How far away are "really good" tiny cameras for still/video?

With the news of patents for new amazing sensors and flat lens elements and such, the promise of extremely small camera systems has us all hoping for super tiny cameras that produce excellent images - and video. But how far away is this?
At the moment I dive with a terrible old GoPro that lets me down in all kinds of ways (wild white balance changes with no controls, multiple camera errors in operation, poor image quality, single wide-angle lens only etc)... but it's what I have. 
On the surface, I am really happy with my micro-four-thirds system with pro lenses from Olympus, particularly because the smallish size and good quality suits my needs... a very different experience to underwater.
I could upgrade to something like the GoPro Hero7 (tiny, simple, cheap-ish, but still lower quality than the Oly and without the ability to use different lenses) ... or I could upgrade to a big housing setup for my Oly gear (expensive, big, clunky and prone to damage, but yielding great images). If I decide to bite that bullet, I will want to be in it for around 10 years, or it's simply too expensive for me.
If a truly revolutionary little system that offers tiny size and amazing quality is only 5-6 years away, I would regret investing in the big housing system. But how far away is it? (I'm thinking of something that could compete with micro-4/3 or APS-C with pro lenses, but at the size of a GoPro)

The thing is, the camera is just a small part of the overall setup. Lighting conditions underwater are extremely challenging, which means that bringing artificial light - and lots of it - is a requirement. Once you hang a few big strobes (and they've gotta be big, because they're dumping lots of energy out) off your rig, the tiniest camera will turn into a boat anchor. I mean, if we're talking science fiction, you could imagine several autonomous, AI-controlled drones swimming around you, positioning themselves to give you the light you need, when and where you need it... all it would take is some revolutionary developments in AI, robotics, energy storage, material science and probably half a dozen other fields. Not physically impossible, but probably won't happen in our lifetimes. Until then, if you're hauling ten pounds' worth of lights, might as well bring a decent camera with them.

You do have a point. 
At the moment I have a really small but very powerful video light on a tiny tray that has been "good enough" for the cold-water, somewhat murky diving I do (with the old GoPro). To really freeze action for quality photos I'd have to add one strobe at the very least and then, as you reminded me, the whole setup really does get bigger.
But lights and arms are fairly robust and foldable. the camera housing is a big lump (at least 6X size of GoPro housing, with multiple, large O-rings etc) and I'm not sure how robust. Certainly any dome port and O-rings need serious care and attention.
Ha ha - there's no silver bullet out there yet...

kelpdiver wrote:

tiny cameras aren't reinventing laws of physics. They have been stuffing a lot of software routines to compensate. So I don't foresee a sudden change in 5 years that will give us FF joy in a GoPro footprint. And if it did, wouldn't be cheap, and wouldn't happen in one generation.

Get the housing for the 4/3rds you already own. It's a perfectly viable UW platform and the lenses pack nicely for air travel. If you're only using an old GoPro currently, you need to get your feet wet anyway. You may find yourself wanting something different after a bit of experience.

Alternatively, you may be able to get housing + older 4/3rds for a nicer price and still be able to make use of the nicer Olympus Pro lenses you have. Though I will say, for that extra f stop, you often incur a pretty big weight penalty, and sometimes the need for a more complicated port than with the 'regular' oly/panasonic lenses. No free lunch.

If we examine DSLR/mirrorless cameras over the past 5 years, the jumps in capability have not exactly been inspiring.
Yet it's still tempting to believe that the sensational promises hinted at in the past couple of years (tech news stories, patents etc) will truly revolutionize cameras.  And I simply cannot ignore the astonishing leaps of capability in mobile phone cameras in those same 5 years.
Regardless, you're absolutely right about there being no free lunch.

Scott G wrote:

You do have a point.
At the moment I have a really small but very powerful video light on a tiny tray that has been "good enough" for the cold-water, somewhat murky diving I do (with the old GoPro). To really freeze action for quality photos I'd have to add one strobe at the very least and then, as you reminded me, the whole setup really does get bigger.
But lights and arms are fairly robust and foldable. the camera housing is a big lump (at least 6X size of GoPro housing, with multiple, large O-rings etc) and I'm not sure how robust. Certainly any dome port and O-rings need serious care and attention.
Ha ha - there's no silver bullet out there yet...

I dive with a Sony A6300 in a fairly substantial SeaFrogs polycarbonate housing with a tray and a pair of strobes, and while this is highly subjective, I feel that most of the bulk - even accounting for an 8" dome on the housing - comes from the arms and strobes. Even folded up, the four 20cm arms with strobes hanging off them are quite unwieldy and add substantial underwater drag.

As for the 'really small but very powerful' video lights, they are several orders of magnitude less powerful that strobes. Comparative testing indicates that it will take about a million lumens of constant output to match the prompt brightness of a Sea & Sea YS-D2. Consider this: a typical strobe running off 4xAA batteries is rated to about 300 full-power flashes, give or take a bit. Each flash is about 3ms long. This means that if you were to run it constantly (physically impossible; it would literally explode if you were to try something crazy like that, but this is just a thought experiment) , the batteries capable of running a regular light for a couple hours would be drained in less than a second.

Regarding the advances in smartphone photography, keep two things in mind. One: phones have a much, much lower starting point - the chasm between the capabilities of phones and digital cameras of early 2000s has certainly shrunk, it's not going away anytime soon - you can't cheat basic physics, and the actual resolution of the camera's sensor is limited by factors such as the airy disk. Two: the huge advances in phone camera image quality have largely come from computational photography - increasingly sophisticated software in the phones interpreting and processing the camera output. This is achieved by huge multinational companies pouring literally billions of dollars into R&D on this - but underwater, the camera output is highly, let's say peculiar, and all that R&D has little relevance. There is the 'Dive+' app that processes underwater images on smartphones, but don't expect miracles.

Finally, keep in mind that no currently available camera sensor is capable of syncing with strobes without the aid of a mechanical shutter. This isn't physically impossible, but currently available electronics - even in extremely expensive cameras such as Sony A9 - are far too slow to read out the sensor output within the duration of a flash pulse. This is something that will probably get solved in the future, but how long it's going to take is basically anybody's guess at the moment. Right now, no smartphone or action camera on the market can sync with xenon strobes, and LED flashes are far too weak to be useful underwater. There were a few smartphones with xenon flashes and mechanical shutters made in the past, but none of them are currently in production.

Barmaglot_07 wrote:

Scott G wrote:

You do have a point.
At the moment I have a really small but very powerful video light on a tiny tray that has been "good enough" for the cold-water, somewhat murky diving I do (with the old GoPro). To really freeze action for quality photos I'd have to add one strobe at the very least and then, as you reminded me, the whole setup really does get bigger.
But lights and arms are fairly robust and foldable. the camera housing is a big lump (at least 6X size of GoPro housing, with multiple, large O-rings etc) and I'm not sure how robust. Certainly any dome port and O-rings need serious care and attention.
Ha ha - there's no silver bullet out there yet...

I dive with a Sony A6300 in a fairly substantial SeaFrogs polycarbonate housing with a tray and a pair of strobes, and while this is highly subjective, I feel that most of the bulk - even accounting for an 8" dome on the housing - comes from the arms and strobes. Even folded up, the four 20cm arms with strobes hanging off them are quite unwieldy and add substantial underwater drag.

As for the 'really small but very powerful' video lights, they are several orders of magnitude less powerful that strobes. Comparative testing indicates that it will take about a million lumens of constant output to match the prompt brightness of a Sea & Sea YS-D2. Consider this: a typical strobe running off 4xAA batteries is rated to about 300 full-power flashes, give or take a bit. Each flash is about 3ms long. This means that if you were to run it constantly (physically impossible; it would literally explode if you were to try something crazy like that, but this is just a thought experiment) , the batteries capable of running a regular light for a couple hours would be drained in less than a second.

Regarding the advances in smartphone photography, keep two things in mind. One: phones have a much, much lower starting point - the chasm between the capabilities of phones and digital cameras of early 2000s has certainly shrunk, it's not going away anytime soon - you can't cheat basic physics, and the actual resolution of the camera's sensor is limited by factors such as the airy disk. Two: the huge advances in phone camera image quality have largely come from computational photography - increasingly sophisticated software in the phones interpreting and processing the camera output. This is achieved by huge multinational companies pouring literally billions of dollars into R&D on this - but underwater, the camera output is highly, let's say peculiar, and all that R&D has little relevance. There is the 'Dive+' app that processes underwater images on smartphones, but don't expect miracles.

Finally, keep in mind that no currently available camera sensor is capable of syncing with strobes without the aid of a mechanical shutter. This isn't physically impossible, but currently available electronics - even in extremely expensive cameras such as Sony A9 - are far too slow to read out the sensor output within the duration of a flash pulse. This is something that will probably get solved in the future, but how long it's going to take is basically anybody's guess at the moment. Right now, no smartphone or action camera on the market can sync with xenon strobes, and LED flashes are far too weak to be useful underwater. There were a few smartphones with xenon flashes and mechanical shutters made in the past, but none of them are currently in production.

Great info and explanations here - thanks! I have no argument at all. I understand that to get really good photos, I would need much faster exposures, which necessitates at least 1 strobe and therefore a proper camera. No question strobes are vastly more powerful than video lights. I currently live with slowish exposures and really low IQ.
My real source of head-scratching is: how big is the quality gap between a bit of money and a whole lot of money? I won't be selling my underwater shots, no matter how much I'd like that. So why not look at an actual nuts and bolts comparison?
 Upgrade 1: upgrade to GoPro Hero 7 and add another Kraken 3500S (+ tray) (already have one of these lights) to double my illumination. I'm already burning out lightly coloured objects close to the camera with 1 video light, but that's mostly because I have to point the light centre-frame. with two lights angled outward I avoid a lot of backscatter and blown out highlights for close objects. I'd get better photos and MUCH better video than I do now for 2 reasons - better camera (than my current Hero 3+), and double the light. Cost for me - in the range of US$1400.
Upgrade 2: get Ikelite housing (+ zoom ring + dome port), tray and a strobe for my Oly. Continue using my Kraken video-light. Cost for me - US$3200 + +
Is the increased IQ from faster exposures and much more light, and the ability to use my super-wide and normal lens (yes - I checked) worth more than double the price (and a lot more bulk, weight, and delicate dome port worries)?
I know it's a subjective question, but I'm hearing about other divers (who I know have big-gear setups) shooting 90% of their dives with a GoPro. Obviously I'll try to track down some of these guys and ask what their logic is (they probably also have larger APS-C or full-frame cams, so it isn't an ideal comparison, but the logic is likely related).

Thanks for any feedback you might be able to offer

Scott G wrote:

I currently live with slowish exposures and really low IQ.
My real source of head-scratching is: how big is the quality gap between a bit of money and a whole lot of money? I won't be selling my underwater shots, no matter how much I'd like that. So why not look at an actual nuts and bolts comparison?

[if you have a red X instead of an image, click on the original size link]

OK, for comparison I've shown a few 'representative' shots I've taken over the years with a series of increasingly-better underwater cameras. Doesn't compare the Gopro, or the current cheap options, but you can get an idea of some relative differences.

Keep in mind that strobe makes a HUGE difference to what you can accomplish underwater. I started with an external strobe for a few shots clear back in 2009, but it wasn't until I got a Sony RX100 II that I added dual ttl strobes.

While you can see a clear improvement as the cameras got better, you can't really get a feel for autofocus performance and limitation. ALL of the point-n-shoots I've used (most modern the Sony RX100 II) are pretty much alike in autofocus performance. Too slow. Can't pinpoint the spot you want to focus on. (Focus areas too wide, too hard to move around). The DLSR's I use (D810 and D850) are just night and day better. Absolutely no compasion. Like a jet plane compared to an Amish buggy.

I started with a Canon SD630 in 2008, but I don't have a good shot to pull up at the moment. It got lost traveling from Maui to Phoenix in 2009. I was happy to replace it with a camera that offered vibration reduction (VR to Nikon, IS to Canon).

Canon SD870is:

Canon s95: About the same autofocus limitations. I shot it a lot with on-camera flash through a diffuser attached to the housing. (Canon dive housing).

Blown highlights are common with older point-n-shoots. Macro capability pretty limited.

Canon s120:

The s95 eventually was replaced with an s120, but I've not used it underwater very much. By the time I got it, I was using a Sony RX100 II underwater. The s95 failed (zoom would no longer retract, which was how both my SD870's failed.) But the s120 is clearly better in low light than the s95. Also, the (Canon, probably made by Meikon) housing came with a fiber optic adapter for external flash, which would be useful.

Pointing at a seahorse

Sony RX100 II:

I finally got sick of shooting without external flash, and I decided on a major upgrade. I would get dual strobes, a focus light, a Nauticam housing and a one inch sensor. I got all I wanted by adding strobes, but the focus speed was still pretty much the same as the Canon point-n-shoots. On the other hand, while the Canon menu system is well-thought out and implemented, the Sony's is just a nightmare of menu options and small print everywhere.

Nikon D810:

One year and three dive trips after getting my $5000 RX100 II rig, I was sick of the point-n-shootiness of the camera. Specifically slow autofocus and inability to pinpoint the autofocus spot. And I was pretty sick of trying to read the menus underwater. I resorted to cranking the LCD display to almost full brightness, which meant my shots looked grossly overexposed. (I just resorted to judging exposure solely by the histogram).

So I spent big bucks on a DSLR rig. Finally. I was able to move the dual strobes and focus light with no more expense than a new set of fiber cables, but the housing was $3400 or so, and the 230mm dome port a bunch, along with 105 macro port. All in I was around $10,000 for the rig. But the results were worth it to me.

Macro was all I asked for. Wide angle is more complicated, and I'm still working on it, but with the 16-35vr and a 230mm dome port (and a bit of edge cropping), I was getting decent wide angle shots.

Nikon D850:

After 3 years of shooting my D810 underwater, I was happy with the camera, but the housing was 'due' for an expensive maintenance. I had been shooting a D850 above water for some time, and I wanted the autofocus ability underwater. Not getting any younger I decided to upgrade to the D850 underwater, with all the same equipment otherwise. (I did replace the focus light with a new Kraken).

The new camera gave me even more cropping power - very important to me in macro as I normally can't get close enough. 46 mp rather than 36 isn't a huge change, but it's welcome. The biggest difference is autofocus performance, particularly in dimmer conditions.

For wide angle, I tried something new, and in my opinion it worked spectacularly well. I added a 77mm Sea and Sea Internal Correction lens to the front of my 16-35 (screws on like a filter), and it massively improved the edges and corner resolution. Without this lens I was invariably cropping the edges off, but this option makes that pretty optional.

Other considerations:

Lens selections: With a full frame Nikon DSLR, I'm extremely limited in my lens choices. I have but two options - wide angle (rectilinear and fisheye) or macro. There simply are no mid-range lens options, particularly zooms. This is different for DX / APS-C format cameras. I shoot mostly macro at this point, and I often wish for a mid-range zoom.

I carry my Sony RX100 rig as a backup for most dive trips as the equipment is small enough to do that. (Can't really pack two dslr housings). I had one occasion to use it between D810 dives in Jamaica last year. I shot a beachside wedding between dives with my D810 and didn't have enough time to swap the camera in and out of the housing. So I dove twice with the Sony again. Pros and cons there - I liked having the normal zoom to get 'typical dive ambiance' shots. I immediately noticed a dropoff in lens IQ though after three years of the D810. Still, I got a couple of shots I liked that would have been the wrong lens on the D810:

Sony RX100 II

Sony RX100 II. I really like this shot, but I can see clear lens issues on the left side there, compared to my Nikons

Packing considerations:

Going to a DSLR will be an exercise in travel logistics. Adding a big dome port for wide angle makes it hugely more of hassle to pack. I show how I disassemble and pack the rig here: http://www.cjcphoto.net/uwcamera/

Wow - thanks Craig for this very ample contribution!
I love your shots - obviously there is a huge difference with a great camera and strobes! But as your gear explanation clearly shows, it sort of becomes its own monster. All good food for thought - thanks again!

Your crayfish in the sponge and green moray shots are absolutely outstanding.

They do become a monster, but in a sense that happens very quickly underwater - more so than above.

What I mean by that is that once you have a camera fitted with dual strobes, the whole diving experience is altered.  You are really tied to holding onto that camera, unlike something small that can be dangled from a wrist easily.   And the drag underwater is more from the arms sticking out than a big housing.   (I turn the rig sideways and stick one arm between my legs and the other along my chest to streamline if in a hurry).

When I switched back from my D810 to my RX100 that morning, I just moved the strobes and focus light over.  It was supremely easier to lug onto and off the boat, but underwater it was about the same, except the viewfinder, in terms of finning along.

But the lack of lens choices coupled with very large and heavy parts does make it a pretty specialized system when you get into an FX camera underwater.

Good points once again, Craig - thanks!

I sure enjoyed browsing through your online galleries! Great stuff - and really nice to see the many travel memories you documented.

There are a lot more I never posted publicly, or linked on my web site.    Travel got so intensive for a while there with Connie that I couldn't keep up with the travel page.  Then again 16 shots per page and 75+ pages is a lot of shots already in that one.

I'll finish that up, hopefully, but it's very difficult to go back and look at my travel shots with Connie, much less spend time editing.   It's why I never went back and finished up my Scandinavian trip shots - too painful.

I'll be heading back to Roatan again in about a month.   Two more weeks shooting the D850 underwater.

completely understandable - sorry for your loss.
Keep up the snapping and memory making though - you do wonderful work! Happy Travels,