inohuri: I see several comments concerned about sensor size.
If they have the dynamic range and variable DOF field they claim it wouldn't matter except for which lenses will work.
If it is large lenses will need to have a wide enough projection out the back. If the sensor is small, a wide enough field of view out the front.
If the sensor really is enormous they would have a hard time getting deep DOF through a lens and still have enough light.
Or could they be using several lenses???? Or no lens??????????? Or just the micro lenses on the sensor(s)?
The Lytro stills camera disappointed me because it would not output a very deep depth of field, just an equivalent of what would be baked into a normal image. That it needed to focus was the biggest hint it was not what I wanted.
Oops. Lytro sent me an email. I went to their website. I found that I did not have a clue regarding the Illum. I must have dismissed it because of the earlier camera's weakness.It has a lens and does shift perspective and the examples show f16.So my guesses are far more probably wrong.Sorry.Oh oh. I might want one. I took a look and prices start at around $400 new. Must check warranty.
How about a wide collection of sensors possibly on a curved panel, no central lens. This could provide the shift in view.
Smart algos may have arrived, at least way far up market.
For those not familiar with this way of reducing noise think of high quality HDR but all exposures the same. The random noise averages out.
Did currently just become yesterday?
The term Light Field says a lot. So does the shifts in view. I'm betting wide area sensor made up of several smaller sensors. That could give a fast per sub-sensor read out. It could also build dynamic range by reducing noise. No single central lens, just light gathered at each sensel which is then calculated into a matrix. No jelly effects either.
Thanks for the Illum output data. I had not seen a shot processed that way nor had I seen it promoted. I don't have files to process so no reason to have Lytro Desktop to discover it would do that. I did subscribe to the emails and look at what was on their website. All of what I saw had fixed DOF, variable focus. So maybe their promo isn't that great.
I see several comments concerned about sensor size.
halfwaythere: The penultimate shot, the one with the dog, is quite horrendous image quality wise: very noisy to the point of asking myself is this was shot at ISO 3200 or higher and the CA in the tree is quite "spectacular".
I see the appeal of a camera like this but I don't understand the hype.
I looked for tutorials on Lightroom CA correction and they just desaturate color. I do that with Hue/Saturation.It is a global correction or requires selection. I selected the upper part above the grass and when the cyan CA was desaturated the color in the trees was spotty or gone. Too blue and the sky posterized. There was also magenta which might have been easy to fix.Check this video to see this problem. At 1:38 the purple stuff (not CA) at the top of the screen is complete. After editing it is spotty. This is why I compare before and after to see if anything I don't want changed winks in and out.
Lateral CA correction just makes this image worse.I did check the samples for the camera and the one similar shot had purple fringing which usually is easy to desaturate.
What is the method please? I would love to know.
Have you downloaded and corrected this?
I don't mind correctable lens faults. I'm used to CA and fringing.
I downloaded and tried to correct the dog shot with my usual tools and could not.
To all of you who say Canon must build more EF-M lenses:
They did. A long time ago. You use an adapter.
Seriously if you want to make honest comparisons compare to a similar sensor size like the APSC Sony. Little sensors allow little lenses.
On a large, really large, lens I would be using live view any way and a mirror really gets in the way.
My biggest is 3Kg. Roger has bigger.
"EOS-M First ImpressionsRoger Cicala"My favorite part of today's tests, though, weren't the lenses. It was the EF-EOS-M adapter. No one has done adapters better. No one."
ADSinger: The history of photography is a study in miniaturization. Cameras and negative/sensor size have been shrinking for two hundred years. Every new technological advance is met with derision by the supporters of the status quo. We saw this within the past couple of years as smart phones decimated the P&S market. The DSLR users view is no different than that of the Rollieflex owners faced with the 35mm Leica. MILC cameras will supplant DSLRs for all but professionals with specific needs. Canikon are faced with adapting or becoming niche products, i.e. Hasselblad, Leica M, etc.
In other words a 1D will do a better job than a gopro on a drone.
Sounds like we should all be shooting 8x10.
I have been wanting to build a proper tilt/shift rig and the flange focal distance for EF is just too big. This was pushing me to Sony which for me isn't really viable. M4/3 sensor is too small and has thicker glass over the sensor.I bought the camera and sure enough I can now use 35mm SLR lenses for tilt/shift. The old 20mm Nikkor will give me 32mm equiv. That 72mm filter size looks odd on the tiny body. Don't know how much tilt yet but on EF with enlarger lens I got 30 degrees through old film camera finders. The adventurous might peek at ebay. Canon brand EF to EF-M adapters run around $50, third party cheaper. There is a tilt/shift adapter but the few reviews knock the quality and it's $300. Adapters to Nikon, Pentax etc. Even Minolta and FD adapters without a lens to reach infinity.Magic / Tragic Lantern is available.
"Who owns the content that users create?"This deserves an answer. Where's the contract?
What's in this for me? I'm not greedy, I just want a fair exchange for hard work well done.
Can I copyright my work? Can I withdraw the article whenever I choose to?
I could write articles for Amazon for free that would bring me what?
skytripper: In the comparisons between jpeg and raw, the raw images have obviously been tweaked but it appears that the jpeg images have not been. A much more meaningful comparison would be between tweaked jpeg and tweaked raw. While it's true that one can get more out of a raw image than a jpeg, jpeg's can most certainly be substantially improved with minimal tweaking.
The one other subject is saving the jpeg after editing. I now save at the highest resolution. I know that I could probably not see a difference higher than 10 (Adobe) but what can the printer see?Also I may want to do a quick edit for something I missed and at 12 I should lose less opening and saving. Storage is cheap. The edited file may be bigger than the original but that is fine with me. For cameras under 8 megapixels (love the Fuji S6000 shot in raw at ISO 100) I double the resolution right after opening from Camera Raw (x1.5 in dimensions) as they edit a little better. Fuji's software for Super CCD did that too.
Many of my cameras will not shoot in raw or shoot too slowly in raw for some situations. I edit the jpegs in Camera Raw and they seem to me to get better. Sometimes much better.
Must I do what others do?
"I said the file size makes no difference, and that does not refer to the compression being used."
I don't understand this.
I apologize. I didn't think to click on the jpeg and when I did the large file came out, probably the original, I didn't check.
I opened it in Camera Raw and got nowhere with the White Balance Tool on his shirt. Then I saw a bit of white in the upper right and one click did what I thought could be done. Better but not as good as the raw as usual.
Was the camera set to AWB? If it estimated the shirt to be white this could have thrown it off. The exif seems to have been stripped (which is what I would do, my name and phone are in there for security).
Also editing a jpeg in 16 bit will do a better job. Try it, it shouldn't hurt. Adobe doesn't know best, they just think they do.
I use Elements because I can afford it and know the workarounds. That means I open in Camera Raw even if I do no changes there just to force 16 bit editing. I would probably go minus one or two in Clarity because it looks better to me when at 100 per cent or higher.
" The size of the JPEG reveals absolutely zero about how much it can be tweaked." As it makes no difference I should reset all my cameras to the highest compression / smallest file size.
Didn't I say "hint"? Of course this is not an absolute. With similar scenes at the same ISO I believe I am correct. You can go shoot black walls all you want. The quality of compression also does indeed vary from camera to camera. For reasons of my own I almost always set the camera to the largest file size.
"A RAW file can be processed to look exactly like the JPEG." Uh, isn't that turned around bacKwards from what was intended? We want the best tweaked jpeg compared to the best tweaked RAW.
I often compare the full size least compressed jpeg to the raw when the camera will shoot raw + jpeg. If the jpeg is better I'm doing something wrong (again). I open the jpeg in DPP and compare side by side with what I (am attempting to) edit in Camera Raw.
AbrasiveReducer: Interesting that Nikon and Canon are going to battle it out to see who can make the best camera with a sensor that's too small. Although I didn't care for the Sony, assuming the G16 is the same price, I'd definitely take the Sony.
Please Canon, give us an updated G1X. The image quality is already there; it just needs the computer stuff.
Sensor that is too small for what?This is my favorite size, small enough for deep depth of field, just big enough for quality. My A640 has too often outdone my Xsi/450D with prime Nikkor or Canon lenses especially if shot in dng raw with CHDK.
The point was tweaking the jpeg for a more level comparison, not how much you can tweak.I find in practical application, especially with Camera Raw 7, jpegs can be tweaked quite a bit. The file size is a hint as to how far you can push.File:Open As:Camera Raw for those who don't know. PS Elements 11 Camera Raw will bring up shadows just the same as PhotoShop or Lightroom as far as I know. Be sure to open in 16 bits, the latest Camera Raw 8 setting looks like a web link, I didn't want to go on the web so didn't click there until I searched the internet. It also sets to 8 bit by default. Another "What were they thinking?" The setting is at the bottom center left.
Why F2.8? The four photos I looked at were shot at large apertures. My experience with sensors of this size is that they are at their best around F4 or F4.5. I want to see the best the camera will do. I stayed with the A640 from 2006CE because the G series shots I have seen (except the G15 studio shot) seemed only slightly better to me. Is this the better camera? I can't tell here. I know the A640 will do better than this.I live in Seattle, I don't have image stabilization and I routinely shoot around F4.5 at ISO 80.