ChristianG
14-12-2009, 23:10
Last updated 01/01/2010
INTRODUCTION
This is an attempt to give a "how to" for the purchase of a camera specifically suited to underwater photography. It is largely designed for users of P&S style cameras although the same basic principles apply to all digital cameras.
If you are buying new, the first choice you are likely to have to make is whether to buy a P&S, a water resistant P&S, a Bridge, a Hybrid, or a dSLR camera and that’s a lot of categories of camera. Confused yet? I hope to explain this, including questions such as "What on earth is “Manual White Balance”? in what follows.
Where I have referred to a website for a general explanation of something I have specifically used Wikipedia, if only for the neutrality of it. There are likely to be better explanations around but they may be a little biased. Similarly, if I give a list of, say, lens manufacturers it is in alpha order.
Also bear in mind the weight (think airlines) of the camera/housing that you are willing to lug about with you as well as how much space you want to devote to this kit.
Do you think you will want to get more technical in future? If you dive deep you will need to keep a close eye on the depth rating of your housing and ancillary equipment (if any). Depth ratings are fairly arbitrary, all housing makers test their housings significantly deeper than the recommended maximum depth. All 10Bar housings, for example, are tested to 10 bar or 90 metres of seawater (hence the name of course) yet the recommended maximum depth for these housings is 60 metres, fully one/third less.
This is not, by any means, a suggestion box of cameras you should look at. It is not in my view possible to suggest a camera, or cameras, given the plethora of them out there, their various price points, their various differences, the individual preferences of prospective purchasers and the ever-evolving camera scene generally.
Similarly any explanations given are on a "rough and ready" basis, so, unless they're downright wrong, if they do the trick please let it be.
Perhaps I should also point out that when I talk of angles of view (how wide the lens is) I have converted all the figures to their 35mm film equivalents for the sake of simplicity.
Bias, we all have it to one extent or another. Where I believe that I might be biased I will say so.
ABBREVIATIONS, EXPLANATIONS ETC
• P&S: Literally "Point & Shoot". Small cameras with permanently fitted lenses. Increasingly fewer of them do not have a viewfinder, which I wouldn’t look through anyway because of parallax error (http://en.wikipedia.org/wiki/Parallax), you view what you want to photograph on the screen on the back of the camera. Note that those screens are fairly tiny (despite the marketing-speak of the manufacturers) and they become tinier still behind a housing in turgid water.
• Water Resistant P&S: these don’t have a separate housing and are fairly new to the game so more variety is being made available all the time. Check the depth rating on these very carefully, you are unlikely to be able to add on any ancillary equipment but they are handy for snorkelling, for shallow dives where there is good light and for leaving around on a sandy beach.
• Bridge: marketing-speak for a more sophisticated P&S. They often, but not always, look like a small dSLR. There are cameras I would designate as “Bridge” cameras but which may not be considered so even by the manufacturer and so I tend to ignore this nomenclature and head straight for the specs.
• Hybrids: quite possibly the coming thing. Briefly put, a P&S but with an exchangeable (not underwater) lens. Not quite a dSLR (you can’t view the image through the lens) but with better options (albeit limited by (dome) port availability on housings of which there aren’t any as I speak - housings that is). The Olympus Pen E-P1 is the first such, virtually immediately superseded by the Pen E-P2. The E-P1/2 series is built according to the Micro version of the 4/3rds Standard (http://www.four-thirds.org/en/index.html) which allows for (a) a plethora of lenses, both Olympic and others and (b) a sensor significantly larger than that of P&Ss. That would be fine for those manufacturers who subscribe to that system but care must be taken to ensure that similar future products from other manufacturers have decently sized sensors.
• dSLR: "digital Single Lens Reflex" camera. These have the ability to change lenses (not underwater) and they have a viewfinder which looks "through the lens", in other words "What You See Is What You Get". Most do not use the screen to show you a ‘live’ view of what you are about to take a picture of, you look through the viewfinder and see the picture on screen after you’ve taken it. As a general rule the expense of the dSLR camera reflects its quality/the size of its sensor. Some do have what they call “live view” where they show what is the potential picture on their screens. There is a growing tendency, particularly from erstwhile P&S users, to want to use the screen to shoot the picture. Don't do it! The viewfinder is a far better option.
• Pixel: "Picture Element". It’s important that you have at least a rudimentary understanding of what happens here, so please bear with me: CCD (http://en.wikipedia.org/wiki/Charge-coupled_device) or CMOS (http://en.wikipedia.org/wiki/CMOS_sensor) (see also CMOS sensor (http://en.wikipedia.org/wiki/Cmos)) sensors make up the vast majority of such devices and are the imaging element of the camera. They (with the exception of the Foveon sensor, a CMOS design but arranged very differently but which has virtually no relevance here) are for our intents and purposes interchangeable. All P&S sensors are either 7.18mm x 5.32mm or 5.76mm x 4.29mm in size, which is really tiny. If you can, always try to buy a camera with the bigger sensor, bigger is better in this case. The other trouble with pixels is that, shamefully aided and abetted by all the manufacturers (to be fair, some were drawn into this by the other, more shameful, ones) the general public continues to think that more pixels is also better, not so in many instances (http://www.yorkshire-divers.com/forums/underwater-video-photography/64746-pixels.html). Particularly in the instance of P&S cameras less is often more.
• Pixel layout: another bone of contention. With the notable exception of the Foveon sensor all the others are laid out in a flat plane of cells, often called sensels (sensor cells), on the sensor. A certain Mr Bayer, then working for Kodak, came up with the concept now employed by all such devices of an algorithm that complements the cells (which become pixels) of the image. Bayer established a combination of R (red), G (green), B (blue) and G (green, again) sensels to make up one actual, real, proper pixel. Note that this is very similar to your TV except that there is no extra G in the TV picture. These sensels are set up in a rectilinear grid on the sensor. Confusingly all the camera manufacturers describe each sensel as a pixel (no doubt marketing-speak driven) so if your proposed camera has 8,000 "manufacturer” pixels you can divide by four to get to the proper pixel count of 2,000 which happens to be plenty in the case of a P&S given its tiny sensor.
• Noise: what in the good old days of film used to be called “grain”, a much better description because it actually means what it says but never mind. Noise means undesirable elements in your picture although these can, sometimes, perhaps partially only, be corrected in post-production. Noise is often a trade off between aperture, exposure and ISO.
• ISO: actually there is no such thing as ISO in digital photography but the manufacturers have decided that this is a convenient term to carry over from film photography for purpose of the exercise. In fact there is now a standard on this, but we won’t go there. The higher the “ISO” you dial into your camera, and some of them, particularly the P&Ss, are very restricted in this, the more “sensitive” the lens becomes - at the expense of highly undesirable noise.
• Exposure: the amount of time (that the lens is open) it takes for your camera to collect the data of the picture it is trying to resolve. Think camera shake and the need for a tripod if using long exposures.
• Aperture: the “hole” that the camera sees through to take the picture. The smaller the hole the greater the Depth of Field (DoF) (usually desirable), meaning that more of the image is in focus. Conversely, the larger the “hole”, the lesser the DoF.
• Lenses: IMO the beating heart of any camera, usually described by "x" mm at "f" something-or-another. The lower the numbers the wider and brighter (technically “faster”) the lens respectively. Wider and faster is the name of the game underwater. To put this into perspective a 24mm lens captures approximately a 136% larger viewing area than that of a 28mm lens and 213% larger viewing area than that of a “standard” 35mm lens whereas an f2 lens is about twice as fast as an f2.8 lens. 24 mm and f2 is the current benchmark for P&S cameras (and probably stood the rest of the manufacturers on their collective ear when it first saw the light of day). Having said that, the widest lens I have ATM for my dSLR is 10-20mm at f4-5.6, depending on where the telephoto of the lens sits. Very roughly speaking, if you use the zoom (we'll get to that) you progressively darken the image. As well, there are lenses and then there are lenses, Samsung uses Schneider lenses and Panasonic uses Leica lenses and I tend to stand at attention whenever I hear those names. This is not to say that the Canons, Nikons, Olympuses etc of this world are "bad" lenses but I certainly wouldn't entertain a camera and lens made by the equivalents of Messrs Yum and Cha (that no-one's probably ever heard of previously).
I should point out that in the Canon and Nikon lens ranges (for example) there are “professional” lenses, in Canon’s case usually designated by “L” and by Nikon as “Nikkor”. Some other manufacturers do the same thing and some use bespoke manufacturers of the likes of Schneider and Leica (you’ll find Karl Zeiss lenses sitting in Hasselblads). “Non” professional lenses can generally be taken to be of lesser quality, sometimes decidedly so. For example if I were to buy a dSLR I would almost certainly not buy any of the lenses, often called kit lenses, which may be available for that camera at little to no extra cost. This is hardly likely to be the case other than for the “entry level” cameras of the various major manufacturers, nor is it likely to be the case in the lenses of any of the alternative lens manufacturers such as Olympus, Sigma, Tamron, Tokina etc. The likes of these, incidentally, make some pretty good lenses which are right up there even though not all of them may be of that quality.
• WB: "White Balance" (or MWB – “Manual White Balance”), also useful above water, allows you to change the hue your camera shoots at depending on the light. Light, BTW, is not just bright or dull but it’s also not a subject for purpose of this exercise. A form of white balance is employed for example when you opt from "scenery" to "portrait" (etc) mode on a P&S.
Something to be considered if you are using (or wanting to use) WB is that it needs to be changed at different depths and in different conditions. Remember that different levels of colour get lost at different depths with red being the first to go? Therefore the WB setting at 5m will definitely not be the same as that at 40m - actually the WB setting at 40m is going to be well nigh useless anyway without an artificial light source! If WB is going to be important to you, choose a housing/camera combination on which you can alter it fairly easily. Try it out in the shop! If you want to choose a camera that has a video option, you may wish to think about whether the WB can also be altered on video.
• raw (http://en.wikipedia.org/wiki/Raw_image_format): some P&Ss can shoot in raw (which is a file format and the suffixes change, depending on the whim of the individual camera manufacturer e.g. Nikon have .nef files as their raw files whereas Sigma uses .X3F). It basically means that all of the information is made available to you rather than having the camera auto-convert it into a jpeg (usually) when a lot of the information is unfortunately lost. Importantly, raw allows you to adjust the WB in post processing and that is just so useful u/w because that turtle you're trying to get a shot of is not going to stick around while you fiddle with WB. It follows that if you have raw, and do post processing (as you should - see below) then you don't really need WB.
• jpeg (http://en.wikipedia.org/wiki/JPEG): stands for the Joint Photographic Experts Group and it was a brilliant concept when it first came out, still is for that matter and today pretty well everyone knows what a jpeg is and everyone can read it whether they are camera buffs or not. The trouble is that it is "lossy", look at the Wikipedia explanation of that, particularly the flower at the right of the screen. jpegs are perfectly adequate, indeed preferable if only because of their intrinsic failing, for sharing images over the Internet.
• Sharing images (a quick word): there are methods of doing this but the best, simplest advice I can give you to ensure that your images are not "acquired" by someone else is to convert them to jpeg at not more than 72 ppi (pixels, real pixels, per inch) at a small size, say 50 x 35mm which is perfectly adequate for Internet purposes, any higher resolution than that can't be seen on screen. At that size and that resolution your images are not really of value other than on the Internet itself - by which I mean that they can't be successfully printed other than at a much smaller size than that because printing an image requires 300 ppi - do the maths. Oh, OK: 50mm at 72 ppi = 72/300 x 50 = 12mm, that's less than postage stamp size, I assure you that most of the postage stamp manufacturers would make very sure of the identity and veracity of their image.
• Focus: is really self explanatory. With the advent of P&Ss excellent autofocus pretty well became mandatory and has come ahead in leaps and bounds. IMO anyone who even attempts to manually focus a digital consumer camera via that silly little picture on the back of it, especially u/w, will only successfully do so by gosh or by golly.
• Zooms: the majority of P&Ss have zoom lenses, some of which are absurdly long. The Olympus SP-590UZ "Bridge" Camera has a 26x lens of which you might be able to use about 1% u/w. My longest lens on my dSLR is 3x or 50-150 mm and that's a "proper sized" lens. Furthermore you can't use more than, say, 3x u/w (and that's stretching it) because unlike air you're not only drawing in the picture, you're also drawing in all the particulate matter suspended in the water in front of the image - commonly called crud (some people are less polite about it) but more correctly called Backscatter (http://en.wikipedia.org/wiki/Backscatter_(e-mail)). Apropos Backscatter, this is a perfectly good reason why a puny pop up flash immediately next to a lens pales in comparison to decent strobe arms (http://forum.divernet.com/album.php?albumid=2&pictureid=43) where the backscatter (mostly) reflects back to the strobes rather than the lens.
• Memory Cards and Batteries: I would assume you know about these but it is far more annoying underwater if you run out of card space or the battery runs down before you finish the dive so consider this as part of your purchase. I consider an empty card and a full battery a requirement rather than an option per dive. I have two batteries for the camera and several cards.
UNIFORM RESOURCE LOCATIONS
Some URLs to help you:
http://www.digideep.com/english - mostly tells you whether there is a housing available for a camera. Mostly - they're not prescient.
http://db.tidbits.com/article/7891 - tells you, in exquisite detail, how to buy a camera but it has no provisos for u/w use so that's very much a caveat. I would urge you to read Charles's other musings (http://db.tidbits.com/author/Charles%20Maurer) on that site as well, the man is a bit good.
http://www.dpreview.com/ - gives you as close to an unbiased opinion as you can get, provided you know how to read the technical jargon, which is pretty daunting for a newcomer. Again, above water only.
http://www.cambridgeincolour.com/ - one of the better "how to/what's that?" sites.
http://www.digitaldiver.net/ - a useful “comparison” site for bits you may, or may not, want to buy.
http://www.wetpixel.com/ - as the name implies - arguably the best underwater digital photography site on the web.
http://www.digigreen.net/forums/ - as the name implies - a UK site specialising in UK u/w digital photography.
HOUSINGS
The most important things to remember are that:
• not all cameras automatically have housings, it follows that before you buy the camera, check to see about the availability of a housing and whether that housing is going to suit you,
• some housings are, sometimes significantly so, more equal than others,
• housings can, and will, drown. That's a fundamental law of underwater photography so either live with it, check your insurance cover or forget all about it and
• housings require O Rings which do need to be looked after rather lovingly (http://www.digigreen.net/forums/other-stuff/7056-o-rings-general-gear-care.html#post35938).
As a generality, all manufacturer supplied housings are made by someone else, mostly Sea & Sea as it happens, which is actually no bad thing because at least Sea & Sea have a pretty decent track record - unlike the majority of camera manufacturers who certainly have a clue about cameras but zero clue about housings. These housings, however, are often fairly limited because the specs are those of the camera manufacturer where cheapest is often best. I do not here mean that they are necessarily built cheaply but I do mean that they can be limiting in other ways, particularly the addition of wet lenses and strobes (see below) or other loss of functionality. I point out also that the Canons, Nikons of this world are not interested in housings for their dSLRs entirely leaving that up to the housing manufacturers.
INTRODUCTION
This is an attempt to give a "how to" for the purchase of a camera specifically suited to underwater photography. It is largely designed for users of P&S style cameras although the same basic principles apply to all digital cameras.
If you are buying new, the first choice you are likely to have to make is whether to buy a P&S, a water resistant P&S, a Bridge, a Hybrid, or a dSLR camera and that’s a lot of categories of camera. Confused yet? I hope to explain this, including questions such as "What on earth is “Manual White Balance”? in what follows.
Where I have referred to a website for a general explanation of something I have specifically used Wikipedia, if only for the neutrality of it. There are likely to be better explanations around but they may be a little biased. Similarly, if I give a list of, say, lens manufacturers it is in alpha order.
Also bear in mind the weight (think airlines) of the camera/housing that you are willing to lug about with you as well as how much space you want to devote to this kit.
Do you think you will want to get more technical in future? If you dive deep you will need to keep a close eye on the depth rating of your housing and ancillary equipment (if any). Depth ratings are fairly arbitrary, all housing makers test their housings significantly deeper than the recommended maximum depth. All 10Bar housings, for example, are tested to 10 bar or 90 metres of seawater (hence the name of course) yet the recommended maximum depth for these housings is 60 metres, fully one/third less.
This is not, by any means, a suggestion box of cameras you should look at. It is not in my view possible to suggest a camera, or cameras, given the plethora of them out there, their various price points, their various differences, the individual preferences of prospective purchasers and the ever-evolving camera scene generally.
Similarly any explanations given are on a "rough and ready" basis, so, unless they're downright wrong, if they do the trick please let it be.
Perhaps I should also point out that when I talk of angles of view (how wide the lens is) I have converted all the figures to their 35mm film equivalents for the sake of simplicity.
Bias, we all have it to one extent or another. Where I believe that I might be biased I will say so.
ABBREVIATIONS, EXPLANATIONS ETC
• P&S: Literally "Point & Shoot". Small cameras with permanently fitted lenses. Increasingly fewer of them do not have a viewfinder, which I wouldn’t look through anyway because of parallax error (http://en.wikipedia.org/wiki/Parallax), you view what you want to photograph on the screen on the back of the camera. Note that those screens are fairly tiny (despite the marketing-speak of the manufacturers) and they become tinier still behind a housing in turgid water.
• Water Resistant P&S: these don’t have a separate housing and are fairly new to the game so more variety is being made available all the time. Check the depth rating on these very carefully, you are unlikely to be able to add on any ancillary equipment but they are handy for snorkelling, for shallow dives where there is good light and for leaving around on a sandy beach.
• Bridge: marketing-speak for a more sophisticated P&S. They often, but not always, look like a small dSLR. There are cameras I would designate as “Bridge” cameras but which may not be considered so even by the manufacturer and so I tend to ignore this nomenclature and head straight for the specs.
• Hybrids: quite possibly the coming thing. Briefly put, a P&S but with an exchangeable (not underwater) lens. Not quite a dSLR (you can’t view the image through the lens) but with better options (albeit limited by (dome) port availability on housings of which there aren’t any as I speak - housings that is). The Olympus Pen E-P1 is the first such, virtually immediately superseded by the Pen E-P2. The E-P1/2 series is built according to the Micro version of the 4/3rds Standard (http://www.four-thirds.org/en/index.html) which allows for (a) a plethora of lenses, both Olympic and others and (b) a sensor significantly larger than that of P&Ss. That would be fine for those manufacturers who subscribe to that system but care must be taken to ensure that similar future products from other manufacturers have decently sized sensors.
• dSLR: "digital Single Lens Reflex" camera. These have the ability to change lenses (not underwater) and they have a viewfinder which looks "through the lens", in other words "What You See Is What You Get". Most do not use the screen to show you a ‘live’ view of what you are about to take a picture of, you look through the viewfinder and see the picture on screen after you’ve taken it. As a general rule the expense of the dSLR camera reflects its quality/the size of its sensor. Some do have what they call “live view” where they show what is the potential picture on their screens. There is a growing tendency, particularly from erstwhile P&S users, to want to use the screen to shoot the picture. Don't do it! The viewfinder is a far better option.
• Pixel: "Picture Element". It’s important that you have at least a rudimentary understanding of what happens here, so please bear with me: CCD (http://en.wikipedia.org/wiki/Charge-coupled_device) or CMOS (http://en.wikipedia.org/wiki/CMOS_sensor) (see also CMOS sensor (http://en.wikipedia.org/wiki/Cmos)) sensors make up the vast majority of such devices and are the imaging element of the camera. They (with the exception of the Foveon sensor, a CMOS design but arranged very differently but which has virtually no relevance here) are for our intents and purposes interchangeable. All P&S sensors are either 7.18mm x 5.32mm or 5.76mm x 4.29mm in size, which is really tiny. If you can, always try to buy a camera with the bigger sensor, bigger is better in this case. The other trouble with pixels is that, shamefully aided and abetted by all the manufacturers (to be fair, some were drawn into this by the other, more shameful, ones) the general public continues to think that more pixels is also better, not so in many instances (http://www.yorkshire-divers.com/forums/underwater-video-photography/64746-pixels.html). Particularly in the instance of P&S cameras less is often more.
• Pixel layout: another bone of contention. With the notable exception of the Foveon sensor all the others are laid out in a flat plane of cells, often called sensels (sensor cells), on the sensor. A certain Mr Bayer, then working for Kodak, came up with the concept now employed by all such devices of an algorithm that complements the cells (which become pixels) of the image. Bayer established a combination of R (red), G (green), B (blue) and G (green, again) sensels to make up one actual, real, proper pixel. Note that this is very similar to your TV except that there is no extra G in the TV picture. These sensels are set up in a rectilinear grid on the sensor. Confusingly all the camera manufacturers describe each sensel as a pixel (no doubt marketing-speak driven) so if your proposed camera has 8,000 "manufacturer” pixels you can divide by four to get to the proper pixel count of 2,000 which happens to be plenty in the case of a P&S given its tiny sensor.
• Noise: what in the good old days of film used to be called “grain”, a much better description because it actually means what it says but never mind. Noise means undesirable elements in your picture although these can, sometimes, perhaps partially only, be corrected in post-production. Noise is often a trade off between aperture, exposure and ISO.
• ISO: actually there is no such thing as ISO in digital photography but the manufacturers have decided that this is a convenient term to carry over from film photography for purpose of the exercise. In fact there is now a standard on this, but we won’t go there. The higher the “ISO” you dial into your camera, and some of them, particularly the P&Ss, are very restricted in this, the more “sensitive” the lens becomes - at the expense of highly undesirable noise.
• Exposure: the amount of time (that the lens is open) it takes for your camera to collect the data of the picture it is trying to resolve. Think camera shake and the need for a tripod if using long exposures.
• Aperture: the “hole” that the camera sees through to take the picture. The smaller the hole the greater the Depth of Field (DoF) (usually desirable), meaning that more of the image is in focus. Conversely, the larger the “hole”, the lesser the DoF.
• Lenses: IMO the beating heart of any camera, usually described by "x" mm at "f" something-or-another. The lower the numbers the wider and brighter (technically “faster”) the lens respectively. Wider and faster is the name of the game underwater. To put this into perspective a 24mm lens captures approximately a 136% larger viewing area than that of a 28mm lens and 213% larger viewing area than that of a “standard” 35mm lens whereas an f2 lens is about twice as fast as an f2.8 lens. 24 mm and f2 is the current benchmark for P&S cameras (and probably stood the rest of the manufacturers on their collective ear when it first saw the light of day). Having said that, the widest lens I have ATM for my dSLR is 10-20mm at f4-5.6, depending on where the telephoto of the lens sits. Very roughly speaking, if you use the zoom (we'll get to that) you progressively darken the image. As well, there are lenses and then there are lenses, Samsung uses Schneider lenses and Panasonic uses Leica lenses and I tend to stand at attention whenever I hear those names. This is not to say that the Canons, Nikons, Olympuses etc of this world are "bad" lenses but I certainly wouldn't entertain a camera and lens made by the equivalents of Messrs Yum and Cha (that no-one's probably ever heard of previously).
I should point out that in the Canon and Nikon lens ranges (for example) there are “professional” lenses, in Canon’s case usually designated by “L” and by Nikon as “Nikkor”. Some other manufacturers do the same thing and some use bespoke manufacturers of the likes of Schneider and Leica (you’ll find Karl Zeiss lenses sitting in Hasselblads). “Non” professional lenses can generally be taken to be of lesser quality, sometimes decidedly so. For example if I were to buy a dSLR I would almost certainly not buy any of the lenses, often called kit lenses, which may be available for that camera at little to no extra cost. This is hardly likely to be the case other than for the “entry level” cameras of the various major manufacturers, nor is it likely to be the case in the lenses of any of the alternative lens manufacturers such as Olympus, Sigma, Tamron, Tokina etc. The likes of these, incidentally, make some pretty good lenses which are right up there even though not all of them may be of that quality.
• WB: "White Balance" (or MWB – “Manual White Balance”), also useful above water, allows you to change the hue your camera shoots at depending on the light. Light, BTW, is not just bright or dull but it’s also not a subject for purpose of this exercise. A form of white balance is employed for example when you opt from "scenery" to "portrait" (etc) mode on a P&S.
Something to be considered if you are using (or wanting to use) WB is that it needs to be changed at different depths and in different conditions. Remember that different levels of colour get lost at different depths with red being the first to go? Therefore the WB setting at 5m will definitely not be the same as that at 40m - actually the WB setting at 40m is going to be well nigh useless anyway without an artificial light source! If WB is going to be important to you, choose a housing/camera combination on which you can alter it fairly easily. Try it out in the shop! If you want to choose a camera that has a video option, you may wish to think about whether the WB can also be altered on video.
• raw (http://en.wikipedia.org/wiki/Raw_image_format): some P&Ss can shoot in raw (which is a file format and the suffixes change, depending on the whim of the individual camera manufacturer e.g. Nikon have .nef files as their raw files whereas Sigma uses .X3F). It basically means that all of the information is made available to you rather than having the camera auto-convert it into a jpeg (usually) when a lot of the information is unfortunately lost. Importantly, raw allows you to adjust the WB in post processing and that is just so useful u/w because that turtle you're trying to get a shot of is not going to stick around while you fiddle with WB. It follows that if you have raw, and do post processing (as you should - see below) then you don't really need WB.
• jpeg (http://en.wikipedia.org/wiki/JPEG): stands for the Joint Photographic Experts Group and it was a brilliant concept when it first came out, still is for that matter and today pretty well everyone knows what a jpeg is and everyone can read it whether they are camera buffs or not. The trouble is that it is "lossy", look at the Wikipedia explanation of that, particularly the flower at the right of the screen. jpegs are perfectly adequate, indeed preferable if only because of their intrinsic failing, for sharing images over the Internet.
• Sharing images (a quick word): there are methods of doing this but the best, simplest advice I can give you to ensure that your images are not "acquired" by someone else is to convert them to jpeg at not more than 72 ppi (pixels, real pixels, per inch) at a small size, say 50 x 35mm which is perfectly adequate for Internet purposes, any higher resolution than that can't be seen on screen. At that size and that resolution your images are not really of value other than on the Internet itself - by which I mean that they can't be successfully printed other than at a much smaller size than that because printing an image requires 300 ppi - do the maths. Oh, OK: 50mm at 72 ppi = 72/300 x 50 = 12mm, that's less than postage stamp size, I assure you that most of the postage stamp manufacturers would make very sure of the identity and veracity of their image.
• Focus: is really self explanatory. With the advent of P&Ss excellent autofocus pretty well became mandatory and has come ahead in leaps and bounds. IMO anyone who even attempts to manually focus a digital consumer camera via that silly little picture on the back of it, especially u/w, will only successfully do so by gosh or by golly.
• Zooms: the majority of P&Ss have zoom lenses, some of which are absurdly long. The Olympus SP-590UZ "Bridge" Camera has a 26x lens of which you might be able to use about 1% u/w. My longest lens on my dSLR is 3x or 50-150 mm and that's a "proper sized" lens. Furthermore you can't use more than, say, 3x u/w (and that's stretching it) because unlike air you're not only drawing in the picture, you're also drawing in all the particulate matter suspended in the water in front of the image - commonly called crud (some people are less polite about it) but more correctly called Backscatter (http://en.wikipedia.org/wiki/Backscatter_(e-mail)). Apropos Backscatter, this is a perfectly good reason why a puny pop up flash immediately next to a lens pales in comparison to decent strobe arms (http://forum.divernet.com/album.php?albumid=2&pictureid=43) where the backscatter (mostly) reflects back to the strobes rather than the lens.
• Memory Cards and Batteries: I would assume you know about these but it is far more annoying underwater if you run out of card space or the battery runs down before you finish the dive so consider this as part of your purchase. I consider an empty card and a full battery a requirement rather than an option per dive. I have two batteries for the camera and several cards.
UNIFORM RESOURCE LOCATIONS
Some URLs to help you:
http://www.digideep.com/english - mostly tells you whether there is a housing available for a camera. Mostly - they're not prescient.
http://db.tidbits.com/article/7891 - tells you, in exquisite detail, how to buy a camera but it has no provisos for u/w use so that's very much a caveat. I would urge you to read Charles's other musings (http://db.tidbits.com/author/Charles%20Maurer) on that site as well, the man is a bit good.
http://www.dpreview.com/ - gives you as close to an unbiased opinion as you can get, provided you know how to read the technical jargon, which is pretty daunting for a newcomer. Again, above water only.
http://www.cambridgeincolour.com/ - one of the better "how to/what's that?" sites.
http://www.digitaldiver.net/ - a useful “comparison” site for bits you may, or may not, want to buy.
http://www.wetpixel.com/ - as the name implies - arguably the best underwater digital photography site on the web.
http://www.digigreen.net/forums/ - as the name implies - a UK site specialising in UK u/w digital photography.
HOUSINGS
The most important things to remember are that:
• not all cameras automatically have housings, it follows that before you buy the camera, check to see about the availability of a housing and whether that housing is going to suit you,
• some housings are, sometimes significantly so, more equal than others,
• housings can, and will, drown. That's a fundamental law of underwater photography so either live with it, check your insurance cover or forget all about it and
• housings require O Rings which do need to be looked after rather lovingly (http://www.digigreen.net/forums/other-stuff/7056-o-rings-general-gear-care.html#post35938).
As a generality, all manufacturer supplied housings are made by someone else, mostly Sea & Sea as it happens, which is actually no bad thing because at least Sea & Sea have a pretty decent track record - unlike the majority of camera manufacturers who certainly have a clue about cameras but zero clue about housings. These housings, however, are often fairly limited because the specs are those of the camera manufacturer where cheapest is often best. I do not here mean that they are necessarily built cheaply but I do mean that they can be limiting in other ways, particularly the addition of wet lenses and strobes (see below) or other loss of functionality. I point out also that the Canons, Nikons of this world are not interested in housings for their dSLRs entirely leaving that up to the housing manufacturers.