Egami spotted the next patent showing a Full Frame lens for the mirrorless NEX or the fixed lens RX system. It is a 35mm f/1.8 lens that has one big characteristic…it has been designed to be ultra compact! A perfect street photography lens! I want that 🙂
And there is one more particular thing to see on that patent: The lens has been designed for a curved sensor! The main advantage is that the sharpness through the full frame is increased.
via PR.
59 Responses
Curved sensors are the future for small full sensor mirrorless cameras
…as long as this sort of future does not include curved image prints that’s fine with me 😀
No more lens or chromatic aberrations at the edges
drool
The curved sensor sounds interesting, but I wonder how it will work with adapted lenses. Will any adapters for M-mount, A-mount, etc. need corrective optics? I assume for APS-C E-mount lenses they can do distortion correction via software in the camera, but for everything else… this might not be the dream-come-true mirrorless camera people were hoping for.
It’ll be progress, not backwards-compatibility.
This is not an endorsement nor a criticism.
It is not distortion that is corrected by a curved sensor (or plate) but curvature of field – the corners and centre are focussed at different distances. This seems to be the cause of the soft corners on the 16mm lens, for instance. It cannot be corrected in software, except by taking several shots and using focus stacking.
If a curved sensor is used on an interchangeable lens camera, all lenses designed for use with a flat sensor will have blurry corners, at least at wide apertures.
So this idea is best suited to a fixed lens camera such as the RX1.
>This seems to be the cause of the soft corners on the 16mm lens, for instance.
It is not. If you pick focus point in corner, corners are still blurry.
This is not for a interchanagble lens camera… for the RX1 successor
Im loving Sony, finally Sony gonna bring sweet cherry to pop, im just amazed, im moving from alpha to nex, that system is so sweet.
ts fixed lens tec. not bad tho
Like in real eye. Great.
Technology never end. But our photograph still the same. (^ ^)
That’s interesting. That’s really interesting actually. Hope I will have enough room to decide whether should I sell my Summilux 35 ASPH FLE.
BTW I love this Sony: Sony that make innovations just as much as I hate Sony that set premium prices for mediocre stuff like Sonnar 24/1.8
just god damn make FF NEX already. stop it with these fixed lens camera.
I have heard curved sensor several times but is there any real life product or equipments that use such kind of design?
Lieca use a curved sensor
No, Leica uses offset microlenses on a flat sensor.
The Ross lens made in 1841 for Calotype photography required the paper negative to be curved in this way, to correct for field curvature.
I don’t see how Sony can patent an idea which has been around for at least 170 years. They might patent a method of making a curved digital sensor. I’ve never heard of a curved silicon wafer.
They are patenting the lens not the sensor. Regardless the patent is for the specific design, why else would companies issue a new patent for every lens design.
I am always surprised that patent offices allow patents for so many lens designs which incorporate no new ideas and are just trivially different.
Design is not invention.
There is history here: AGFA Clack amongst others
35mm F1.4 for RX1 would have been just… perfect !
These patents are filed periodically by various makers. Concepts have been worked on in research labs. Somebody said maybe 2016 on the counter top. Sony would make a big splash if they did it before then.
Worth watching.
Thanks SAR.
Sony pushing forward … thats nice ! …. the RX systems are going to be remember like killer game changers cameras , maybe they are not cheap … but to be honest I like them a lot
if this lens its make for the RX1 successor its has to be something really special
Yep, when Sony needs to replace the fantastic RX1 in a year or two, they sure will need something special to really make a bang. An improved faster lens and a curved sensor should do the trick!
I can’t see curved sensors making sense on interchangable lens cameras, but if they can make it work financially on a fixed lens camera, then I’m all for it.
If they produce 3 small primes optimized for that curved sensor it maybe makes sense..
Yep. Totally agreed with Carl.
Curved sensor is a brilliant solution for compacts (even compacts of the RX1 class), but rather poor solution for ILCs.
Why? Because ILC with curved sensor would be incompatible with any existing lens, and the curvature would significantly limit the focal lengths. So it’s better to create brilliant compact than poor ILC system (note word system: as far as individual camera with certain lens might be great – system is a whole is rather pointless).
>Because ILC with curved sensor would be incompatible with any existing lens,
Who will need large lenses made for flat sensor, after small and high IQ lenses for curved sensors will be available? Only retrophyles.
>the curvature would significantly limit the focal lengths.
It cannot (unless you want to have lens with single element). Optical designer can deliberately introduce any amount of field curvature to match sensor (essentially same which they already do for flat sensors).
Anyways, many of telescopes (simple Maksutovs, etc) have field curvature too.
It does not limit focal length in any way. Why do you think it would?
At most, advantage with longer focal length would be not as big as with wideangles.
The lens design shown looks very similar to the fabulous and inexpensive Sigma DN 30/2.8.
Hope the Sony stuff will also be fabulous and very inexpensive.
At the bottom of the egami-page it lists a 10-25mm f/2.8-5.6 (1inch mirror-less). In Sony-terms that would be related to something like rx100, but 10.79-24.66 with 2.7 crop equals 29 – 66.5 mm. Rx100 had 10.4-37.1 f/1.8-4.9. If they are planning a 1-inch camera with less zoom and weaker light specs, are they going to make a camera which makes rx100 look big?
The FF NEX is cool but what about the 7n? Any rumors regarding its release?
Obviously nothing to do with FF NEX because current A/E-Mount lenses won’t work.
not good news for adapter crowd…
Curved sensors sounds great. About time they started copying what billions of years of evolution in our own eyes has proved to be the best. I wonder if they’ll ever go the full way and create two round sensors!
Interesting. So, this looks like the replacement combo of the RX1 coming up in the future….maybe.
But why a curved sensor? surely it may be more expensive than the normal flat alternative. The lens in the RX1 is already so good that the performance from centre to corner cannot be matched by any other lens on any other camera of its size or close (except Sigma 35mm 1.4 and a big SLR maybe)
Giving us 1/3 of a stop more of light is nice but, will it be as sharp as the current f2 version today wide open too?
However, I believe the design for a curved sensor is actually solving the problem of the NEX 7…very small pixels. So maybe, this sensor will be a 30+ or 40+mp sensor that will be optimized from centre to corner sharpness ala RX1 with the new 35mm lens.
I would prefer, however, an f2 that is as sharp as today’s sonnar instead of a 1.8 that isnt.
Great idea for compact fixed lens camera.
Problematic for interchangeable lensrs.
Why problematic? After autofocus mounts came out, Canon and Minolta users switched to completely new lens lineup for much less reason…
This, actually will be good for lens makers because every mount will have its own field curvature, like nowadays they have their own flange distance
Each focal length eould have s different curvature of field and require a different sensor. The sensor fits the lens
The purpose of a curved sensor is to eliminate the need to correct distorions by having a deliberately distorted sensor.
This allows a very simple(4 elements ) inexpensive lens with excellent optical quality.
It also eliminates the need for exotic glass or aspherical elements to achieve apochromatic quality. This again reduces the lens cost.
Correcting a different focal length lens to match a specific sensor curve would require all the stuff we eliminated to come back.
This system works with the SENSOR matching the lens otherwise it is not practical.
If you want interchangeability the way to do that would be to exchange lens/sensor sets. Not just the lens.
I am sorry, you’re speaking about things you do not fully understand.
Apochromatic correction needs glasses with different Abbe numbers, is has nothing to do with sensor curvature. Sensor curvature has nothing to do with distortion, too (distortion can be ‘fixed’ by making rows and columns or sensor curved, that is, regardless of sensor curvature itself; but distortion is corrected in software nowadays).
It is true than each focal length has the optimal sensor curvature.
But, if, for example, sensor curvature is optimized for 35 mm FOV, then making 24 mm and 50 mm FOV would be slightly easier too. And because longer focal lenses typically have long back focus too, and this designed doesn’t have focal-plane shutter, it allows easy placement of Smith’s corrector plate (which is inconvient in wideangles).
WRONG!! Schnieder was making 4the element symmetrical apochromatic lenses that I myself used over 50 years ago. Long before exotic glass was invented.
A positive then negative set produce opposite chromatitic abberations that partly cancel each other resulting in a achromatic pair.
The light then passes through a diaphram an on the other side is the exact reverse achromatic set.
The result when used at 1/1 magnification is a perfectly flat field whith no chromafic abberation.
Thats a TRUE apochromatic lens and they have been used in almost every graphic arts copy camera ever made.
When we greatly enlarge or reduce the image the symetrical configuration does not work as good.
So for camera lenses manufacturers made asymmetrical versions with the focal lengths being different in front of and behind the diaphram.
Now of course the corrections become more complicated and include perspective and field abberations.
Continued.
The asymetric desighn introduces new abberations including perspective snd curvature of focal plane abberations.
If we could ignore these than a simple 4example element glass lens could easily be configured to be apochromatic.
The introduction of a curved sensor allows us to compensate for perspective snd field abbreviations on the sensor snd we can therfore ignore them in our lens desighn. Now by correctly mismatching the front and rear achromatic sets to fit the magnification range we choose to use wr can easily make s simple 4 element glass apochromatic lens that would otherwise not be possible.
Thats what this desighn does and thats why the sensor must match the lens rather than the other way around
That is also why you were mistaken when you said I dont understand ehat I am talking about.
This is not a new lens desighn.
This 4 element desighn is actually one of the oldest, most popular and successful ever.
Symmetrical apochromatic flat field versions were used on almost every copy camera ever made.
Asymmetric versions were used for most press and view cameras long before slrs or rangefinders existed.
This lens will be apochromatic wihout needing special glass or assymetric elements as long as the elements are properly matched to the magnification ratio.
It will be an inexpensive and very high quality lens.
I dont quite know how it can be patented again. The old patent expired long ago.
Being a desghn patent however only protects sony from an EXACT COPY IN EVERY DETAIL and not from any SIMILAR lens.
It’s going to be the new full frame xperia phone.
The point of the Sony spherical image plane is to depict a spherical object plane. In other words the focus is on subjects with the same distance from the optical center of the lens. The result is a much more natural lens design, simpler, smaller and less costly.
Another advantage for you as a photographer is that when you focus and re-frame, the originally focused object stays in focus, which is not the case with current flat object to flat image lenses.
>Another advantage for you as a photographer is that when you focus and re-frame, the originally focused object stays in focus, which is not the case with current flat object to flat image lenses.
MISCONCEPTION
Even the focal plane in image space is curved, this does not mean the focal plane in object space will be curved too
Note that there are cameras that handle this problem in software (tracking the object)
Of course not. But I am telling you that is the purpose behind the patent.
>Another advantage for you as a photographer is that when you focus and re-frame, the originally focused object stays in focus, which is not the case with current flat object to flat image lenses.
You can have it now with flat sensor, using Minolta 28/2.8 (IIRC) variable field curvature lens.
You can buy that effect on several cameras, including the Hasselblad with its patented “True Focus”. The advantage of the Sony design is that you get simpler, smaller and cheaper lenses plus “True Focus” for free.
A patent’s existence does not always lead to manufacturing…
Please Sony, could you please just make a compact 35mm equiv (24mm) so that I can buy a NEX-6!
I think they will not. If they do, who would buy hugey and expensive 24/1.8 Zeiss?
could be that from 2020 all lenses will had a curved sensor built in?
surely it will make wonders for exposure and autofocus.
so, the body will only be a box for grip with the battery, screen, chip… wait… nah… the mount should had like 100 contact pins.
but dreaming that every lens is perfectly matched to a crafted sensor to give us best image quality, its a good dream… isnt?
maybe for ff and aps-c cost will add up (dont know if the savings from expensive exotic glasses and fewer elements will equal the cost of the special built sensor) but for 1″ sensor i keep my dream alive.
…. and this would mean that you would only be able to use proprietary lenses and not adapt existing glass… No thanks!