The first thought I had when I saw the new Google Camera app was "wow, this might kill that light field camera thing that came out a couple of months ago that I haven't heard of since". I then installed Google Camera, had a play, and it confirmed my thoughts. The new Google Camera app is awesome and it works with my current phone. Extra hardware to carry around = zero.
The original Lytro was a cool idea, but I think the slow time-to-market caused it a dragged out death. Google put the last bullet in.
The difference is that Google's effect is faked in software and relies on the huge depth of field of a tiny sensor. The Lytro actually does what Google is simulating.
That's like saying my MP3 player from the early 2000s had a microphone and cost $200, so why would anyone buy thousands of dollars of professional audio equipment anymore to record music?
>> The Lytro actually does what Google is simulating.
But wait a second. We're talking about the Google Camera lens blur function, right? That feature is only really simulating out of focus blur, which is not really the same as what Lytro does.
What Lytro does more akin to focus bracketing, isn't it?
Right. What I meant was that when the Lytro's image is 'focused' on a subject, the background actually is out of focus instead of just having a fancy blur algorithm run on it to simulate being out of focus.
Well, but maybe it's possible to make a compromise? I mean, for a stationary subject the Google approach (camera movement) and the Lytro approach (microlens array) captures the same data (with the caveat that Google apparently only uses a linear path).
I could imagine a system which captures a whole lot of frames with some camera shake, identifies which parts of it moved, and then computes "real" lightfield bokeh for the stationary background while filling in with just image-space blur for the gaps.
It's definitely expensive, and it has been slow to iterate. But the nasty-looking fake bokeh produced by Google's app doesn't even make conventional large sensor cameras obsolete. This thing is actually pretty impressive and lets you do stuff you cannot do with any other camera.
>> But the nasty-looking fake bokeh produced by Google's app doesn't even make conventional large sensor cameras obsolete.
The thing is, the fake bokeh is nasty looking to people who know how to judge bokeh, which is a fairly small portion of camera phone users. To most of them, it's probably good enough for what they are trying to accomplish. And in that regard, while it won't make a conventional large sensor camera obsolete, it'll still have enough people saying "my smartphone can do that, why would I bother buying a big camera that doesn't fit in my pocket?"
>> This thing is actually pretty impressive and lets you do stuff you cannot do with any other camera.
I agree with this, but I often wonder if the Lytro is a solution looking for a problem.
Hate that you are getting downvoted, as I think this is actually a discussion that needs to happen. Many people will have a hard time distinguishing these technologies. Especially on paper.
My take is that this camera will have some major advantages in shutter speed and quickly taking successive pictures.
As an example, at my child's soccer practice, it is not uncommon to fire off about 5 shots in any 2 second interval. It is also not uncommon to have 4 of those shots be out of focus. Heck, all five.
This camera could solve that. The phone solution fails, as I can't take that many pictures that quickly.
>> As an example, at my child's soccer practice, it is not uncommon to fire off about 5 shots in any 2 second interval. It is also not uncommon to have 4 of those shots be out of focus.
I'm not sure Lytro's necessarily the right the solution for this either, at least not today.
I know I'll get laughed at and mocked for saying this, but I bought an inexpensive Nikon One camera for a very similar use case. I got tired of carrying my DSLR around on doggie play dates, and while the Nikon One may suck at an incredibly long list of things, the one thing it still does better than just about every other mirrorless camera out there (even with the Fuji XT1 and Sony A6000 now on sale) is AF on moving subjects. It can track a dog running towards me (the hardest part about AF in this use case) just about as well as my DSLR.
If you don't believe me about the moving subject AF, Thom Hogan agrees:
Oh, I should have made this clearer. I have not used this camera. I have tried my camera phone in frustration a few times, but I am usually only happy with the shots I get from my DSLR. To the point that I picked up a used 5D Mark II. I'm very happy with this camera, but getting focus on far away shots with a zoom can be difficult. Solving that would be nice.
Which is why I think this conversation needs to happen. I'm not sure what the benefits are. I'm just not dismissive, either. Nor do I think the camera phone will completely destroy the DSLR any time soon.
(I do think it will eventually happen. Maybe not the camera phone, per se, but camera sensors could advance such that having many different cameras will be for novelty more than utility.)
The 5DMKII isn't going to win any awards in AF speed (it is still a great camera, imo) but you can do some stuff to mitigate your issues. Try shooting with a smaller aperture to get more depth of field, and consider learning how to manually zone focus. If you know the distances for which your lens is "in focus" for a particular aperture, and can evaluate how far away your subjects are, you might get better results. I think this is how they did this in the "good old days".
It is more work, requires learning (I think it's a useful skill, but that's just me), but if you've got AF limitations and don't want to throw money at the problem (i.e., get an MKIII), it doesn't hurt to try.
Oh, certainly! Learning to use my camera has been paying dividends that I can't really explain. I expect I'll keep getting better.
Biggest thing for me to learn now is how to work with a zoom lens. Then, get a better one. (Well, I say zoom, I really just mean telephoto. Right now I have a zoom one, but expect to move to a prime one eventually.)
And I think you are dead on, learning how they did things in the "good old days" is a huge skill that has been helping a lot. Things are a little tougher when my child is effectively running around at random right now. I expect that to change, as well.
The "big trick" to sports photography has always been learning the sport, and that's only a little less crucial now with, say, a 1DX or a D4/D4S than it was back in my manual-focus days. However, that's not going to be of much help when the kids you're photographing don't know the sport. Expect a comparatively large number of failed shots, and learn to laugh. A 400/2.8 L isn't going to help a whole lot until there's some statistically-valid chance of anticipating the action. (And it's godawful heavy and awkward as well as really good used car expensive.) As your photography develops and your athlete develops, you'll know when (or if) it's time to go for the big guns. In the meantime, a fast 70-200 (with a good teleconverter for some shot types) will fill the bill, and unless you're printing huge, don't be afraid to crank the ISO a bit to keep the shutter speed down. (Look at the pictures, not the pixels. There ought to be some sort of license required to zoom in to 100%.)
I've definitely learned this one the hard way. At first I was excited about having everything at super low ISO. Now, I take a few quick shots to see what the lowest I can get away with and still get quick pictures.
Heck, often times I'm happy to just take videos. Really liking how well they turn out with this camera.
Actually, if you're on a soccer field, and the kids are far enough away, there's a good chance they're focused at infinity anyways, so you may not even need to focus (although your camera should be set to MF) as long as you've got the aperture small enough to give you wiggle room on the DOF.
The Lytro still needs to be focused; the trick it uses is to focus to the hyperfocal distance as you zoom, and then it can capture the equivalent of focus stacking fore and aft of that point; but only on a finite number of focal planes ( fewer than 6 if I recall ).
So... your photos could end-up being nearly in focus, but also lacking in detail ( low-res sensor ) and blown-out ( poor dynamic range ).
That makes sense. I'd be curious to know of any wiggle room available in that regard. Seems it would be nice to do a few focal points that are "close" to each other near where you are focused, combined with a few far away ones. If that makes sense. Would make it possible to do sort of HDR images from a single shot, right?
And, regardless of that, this still goes a much farther way towards "solving" this problem than the camera phone. Right?
The simplest solution to your soccer practice photography dilemma is to get a camera that can focus quickly enough (or maybe to learn how to zone focus). There are many cameras that can do this, but the problem is that none of them are a smartphone.
As I noted in the sibling post, I have a decent enough (I think it is a dang good one, actually) camera for this. Right now, I just need more practice with my tools. We have only done three practices, my daughter is only four. :)
The example was more to get an idea of what this could help with. In short, if you are taking pictures of things in motion, or multiple shots in succession, the camera app that was recently displayed is nigh useless. Right?
The 5D Mark II is an awesome camera in a lot of ways but it has fairly slow autofocus speed for a camera in its class.
It is outpaced quite a bit by even "lesser" (more consumer oriented) cameras like the 70D (though as with anything this is a trade-off since the 70D is a cropped-sensor, so IQ not quite as good under ideal situations, worse low-light performance, etc).
Here's a photo I took a couple of months ago with the 70D, one of a series of 5 that were all pretty well focused (considering the subject was a cheetah running at nearly full speed) using the "AI Servo" autofocus mode:
(To be fair, such photos still aren't exactly point and shoot on my 70D, you have to be fairly decent at panning at that speed)
You do have a point though in that the current situation for me is needing to have multiple cameras (70D for fast action, Sony A7 for relatively still shots with very high image quality, Canon EOS M for carrying around everywhere). I don't think this Lytro comes anywhere close to changing that situation, though computational imaging in general will probably solve the issue eventually.
Yeah, I don't fully understand the tradeoffs yet. I know that it has much better noise level than my last rebel. And the full frame sensor is very nice. (Though, now I want a good 50mm lens).
I'm definitely game for any advice I can get. I hate that I let that example detract from the discussion, which is for average photographers, something like this could help a lot. Doesn't sound like it is fully there, yet. But it is promising. Especially in tandem with the app idea. Exciting times are ahead for getting folks into photography.
The original Lytro was a cool idea, but I think the slow time-to-market caused it a dragged out death. Google put the last bullet in.