asked this on December 20, 2011 03:09 pm
I have been wanting a camera that has a depth of field from the tip of my nose to the far horizon for 35 years. With the refocusing ability the Lytro has, it would seem like all the information is there in the file, so is it possible to have a great depth of field? everything in the picture in focus at the same time?
I'm looking for a similar capability. In theory, a light field camera should be able to focus at any point in the image, but the Flash demonstrations on your website and others only allow focusing at two points--effectively, foreground/background. When you release your cameras, will your software be able to focus at any arbitrary point in the image, or will it also be limited to just two points of focus?
Playing with the sample gallery, its doesn't seem that the depth of field can be increased. And the depth of field is very limited because of the fixed f/2.0 lens. Perhaps the company's software will have a solution. But until then, I would guess that you would have to make 2 or more JPEG and then combine them using Photoshop. Clearly if you are a fan of the f/64 school of photography, this is not the camera for you.
Mark, Greg - We will support "all in focus" sometime after we ship:
lfeldman - "the Flash demonstrations on your website and others only allow focusing at two points": Not really. What's going on is that the pictures you are playing around with are web optimized versions of the light field file. The amount of depth information that was captured -- which is based on the skill of the photographer's 3D compositions as much as camera itself -- will result in more or less refocusability and thus more or less refocus points. Note that some of our web gallery pictures have 5 or more focus points.
The original file has all the light field information in it; as a result, it's pretty big (about 16MB) and would be an unwieldy beast to share.
"When you release your cameras, will your software be able to focus at any arbitrary point in the image" - Yes, you'll be able to do this on the Lytro Desktop software. But again, there is skill and learning involved; if you take a picture at 1x zoom, and your near subject is 5 feet away, they are effectively at infinity as far as we are concerned. There will be very little depth (and therefore refocusability) in such a picture.
A great lightfield photo at 1x will have your near subject about 4-6 inches away.
Eric Cheng just did a great video with the Verge at CES today I think you may enjoy (15 min). Eric does a very effective video demo and you can see a lot of what we can do (and even some things we wont' do at launch but are may happen in the future like perspective shift and all in focus)
I had the same question, but I see the obvious answer (assuming that the DoF isn't as great as you want it). Save multiple JPGs from your desktop software using multiple (and different focus points), then open Photoshop in automate mode for stacking (merge) and stack the images. Each of the focus points will be merged into one multi-focal image. It's a little more work, but you can set up a batch process to do them quicker once you save the multiple-focus images. That's my solution anyway.
@wayne, that's correct. Things like 3D, parallax, tilt shift, EDOF, etc are all computational features, that is, they are made possible by the software, All the data we need is in the .lfp file, we just need to get it out >;-)
And what that means, is that when we release a new feature on Lytro Desktop, you'll be able to apply those new features to every living picture you've ever taken, if you want. Refocusability has always just been the beginning...
And based on what Adam said above, you don't ever want to delete a picture from your desktop program just because it wasn't focused the way you wanted!
From Adam Gould: "The original file has all the light field information in it; as a result, it's pretty big (about 16MB) and would be an unwieldy beast to share."
In saying this, isn't one of your major selling points; "Friends and family can explore your living pictures – having fun interacting with them like you did – on nearly any device, like your mobile phone, your web browser or your tablet."
So they will only get a semi-useful living picture like displayed on your website then? As you have said the original file is just not feasible to be shared due to it's size.
The difference between the full 16MB file on your Desktop, and the web-optimized version you're sharing is marginal, from the perspective of most living picture consumers. There's really just not huge refocusability differences between Desktop and your shared images.
You want to keep everything you've taken, in raw form, for when we release new features like parallax. Those new features need access to all that juicy lightfield data that is not contained in the web-optimized versions, to generate the living pictures with parallax. In other words, you'll need to re-process and re-share any pictures you've already published, that you want to have that new effect.
Does this address your concern about only getting "semi-useful" living pictures?
It seems to me that as the rendering engine progressively focuses from near to far that only rendering those pixels and adding each layer iteratively would produce a near perfect image or you can set a depth range to compose the subject with the remainder out of focus. I can think of a way to primitively do this by using layers in photoshop. I hope this feature becomes available, as I've been hesitant to purchase this camera without it. Also I want to produce a still image in jpg or raw to work with elsewhere.
any idea on when new software will be released ?
and if it will be available to those who purchased now? I don't want to get myself
involved with another 'Apple' type company. (a few months after buying original iPad, out comes
iPad 2 with camera...there goes my $500. down the drain. Apple is now dead to me and many others)
I do not know the date, but I know that as I had my camera since about April, I have already received number of firmware updates (I think it was 2) and few Desktop app updates for both Mac and Windows. I totally agree with you about the Apple approach - even though you support them, they'll just keep making you want to buy new release after a new release. I do believe Lytro will have to, at some point, release a new hardware platform, to improve performance, resolutions, and so on... but at this stage it looks like they do have a lot they can change via software on the computer and on the camera, to keep growing capabilities of this device... As prices of everything go up, $400-$500 for a camera is not that much.. but as this is the most expensive camera I have, I certainly would not want to be pushed to buy a ver. 2 any time soon..... Let's just hope they think that way too..... ;)
I'm glad to hear the file size is manageable at 16MB- obviously large for convenient web toting- but it gives me some idea.
Thanks Adam Gould
@joeciccone, our software updates are available for all Lytro camera owners.
@max, the 16GB original files are not designed for sharing; we create a web-optimized version that is specifically for sharing, at about 200-400K each. The original files are really just for viewing and working with in the Lytro Desktop software.
This could be a pain in requiring a lytro-specific plug-in to parse, but at this point couldn't someone develop a nice multi-resolution analysis basis set (like curvelets) optimized for light field data?
Or is something like this already being done to allow sharing of the images? Would like to see less-lossy compression.
If you go into 'perspective shift' mode the entire image appears to be in focus. You could simply copy the image at that point in order to have the maximum DOF for that image.
What I would like to see in the desktop software would be a 2-slider system for DOF control; one for the 'depth' of the DOF and the other for the center of the DOF. This would make composition easier and more intuitive.
Jeff! A good idea! Of course not every image has as many focus layers as others.. or at least they do not differ.. but in once that do have the differential, it would be cool to use your idea to decide what's in focus, and how deep of a layer of "all in focus we got"... That would be quite cool for those who do some really creative work...
I get all-in-focus images by p-shifting them, then clicking and dragging to shift the perspective to the one I like. A screenshot will have the image all-in-focus. However, a jpeg extraction still will not. This is a feature that I hope will be included in a release soon.
Support Ticket System by Zendesk