and down more. I'm sure that you could learn to overcome this but video housing are definitely more stable.
- Controls: On the DSLR housings (and cameras) that I've seen, the buttons are well placed for photography but the critical video buttons are not as well placed. start-stop, af-lock, zoom. Also the monitor window, being vertical and lower down, makes the handling more awkward for comfortable video shooting.
- Depth of Field: You could argue that the shallower DoF on DSLRs makes for more artistic footage with differentiated focus. But that also mean that you'll need to pay a lot more attention to focus. When that shot of a lifetime swims past, your chance of messing up is greater. Not a problem for an photo but more complicated for camera movements, eg trucking into a macro subject.
-Mindset: With a DSLR, there is a risk that you'll adopt a mindset of "I'll shoot some clips and then take some photos". Very tempting to exploit the versatility of both options while diving. Isn't this a good thing? Maybe / maybe not. Flipping between these modes could possibly take you out of the "story telling" mindset needed to get the footage to weave together a story back home in the editing suite. Your videos may become moving photos rather than a story. This might not be an issue for you but certainly something that I've heard others say.
For me, I'm watching this space and waiting to see if any of the housing manufacturers make a DSLR housing that's great for shooting video...
The red filter replaces lost red, the first colour to be lost in the colour spectrum. After 70 ft other colours are also lost and you would only be replacing red with your filter but not the other lost colours...
Don't want to be picky but one of the common misconceptions about the red filter is that it "replaces lost red". It can not. Red light that has been absorbed by the water is gone. Can not be replaced by a filter (only a artificial light can do that - and only over very short distances). The red filter looks red because it does not allow blue and green to pass through it.
Why is this important?
Because, if you don't understand what is really happening with the light and colours, you'll make wrong choices.
The red filter is reducing blue and green.
If there is enough red left in the ambient light, it will reduce blue and green, shifting the balance between the three primary RGB colours, creating the appearance of more red. If there is none or little red left, it wont help at all.
But, by reducing blue and green, it is reducing the total amount of light that the sensor is receiving. So, if the sensor is struggling to get enough light to get a decent image, you are just making it worse. Different cameras react differently. Some try to boost the light by amplifying it (gain). This can cause ugly fuzzies.
So a red filter (should really be called a "blue and green reduction filter") is only useful if there is a lot of ambient light and some red left in the mix of ambient light.
The depth range in which it is actually useful varies a lot from dive to dive and even during the dive as the sun disappears behind clouds. As a rule of thumb, you'll get some improvement between 5 to 15 meters with a red filter. If it very sunny, don't swing it in until 8 meters deep, otherwise some things will look too red. If it is cloudy start using it from 5 meters deep. Definitely have it swung out by 15 to 18 meters deep as you are just losing light to the sensor. There is no R left to "balance" with the GB.
If you turn on video lights, flip out the red filter. if you do a silhouette by pointing the camera towards the sun, flip out the red filter.
Better than a red filter, if you have a camera that can do white balance in low light, I suggest that you forget about the red filter all together and just set a suitable white balance for the depth that you are at. But that's a whole new story...
Not sure what you are trying to do but I doubt if you want to get the WB up as high as 15,000k.
(all the below assumes ambient light only. if you add artificial light, the story changes)
Remember what the camera's white balance algorithm is trying to do. It is trying to make the subject white. But even if the subject is bright white (e.g. a slate) on the surface, it is never going to be white at depth. All the red and most of the yellow is gone. So it is naturally going to be blue-green. Trying to push it to white is impossible (without video lights).
Underwater, what we are trying to do is shoot nice video that is acceptable to the viewer. This usually means getting rid of any green, having the water a nice blue and, if possible, trying to get whatever red or yellow that is available in the subject to be emphasised.
Can you do presets on that camera?
If so, forget manual white balance while underwater. Set some "close enough" presets for different situations and just switch when necessary. Doesn't need to be too accurate. If you get it "close enough" you can do final tweaks in post.
I use the Canon XF100 underwater. One great XF100 feature is 4 preset WB settings. I just step through these until I get to the one closest to the situation.
1 = Daylight (close to the surface or when video lights dominate - night, cave, wreck) 2 = 6500 kelvin (5 meters to 15 meters - just warms up a little - no red filter required) 3 = 8000 - 9000 kelvin and set to reduce a little blue - for deep blue water 4 = 8000 - 9000 kelvin and set to reduce green - for deep water when there is a lot of green
Presets 1 and 2 are just set in the menu. Presets 3 and 4 are done by printing an A4 page from Photoshop, filled with either pale aqua blue or pale aqua green. I put the page in sunlight and shoot a manual white balance off the page. This gives me about 8,000 to 9,000K. I move the paper, relative to the sunlight, and camera distance until i get a K setting in the 8000 - 9000 range. I save them into the two MWB slots.
When diving, I just use the WB button on the housing to step through the 4 settings as appropriate to the situation. I don't touch the MWB button as it'll throw out one of my carefully prepared presets.
These days I rarely do anything other than the slightest tweaks in post.
Which video light you would choose, to make great videos and good photos ?
First, make sure whichever light you buy has enough coverage for the widest lens you plan to use.
For photography, you need the most powerful light you can find. Even that won't come close to the intense blast of a strobe but, if you can get a really bright video light, it'll work for some close-up and close-range photographs. But, remember the big difference is timing. A strobe fires for a tiny fraction of a second (e.g. 1/10,000), freezing all movement. Using a video light, the light will be on for the entire shutter opening. If you are using say 1/60th of a second, a lot of movement (i.e. blurring) can occur in that time. Good for nudibranchs; Hopeless for tuna...
Posted by peterbkk
on 03 September 2012 - 04:00 AM
For Peter: I had a look on the BS Kinetiks web site, but I could not find xf300 on the camera list and neither xf100. How did You do to have the housing? Are the controls fully remote or do you know if it is possible a mechanical controls for manual focus? This housing should be fine for light weight also because xf300 is kg 2,7. How is the footage with the xf100 in low light and bad light and water conditions? Are there issues, noise and aberration? I have read you have two uw housing: an electronic one and a mechanical one. Which is the brand of the second one?
BS Kinetics (Bodo) can engineer a housing for many different types of video camera. I did not communicate directly with Bodo. David Cheung of ScubaCam helped me put this system together.
BS Kinetics has several housing models. For the XF-300, you would need to use the housing that is slightly larger than the camera. He already has the electronics control profile for the camera because it is similar to the XF-100.
These housings are completely electronic. All operations are done by pressing buttons. The buttons are magnetic, through the housing skin. This is good because then the housing only has two o-rings, both big, fat, round o-rings; one for the port and one for the main opening. This means that the risk of leak is much less.
There are control buttons for every camera function. The most common are direct, one-purpose buttons near the hand grip: start-stop, focus, zoom. The less common are below the monitor and work in a "mode" approach. e.g, mode 1 = WB, mode 2 = iris, etc. Very easy to use.
They do have "manual" focus but it is operated through 3 buttons. One button switches between manual and auto focus. Once in manual mode, another button moves the focus to "near" and one button moves to "far". With a bit of practice, manual focus is possible. However, I find it more convenient to use the "lock auto focus" option.
The XF100 is OK in low light but not great. It does have small CMOS sensor so it has small pixels. To keep noise low, I use manual gain control and keep it zero or very low. I also turn down the iris to manual in dark environments and keep the recorded image dark. I would expect that the XF-300, with it's 3 sensors, would be better in low light. The 4.2.2 color encoding of all the XF cameras is great.