Lytro’s Illum Is a Camera for Serious Photographers | Hacker News Reader

Lytro’s Illum Is a Camera for Serious Photographers

nytimes.com

"Lytro’s Illum Is a Camera for Serious Photographers"

Meh, not really. I mean, it is interesting, but the title isn't really correct and also not justified by anything written in the full article.

I'm a fairly serious hobbyist photo-taker and while the Lytro appeals to my geeky experimental side it has some major issues for use in "serious photography":

While being able to refocus the image is nice, no matter where you set focus the sharpness is pretty shitty relative to nice, sharp modern conventional sensor/lens combos. The images I've seen from the Illum are better in this regard than the previous oddly shaped box camera Lytro released, but still quite a bit off of conventional digital cameras.

4 megapixels, while perfectly fine for web-based images (though with the acceleration in adoption of HighDPI displays, even this is changing), is seriously limiting if you intend to make prints much larger than a 4x6", especially if you also have to crop at all.

Also, while the Canon 70-200mm is ~$2500 as noted in the article, Tamron makes an f/2.8 70-200mm that is for all practical purposes just as good for $1,400 which is less than the price of the Lytro even if you add in the cost of a used older low-end APS-C DSLR body (and even an older low-end DSLRs will have many more megapixels and overall image quality at HighDPI or print sizes) to go with it.

I still find the Illum (and light-field photography in general) interesting from a tech perspective and and would probably buy one to fuck around with if they cost like $400, but in the price range the Lytro retails for I would recommend anyone "serious" about photography pick up the Sony A7 with the 28-70mm kit lens. You won't be able to refocus or perspective shift the images, but the images will be sooo much nicer and more flexible if you intend to print or display them at very high resolutions.

I really wish I could be more positive about the camera because I want to see where light-field photography ends up down the road if it is pursued, but the $1,600 price really puts it in an odd place where it is too much (for most of us) to buy as a toy but too limited to buy as a serious piece of photography gear.

RE the lens, yeah, it's worth pointing out that these are much higher performing optically and intended for use on full frame or APS-C size sensors. Rather a different animal.

You don't even have to spend that much. The f/4 version of Canon's 70-200 can be had for around $600 new and it's a fantastic lens.

Yup, this is true.

The way the article used the lens was as comparison to something similar to the effective f/2.0 the Lytro camera uses; if you're willing to go 2-stops slower (f/4) instead of 1-stop slower (f/2.8) then the lens costs can be reduced even more substantially.

In any case, the original comparison isn't all that useful anyway since the Lytro is a very different beast than a traditional digital dslr, but the bottom line remains while I think the Illum is cool, $1,600 is about 4x what I'd pay for it (which isn't to suggest the cost isn't justified based on BOM and/or R&D costs, it probably is, but that is irrelevant in calculating its value to me personally).

I love this technology and it is fascinating as hell, but as a photographer it leaves me absolutely cold. I just cannot seem to find a good reason to use it other than the simple novelty one feels when first navigating one of the photos. That doesn't remain novel for long, and the contrived compositions that make the best use of the technology tend to reflect a lack of what, in my own opinion, goes into a good photograph. The results of the camera are simply not compelling to me.

A friend of mine does event photography for swing dancing. He's looking at getting one because it's really hard to catch focus when things are moving so quickly (in dimly-lit halls). There are photography niches out there where features like this can be a necessity.

Another use-case I've heard about was wedding photographers using depth of field as a Ken Burns-like effect for videography or on their webpage.

Both are narrow use-cases, but I too find the technology fascinating and hope it finds a sustainable use-case.

Why not set up a strobe or two at relatively low power in the corners and use those instead?

this won't really help him catch focus of moving objects... because they are moving. only a faster shutter speed can help there.

But when someone is moving closer/further away from you (as probably happens when you're a stationary photographer among dancers) the ability to not worry about what's in focus right now seems huge.

> But when someone is moving closer/further away from you (as probably happens when you're a stationary photographer among dancers) the ability to not worry about what's in focus right now seems huge.

But the subject is still moving quickly - the Lytro can't compensate for that.

Dark, high movement indoor scenes (dancing, sports, etc) are always some of the toughest situations to shoot in, and are probably one of the times where gear can make a substantial difference between getting a "passable" shot and getting a completely unusable one.

Stop down as much as you can, zone focus, push the ISO up to maintain the minimum shutter for the activity at hand and shoot. You can get push ISO higher if you're not looking to print big (as in, 16x20" and up) too, which can help give you more breathing room.

Agreed entirely, although it has to be said the f/2.0 will help a little.

(Of course a 4MP image will not be as good as a real cameras, but low-light performance should be reasonable with that aperture - though I've not seen anything mentioned about ISO ranges.)

I will go hug my 70-200 f/2.8 now, it truely is an amazing lens :)

I took a Nikon 180mm f2.8 [1] to the world cup and it was phenomenal[2]. It's less bulky than the 70-200. Now, if this Lytro has real good ISO range and decent glass, then it might be good for a travel kit. Just depends on how big it really is.

-- 1: http://imaging.nikon.com/lineup/lens/singlefocal/Telephoto/a... 2: https://www.flickr.com/photos/djkianoosh/sets/72157644825335...

I agree with the other poster - I think your problem is motion blur rather than focus - if you have a decent lens (one of the "pro" lenses, rather than kit) the focus is very fast, smooth and does well in low light for both Nikon and Canon.

I agree completely. Despite submitting the article, I don't agree with its conclusions at all. Give me the ability to fix my focus fuckup on a 20MP image at a wide aperture like f/1.4, and then I might be interested. But for now this still seems like a novelty item.

That might not be that far off.

The D800 is 36 megapixels, if we ever hit 50 megapixels then it might be very possible for a Lytro camera to be 15MP or so. That's perfectly acceptable, resolution-wise. The current DoF of the lens is limited by the fact that it's a superzoom, which trades off aperture for focal length. Considering that they capture the entire light field, a traditional 24-70mm might very well be effectively faster than the f/2 superzoom they currently have, giving us a f/1.4 DoF and brightness.

I agree, the refocusing is somewhat gimmicky, but how would you like a 30-250mm f/2 lens that never misses focus. To me, this is where the Lytro shines. It enables you to use a lens that would be unusable with a normal sensor.

At least with the original Lytro camera I have, "never misses focus" is rather overstating the case. When the pictures come out well, they're stellar. But that constant f/2 aperture means the depth of field is minuscule, and it's surprising how often it takes pictures where none of the focal lengths you can choose from is particularly compelling.

I think it would be much more interesting for film than for photography, adding a huge amount of creative power to the editor and cinematographer in post-processing.

EDIT: I wonder if there are any stop-motion animators who have tried working with this camera.

In related news, people are working on light-field display: https://news.ycombinator.com/item?id=8109515

What are the issues with applying this to a video camera instead? That seems like a more interesting use case.

IIRC, Ng told me the team is working on it but the amount of storage and processing necessary is pretty massive.

On a similar topic, does anyone watch 3D television?

What? No. Cameras are complicated enough.

Reply on news.ycombinator.com