Sony is heavily investing in a brand new sensor tech that will likely find his way in production cameras as early as 2015. I already told you a couple of weeks ago that Sony developed a 54 Megapixel sensor with on sensor PDAF. Well today I have been told by a very trusted sources that this sensor is also the first that has not been designed with a classic RGB (Bayer) pixel structure.
The well trusted source told me that Sony’s plan is to keep using 24MP and 36MP FF sensor on the new A and E-mount cameras that will be announced in 2014 and early 2015. It will also depend from the competition whether Sony will launch the camera with the new sensor in late 2015 or during 2016.
In short, after many years of classic RGB sensor structure (with the exception from Sigmas Foveon tech) we will get a real major leap forward in sensor imaging tech.
I hope to get details soon!
—
Message to sources: You can send me more info anonymously by using the contact form you see on the very right of your browser. You can also contact me directly at sonyalpharumors@gmail.com.
For readers: Subscribe the RSS feed (Click here) and Twitter (Click here) stream and the Facebook fan page (Click here) in order to not miss any news!
—
Reminder (SR = Sonyrumor):
SR1=probably fake rumor
SR2=rumor from unknown sources
SR3=50% chance it is correct
SR4=rumor from known sources
SR5=almost certainly correct!
156 Responses
How will that be compatible with RGB standards in software programs? Anyways, great news and let’s hope colors will be as real as possible.
My guess is that Sony will be selling these new sensors to Nikon and Co. So there will be quite alot of cameras out there with this new technology. Thus, the software developers will have to get this tech supported fast, or they will loose their clients.
Why should they do that? After the A7r the game has changed. Sony is on the hunt. They want the whole cake.
Sony is the industry leader in sensor sales.
Sensors are MUCH MORE IMPORTANT to sony than cameras.
Today, but what about tomorrow? Will they still again wait a year to get their own cameras out? Not sure if Sony can survive like that. They have some huge financial problems, and to me it looks like the photography market is just what they need. They make great sensors, have tremendous computing knowledge and they have always been very innovativ. I doubt they put all the effort into building a A7r , via NEX 7 and Rx1, just to sell sensors to Nikon.
…because Sony build Nikon tech. Sony don’t just sell Nikon sensors, Nikon license Sony their sensor technology (in use on non-sony sensors with amazing performance like the D4/Df chip – with the best noise performance of any sensor ever made).
That’s why Nikon can demand to get first use. It’s not a choice for Sony and it’s why not every Sony chip performs as well as the ones they share with Nikon. It’s also why those chips aren’t sold to Pentax and others.
Nonsense from start to finish. Sony Exmor technology was first used by Sony themselves. The D800 sensor is the Sony 16MP Exmor sensor scaled to FF, nothing else. And that 16MP sensor was first found in the A55 and A580.
And talking about that A580, it performed just as well in terms of noise as the D7000 which used that sensor later, the latter only had a tad more DR because Nikon opted to implement 14 bit RAW when Sony (openly) felt it wasn’t adding enough yet.
All other differences are down to SLT, on sensor PDAF and CFA differences.
So is this rumored sensor a Nikon technology?
Highly dubious, Mr Skelter. However, there is no question that Sony and Nikon work together in an elaborate pas de deux. Nikon makes the machines that Sony uses to make sensors. Designing them is either an entirely Sony function, or, in some cases perhaps, a collaboration. After all, the common enemy is the dominant market leader, Canon.
Sources please
D3, D3s and D4 sensor is a “Nikon sensor” produced by Renesas
All other sensors are made by Sony and now also by Toshiba in the APS cameras
“After the A7r the game has changed.” BHAUAUAHAHAHA for nerds maybe, only playstation can change the sony camera market lol
Convert it to jpg, tiff or dng. Those are universal formats. If you’re referring to raw files, it’s Sony that has to provide their raw specs to software makers out there.
I just hope that it will not have the sometimes terrible artifacting of the X-Trans.
it’s the end of 2013 – artefacts, although still can be found in ACR, are not problem – CaptureOne all the way 😉
Agreed, even the much touted jpeg engine still shows them (straight lines becoming wavy, some weird edge haloing from sharpening that I’ve never seen before), see for example the XF23mm samples posted on Dpreview yesterday. High ISO looks good regarding noise, low/mid ISO does not, regarding artifact free detail.
Nice to know that some converters do better than others, ACR/Lightroom is still the most used one out there, by far. And for more than 1 reason.
There’s no point to look at DPreview’s comparisons when it comes to X-Trans – they use ACR for RAW
I was talking about the jpeg samples posted from the new 23mm lens, in the context of the much praised Fuji jpeg engine.
As for ACR studio samples, see my earlier comment regarding ACR/Lightroom.
+1 Agree
The X-trans is fine for the most part, the only thing I don’t like are some smudging of fine details that still goes on. I just hope this new Sony sensor won’t be the useless effort that X-trans was, because it ended up being not better than a regular sensor, yet it required users to jump through hoops for a year. And it still isn’t there.
Will it be the RGBCMY Pattern Sensor?
http://egami.blog.so-net.ne.jp/2013-06-25
http://www.sonyalpharumors.com/sony-patent-discloses-hexagon-and-triangle-pixel-design/
Or it may be the RGBW Pattern Sensor.
which can improve the high ISO ability.
http://egami.blog.so-net.ne.jp/2011-04-06
RGBW got an issue with colors. You can get similar effects by increasing transparency of CFA. That’s one of the reasons why Sony A-mount cameras got better colors than Nikon F-mount: Sony is using stronger CFA for better colors separation. I know that recent crayziness about high ISOs is what ppl like to talk about, but…. if I would want that – I would switch to Nikon long ago. One of the reasons why I’m shooting Sony is because of it’s colors. So RGBW sensor would be a very underwhelming evolution satisfying only a certain group of customers.
(sorry, that was a replay to lhkjacky)
(replay function doesn’t work >_<)
On a forum, indeed replay does not work. You need video for that.
😉
Is that so? I tot we can have the best from both worlds – thicker CFA for better color, white pixel to compensate light loss.
Yes, I love Sony colors too, and hope the new sensors preserve them. A quick example: http://www.modelmayhem.com/portfolio/pic/34112728
RGBW (or any CFA with more than 3 types of filters) could in theory have color accuracy better than Bayer because of extra degree of freedom (but only if demosaic supports it)
at low exposures, pixel readout noise becomes significant, so it’s where RGBW is NOT equvivalent to making less thick color filters.
There’s no extra degree of freedom. In Bayer sensor 100% of pixels got an information about color. In RGBW only 75% of pixels got an information about color. It’s simple as that. You can compensate quite a bit of that loss with software, but… yea… we all know how customers love companies baking RAWs.
in rgbw, “W” pixel has no color information if, and only if, its spectral sensitivity is a linear composition of R,G,B (which won’t apply)
Will they give us also a 1TB hard disk for free, when buying the camera? 🙂
I doubt 1TB spinning platter hard drives will be available beyond 2015, minimum capacity likely to be 2TB or greater.
Even better: that crazy race to highest MP possible is forcing us to upgrade computers, hard disks, SD cards. Only my wallet is downgrading.. 😉
While for post processing large numbers of files a speedy PC helps { i have found that SSD have made the most notable advance in system speed in recent years]. I can easily PP files from my D800 on an old Dell laptop { that is powered by coal :-)].
For event shooting where there will be high volumes I typically shoot JPEG+RAW only using RAW for files that have particularly difficult lighting.For applications that require the most meticulous RAW processing are more often much lower in volume landscape etc
Storage gets cheaper and larger all the time.I have found that higher MP sensors [ i have the D800 are excellent at low ISO and with the extra detail you have more flexibility with NR at high ISO.The D800 delivers results that are at least as good as my D3s at same output size and detail level.
Is it a USB2 SSD drive that you’re using?
FF camera + lenses + batt + tripods + stuff = $9000
1 TB drive = $200
More like $80. $150-$200 for 4TB
like 🙂
@boooe does it mean you will buy one for me? 🙂
so, you’d be able to afford this super camera with high mega pixels but can’t afford storage? lol… you’re like those idiots that buy high performance sports cars and then whine about the high price of (high octane) gas. If you can’t afford $150 for storage, you shouldn’t be buying high end cameras with high MP
And one more for a backup.
Make it a redundant Raid 5 Array 🙂 oh wait, Since we’d be hitting 8TB easy, might as well use Raid 6 Array with dual parities.
“we will get a real major leap forward in sensor imaging tech” – we’ll see if it’s going to be a leap or not. So far non-bayer sensors got rather poor record – especially in terms of high ISO image quality.
That’s a reason why we all use bayer sensors – they’re so good that – despite of dozens experiments made since – nothing manage to beat old, good technology.
If anything – perhaps graphene sensors will really be a major leap in sensor tech. But we’re much further away from that than just 2016.
Yes, the Fuji X-Trans is really bad at high iso, and so are the RGBW sensors…
khuzul – obviously there are exceptions. But yea – X-trans is rather underwhelming comparing to Sony Exmor – only 99% of reviewers can’t see that because they compare photographs at the same ISO settings completely forgetting that Fuji got different ISOs than Sony or… well: any other manufacturer for that matter.
Look at DXO mark ISO score – There’s not even a single X-trans sensor in Top-20 cameras by ISO score.
(ok, nevermind – that was a poor example – DXO doesn’t score X-trans sensors)
very poor example indeed 😛
“Exmor” in the name has nothing to do with CFAs. “Exmor” is about pixel architecture and ADC for each line. “Exmor-R” is about backside illumination. Any of these can come with any CFAs.
Hmm? I never said anything you speak of. Just compared X-trans to best of bayer sensors.
This will NOT BE A FOVEON SENSOR it will be a very different stacked sensor.
Foveon succeeded only in stacking the colors with inherrent advantages for doing so. Everything else about the sensor was mediocre. This does not mean that other stacked sensor technologies will share foveon difficulties.
As far as resolution is concerned 54mp is very much entry level fpr this technology. 54mp stacked is the same pixel size as 13.5 mp bayer.
The stacked equivelent of 36mp is 144mp. The stacked equivelent of 24mp is 96mp.
The effective resolution of lenses will also be 4x what they would be for bayer.
Lastly.. any edge softness will be much less noticeable because adjacent pixels are all the same colr…in bayer they are differend colors so pixel edge softness causes colr shifting.
It seems to me that a 54MP stacked sensor will have 18 MegaPixels. Bayer would need 72MP to get 18M sites, but 54MP will mean 18M x 3 colors. I would hope for more in 2015. Lenses now are starting to support what Sony would call a 108MP stacked sensor (that is 36M real pixels each with R, G, and B.
You are correct but you misunderstood my statement.
I was comparing pixel sizes, Since a bayer array combines 4 pixels to make one rggb
Your statement is correct in that a 18mp sensor would be called 54mp by sigma,
In fact looking at my response elsewhere to Christopher I did make that exact mistake in that post.
I NEVER SAID IT WILL BE FOVEON SENSOR nor that it will be a stacked sensor. It might be stacked – but I find it rather unlikely. More likely they experiment with different CFAs or different materials / manufacturing processes.
My guess would be either stacked 3layer orgsnic/silicon hybred or
Stacked 2 layer.y c top layer in checkerboard patter and all green second layer also combining silicon and organic components.
This is a guess however as I do not know which desighn they have chosen.
I hope it is not the 6 color hexagonal pattern because I do not believe that desighn is practical for higher resolutions.
Comparing photosite sizes between bayer and 3xlayer (“foveon”) sensors doesn’t make the same sort of sense you get when comparing two bayer sensors of different sizes… because of sensitivity and resolution differences.
Also I’m guessing you’re unaware of the colour shift issues that Foveon sensors are having in the corners, likely as a result of having deep photosites?
This will NOT be a ” Fovion” sensor
Sony will have a new and differrent sensor. Sony is aeare of all foveon shortcomings and I am sure desighners are taking them all into consideration.
Microlense desighn can compensate for many problems. In addition to offseting the can angle different areas similar to a fresnel lens.
Also it is possible to offset layers
Super!
And we will first see it in a Nikon Camera…..
I think it is smart to put the new sensor in a Nikon body first.
So they can find out the bugs and make a better sensor for their own.
Also,
If they put the sensor to their own body first, you won’t get any raw support from LR or ACR,
because Sony’s market is too small.
But it will be totally different story if they put the new sensor to Nikon body first,
it will get LR, ACR support immediately.
Yeah, as anyone would like to use brand new high-technology image Sensors with Adobes Outdated LR/ACR Technology 😀
SOOO… I’m fine then with my A7 as for at least a year or two, no better sensors from SONY will emerge :> No groundbreaking progress or major steps in sight. So I’m good for now ;>
Yes you have at least over a year until Sony remotely transmits the self-destruct signal to your a7.
So are we talking about new sensor technology like Foveon or new CFA (colour filter array)?
That is exactly what is not clear in this rumor.
Maybe Andreas can find out more.
In the meantime Sony’s current sensors are really very good.
Probably B&W sensor with sequential on-lens color filter. That will be something. Great for landscapers, useless for sporters.
If the rumor is true, it would almost certainly be a different CFA arrangement, not a layered sensor. All the layer methods currently being explored other than perhaps Nikon’s would have high ISO issues, especially with the small photosite sizes implied by 54mp.
Kodak themselves–indeed, I seem to recall it was Dr. Bayer–said that alternative CFA arrangements really only start to make sense at high pixel densities. Kodak explored a whole host of these alternative layouts, which tend to be laid out in 8×8 pixel blocks.
54mp is 9000 x 6000 pixels. Divide by 8 and you get 1125 x 750, so I’d think that there will be some video subsampling issues with an 8×8 block to get back to 1080P, though the RX10 is doing downsampling, so maybe Sony has a solution for this.
In theory with CPU powerful enough – you can do perfect downscaling to 1080p. So far noone does it – everyone are skipping lines – that’s why stills from DSLR videos got such a sub-pair image quality to even worst of JPGs (and all of the issues with aliasing and moire).
You don’t really need different CFA for brilliant video image quality – just really, really great bandwidth and CPU.
But yea – Kodak did tons of experiments with CFAs and all of them got issues that made them inferior to Bayer sensors. Perhaps they did solve it with high MPx camera. Perhaps they decided to go for supersampling like Nokia Pure View? Or perhaps they got some other trick – we’ll see 🙂 Admin mentioned more rumors about this to come, so… let’s wait and see. Still though I’ll remain skeptical till detailed reviews are out.
Quote: “The well trusted source told me that Sony’s plan is to keep using 24MP and 36MP FF sensor on the new A and E-mount cameras that will be announced in 2014 and early 2015”
And the APS-C cameras? Or won’t we be seeing any more of those…
Sony A mount and E mount come in both APS and FF forms from Sony. The A mount in particular is easily adaptable to either frame size. E mount took a little fiddling because Sony initially designed it for only APS.
Fiddling how? The mount was designed with the option of FF in mind. Usual fact free Walt talk.
The e mount could have been desighned with a larger diameter like the a mount.
This would not have increased the size of the camera and would have allowed for the future introtuction of ibis if sony decided to do so at a later time.
This was indeed poor planning by sony therefore we have to contend with the frustating fact that e mount is compatable with canikon but not with sony a mount.
Current sony customers would have added to the potential sales base for e mount.
As a result of this blunder sony must now produce a hybred a/e mount strictly to remedy thier inability to properly plan ahead.
P.S. WK will not answer our responses to his comment…he has never done so since I have been reading sar.
Admin
Sony patented 3 new sensoe tech
Which will it be
3 layer stacked organic/silicon
2 layer ycg
Or 6 colr hex…this one seems least likely
I doubt the benefits. May be measurable — but noticable? The main point is probably to have new sales arguments, no matter how insignificant they may be. However, definitely significant will be the trouble for all the photo editors to support the upcoming chaos of pixel alignments. And it certainly will not become faster.
If they can get 17-stop of DR, that would be something in sensor development.
If.
A while back i made the following projection/ prediction which now seems even more likely
2014 will see numerous camera inovations but only minor sensor improvements no more than 24mp c and 36 mp ff.
Sony will not do much more to develope bayer sensors and what they do develope..like gapless microlense.. will be carried into next technology.
Consumers will demonstrate thier prefferences for body types.
In 2015 the next technology sensor will come out and sony will have body styles ready and cansumer proven to accecp the new sensors.
I was of course projecting and hoping this would happen in 2015 .
This post seems to indicate that all thes predictions are coming true but are happening more slowly than I wanted or expected.
OK
Now give me 6 numbers between 1 and 49
:-/
1,2,3,4,5,6
Just as bad as Tom Whogan at predicting the future.
😉
You asked for six numbers i gave you six numbers. I did not predict anything.
Lower Prices for new cameras until the introduction of the new tech sensor.
As is typical of most high tech devices as we near the end of one technolgy..bayer..and the beginning of the next prices drop and we are buying a mature technolgy. This will result in 2014 cameras being a relative bargain.
Of course there will be a huge tmptation to replace them with the next generation when available.
Given that the a99 & rx-1 were announced October or November of 2012, it makes sense that the next big jump is Oct/ Nov 2015. These things seem to be in three year cycles. Just pray that Nikon doesn’t get a year of exclusivity (or if they do, hopefully they will get the sensors in 2014).
“A while back i made the following projection/ prediction which now seems even more likely
2014 will see numerous camera inovations but only minor sensor improvements no more than 24mp c and 36 mp ff.”
With 3 layer sensor we have a hope to get rid of color shift problems (with some math applied, sure). Plus a possibility to decrease light loss.
haha what makes you think so? Foveon x3 sensors are more prone to colorshift 😀
Solution for getting rid of colorshift is BSI.
These will definitely NOT BE FOVEON SENSORS
They may be stacked…probably….but most assurdly not be foveon.
Any performance assumptions based on foveon performance..or lack of performance….are completely baseless.
It’s the high-resolution that creates the color shift in the A7r or NEX7. Conversely the A7 is fine.
Color shift results in the sensor from light from a pixel leaking into the pixel next to it.
Since thes adjacent pixels in a bayer sensor are different colors even a reletively small light leak can result in a significant color shift.
In a stacked sensor all the different colors are in the same layer of every pixel. Any light leaking to an adjacent pixel will strike the same colr. Therefore theoretically color shift in a stacked senser will be greatly reduced. Also the sensor will visually also appear to maintain higher resolution because stray light will only slightly soften contrast rather than produce highly noticeable color shifting.
>In a stacked sensor all the different colors are in the same layer of every pixel.
I cited foveon x3 as an example which is like that, but it still vulnerable to colorshift.
It will not work the way you describe! There still will be pixel vignetting. And because pixel well becomes deeper, its deepest layers are more prone to vignetting.
Some people are suggesting difficulty because file type/ pixel arrangement may be differrent.
This is highly unlikely.
Probably a new raw format will be necessary and therefore updated conversion software.
Tiffs will still be tiffs and jpgs will still be jps..the basic fileformats are not going to be different.
Unless the hex desighn is used..which seems unlikely to me… the arrangement of pixels will be the same except there will not be the missalighnment of clors inherrent in bayer.
Yes, but first, you must get to the TIFFs. 😛 For example, I use Darktable for editing photos. Although this software is great and the programmers really diligent, they don’t even support X-Trans because of the difficulties.
If you’ll try to force ppl into using weird 3rd party software to generate TIFFs – you will fail. Professional don’t have time nor money to learn new software, and amateur photographers often don’t have a skill nor will for that (and Sony got long history of rather poor software on PCs). Either they’ll make Adobe, DXO and others support their new format, or they should prepare to a very slow rate of adoption.
FYI: “Bayer RGB pattern” is not a physical design of the sensor, it is tasks dedicated to each pixel by the code (programming). Take f.i. Fuji X-trans, it is a Sony 16 mpix sensor. On the X-A1 the same sensor is used but with a Bayer pattern read-out.
It is difficult to comment on the 54mpix sensor read-out before you know what it is. – Other than assume it will improve image quality.
54 mpix is 50% more than 36 mpix. That means the resolution will be visibly better.
One thing that needs to be enough: Processing power.
I sit with my D800, EM5 and X-A1 and will wait until 2015 to assess the camera system market and if it is meaningful to change. What lenses will Sony have? At what price(s). Nikon and Canon has everything, at different weight, cost and quality levels. Will Sony ever get there?
I think Sony concentrates too much on cameras and features – and too little on lenses. I buy systems, not cameras.
The sensor has a certain colour filter pattern on its top. You cannot switch the pattern just by software.
If you need to improve your workflow/technigue etc to cope with the new improved image guality then do it..problem solved.
Resolution about 22% higher.
Assuming you don’t have artefacts. If you do than a real resolution is much lower than in Bayer sensors.
So you think a 54mp stacked sensor will have lower resolution than a 36mp bayer sensor…get real…you argue for no other purpose than to argue
@ Per K
“I sit with my D800, EM5 and X-A1 and will wait until 2015 to assess the camera system market and if it is meaningful to change. What lenses will Sony have? At what price(s). Nikon and Canon has everything, at different weight, cost and quality levels. Will Sony ever get there?”
I am in much the same position though I am getting the A7r for a bit of fun , i will use it old school with some of my manual focus Nikon lenses. I also use the D800 and my mFT gear is based around the GH3 mainly for video a new interest for me.There is no pressing reason to upgrade though a wonderful machine I was a little disappointed that the E-M1 didn’t come with better image quality.
A 54 Megapixel sensor could be a three layer 18 Megapixel sensor, in the same way as the current Foveon is quoted as 48 Megapixels.
Another possibilty would be to have a two layer sensor with a two-colour filter array (Magenta and cyan). This would avoid the problem the Foveon has with its very insensitive “red” third layer, while still giving a substantial increase in resolution.
Agreed, there is much more room for improvement for the accessories, lens selection and camera software than camera bodies.
– No wireless triggers for multiport shoe
– f60 flash is in dire need of an upgrade due to overheating issue
– no tilt/shift lenses
– lenses due for updates (Sony zeiss 85 &135 deserve SSM)
– better software/firmware
– better JPEG processing
– better video encoding
…and more
– No wireless triggers for multiport shoe
I’m shooting wireless
– f60 flash is in dire need of an upgrade due to overheating issue
Only an issue at full power
– no tilt/shift lenses
Samyang has E-mount T/s etc.
– better software/firmware
New cameras and the firmware is far less buggy than you find in other competitors eg Fuji
– better JPEG processing
There’s new jpeg engine in A7/r – very easy to see the difference
– Would you send me the link the multiport triggers? I’m currently using an f20m but would like to look into the wire less multiport triggers that you are using.
– the f60 overheats when I blink at it, certainly not just at 1 to 1. It is a defective product, well chronicled by many users on sony’s own forum:
http://community.sony.com/t5/Alpha-SLT-DSLR-Cameras/HVL-F60M-Flash-Overheating-Unusable/td-p/139177
– There are many illogical choices in the current a-mount firmware/software, such as: marrying drive mode with timer mode (either you shoot continuously Or you can shoot with a timer), inability to select which card a user can save video files when shooting JPEG+raw, no PAL option on North American a mount cams (Europeans have both pal & NTSC frame rates) et al.
– The JPEGs look fine to me but I shoot raw for processing and small JPEGs simply for transmitting. Many on here hate the JPEG output but many more agree that the video codec is not good (surprising given sony’s video credentials).
I’m not maverick, but using the http://www.amazon.com/Yongnuo-Professional-Speedlight-Flashlight-Olympus/dp/B00BXA7N6A/ref=sr_1_1?s=electronics&ie=UTF8&qid=1386714191&sr=1-1&keywords=yongnuo+yn-560+iii with the Yongnuo trigger. I had to slightly mod the RF-603c trigger to work with the A99, but seeing as you’re giving up TTL for 1/8th the price of the HVL-F60, it works out pretty well. It turns out manual flash is pretty easy.
Thanks for the info 99. I sometimes use my canon 580 on the a99. The greater point is though that there are no native multi-port triggers available. Either Sony should produce these or get 3rd parties on board. It is unclear if the multiport shoe is locked down or if 3rd parties are free to produce accessories that can communicate with the camera. Thus far, there are no 3rd party accessories.
> think Sony concentrates too much on cameras and features – and too little on lenses.
Bu%%%%%t. Many of Sony lenses (e.g. SEL55210, FE70200 are of limited utility because camera/AF is not quick enough.
The E-mount cameras are feature-poor compared to, e.g. EOS DSLRS at the same price point.
Who is talking about e-mount? We’re talking about the ecosystem, which includes but is not limited to lenses as you could see if you read my post.
Maybe you should compare them to the EOS-M then 🙂
I think Sony should focus on cameras and let Zeiss focus on lenses…
Pun intended?
Sony Zeiss lenses are Sony lenses.
They are designed by Zeiss.
I initially thought that too.
– Sony designed, approved by Zeiss
– Sony manufactured, quality control parameters set by Zeiss
– Zeiss t-coatings
– Sony distributes and sets pricing
– Sony also holds the patents for the designs of the Sony Zeiss lenses
Compare the Zeiss 50mm 1.4 Planar design to the Sony Zeiss 50mm 1.4 Planar design:
Zeiss – 6 groups / 7 elements, filter thread 58mm
Sony – 5 groups / 8 elements, filter thread 72mm
Consider the Sony Zeiss line to be akin to Canon’s L lenses. Just as the L and red rim mean high build quality and best IQ for a given focal length speed, so too does the Zeiss branding and blue badge for Sony lenses. They are not pure Zeiss lenses but the build quality is Sony’s best and the color rendition is akin to the Zeiss style.
It varies on case-by-case. As far as I know most of E-mount “zeiss” lenses are in fact Sony lenses. Zeiss in RX1 on the other hand was designed in Germany. Most of Sony Zeiss A-mount lenses were also designed in Germany to Sony requirements, but the lens housing itself was fully designed and manufactured by Sony.
Per K, you’ve misunderstood how Bayer and X-Trans sensors work. The CFA is a physical filter on top of the sensor, filtering away all but one colour for each photo site. That’s why you have to do demosaicing by software, to interpolate the full RGB values for each pixel based on colour information from surrounding pixels. Take away the CFA filter, and each photo site receive all wavelengths, which basically makes the sensor colour blind. This is what Leica did with the M Monochrom.
No, he does understand. You need to read again.
Since this will be both 54 mp and new sensor technolgy…
Double the resons…this should be called the Barry Sensor.
Well Barry does sound similar to Bayer and since it might be non-Bayer it would make sense given Sony’s naming confusing schemes.
Sigma insists thier foveon is 46mp.
It is really 15mp…3 colors of 15 mp….
This is based on the 3 ccd precident…in my opinion a discracefully missleading way to falsely inflate thier claims…
I sincerly hope Sony will not follow this precident and actually call a pixel a pixel.
As a result of what sigma did we have to ask is this actually a 54 mp sensor or really 3 layers of 13.5 mp.
If this is a real 54mp sensor will Sony call it that or call it 162 mp?
My hope is Sony will more honest than Sigma and call a 54 mp sensor 54 mp.
You’re confusing pixel count with resolution. A bayer sensor doesn’t record as much information as the MP count suggests. Color is averaged which mean a lot of detail is lost (and re-synthesized). It works pretty well but when you see the Leica’s monochrome sensor and compare that to one with the same pixel count in color, or look at a Foveon’s output, it’s easy to see how much is lost.
There’s nothing dishonest about Sigma’s claim. It’s just that if you want to talk about acuity you shouldn’t be counting pixels or, if you must, can only compare like with like.
The Foveon is as sharp at 46MP as a Bayer sensor is at 30MP and has more ‘bite’. A 54MP layered sensor would produce a much sharper file than an 18MP bayer and if each layer were 54MP it would greatly exceed a MF digital back at low ISO.
Sony have patents for this kind of sensor, Nikon has better ones with less noise. If Sony build something of that type it will be interesting. I’m sure this isn’t that kind of sensor though.
If Sigma called a pixel a pixel we would not be having this debate .
It is misleading confusing and many people falsly believe it is 46mp…that makes it worse than a lie.
Please sony take the honorable..high road…do not follow sigmas bad example!!!
Bottom line..
Foveon pixels are better than bayer pixels..agreed though opinions will naturally vary as to how much better….that is also the claim sigma should be making.
Sensor will ony produce a 15 mp image unless you want 3 seperatre one color 15mp pictures. It therefore is not a 46 mp sensor and should never have been promoted as one.
I hope other manufacturers acurately describe the pixel count but indicate that they are better than bayer pixels..that would be more honest and less confusing….then make your exact same arqument and I will agree with you.
Link to relevant Nikon patents please.
How do you know the Nikon patents have “less noise” than the Sony ones, since neither has been built?
Non Bayer can mean a lot of things. X-Trans is non Bayer, Foveon is non Bayer, they are not really game changing. I doubt whether the new Sony sensor will really be game changing, or just some marketing gimmick
Sony has some nice products lately, but for A-mount, i think what is needed:
– lenses, lenses, lenses – there is too little choice and you either get rubbish cheapo lens or HQ Zeiss glass at a crazy price. The mid range is completely missing. Specialty lenses are compleetely missing.
– APS-C and FF cameras without an AA filter
– lack of long term product support. Canon is supporting their flagship models for much longer time than Sony. Their firmware updates give new life to old cameras like the 7D. Sonys firmware update for the A77 will be called A78
I suspect that Sony make little or no profit on lenses.
If they could they would provide the camera with a pinhole and software correction.
Oh come on, the prices for Sony lenses are always much higher than comparable offerings from Canon and Nikon. And those have still to incorporate optical stabilisation in lens – which Sony does not have to and can thus save cost on lens production. And I dont even have to mention that many Sony lenses are not on par with Canons lenses. If they are not making any money on lens sale, than they are doing something awefully wrong.
I have to agree – I would wager sony’s lens margins for a-mount lenses equal or exceed Canon or Nikon’s margins based on their pricing and lack lens IS. Plus many lenses are pretty old Minolta designs that have been manufactured for decades. Worst case scenario: the margins are the same as Canon since Canon ships in higher volume.
don’t forget that these (Sony’s) lenses sell waaay less than Canikon. Sony (on A-mount) has a pretty big scale disadvantage vs. its two main rivals. Not that easy to increase dramatically the refresh cycle of lenses, cut prices and widen the lens offering altogether.
But you’re right that its the BIG weakness of the A-mount.
Let’s hope it can shoot at variable RAW mp (like Canon) or use extra pixels to improve ISO.
If its like Foveon then 54mp would be divided by 3 for final file resolution.
Sony chasing Leica (MF legacy lenses obsession with A7). Sony chasing Sigma (Foveon sensor). What the hell is going on? Maybe Sony will discover Moon one day?
Sony isn’t chasing Leica. Some Leica lens users are simply excited about a body that may prove useful to them that is superior to Leica’s offerings at a fraction of the price. I’m confident that Sony would prefer you buy and use their lenses but they are supportive of adaptation (they gave away a Metabones EF adaptor for free in a promotion last summer).
The rumor doesn’t state that foveon will be used, commenters on here are speculating. Sony imaging is a sensor company first and a camera maker second so it is in their best interest to continue to innovate and push sensor design, whatever the underlying tech is called.
Finding profitable markets and then beating the the incumbent/ or buying them out. Sounds like a reasonable plan to me.
I’ll stick with my A900 and A77 for the next year. I just bought the A900 a month ago brand new from a Sony outlet store. I’m missing out on a generation of newer technology but oh well. I never was feeling the A99 unfortunetly. If the new alpha for sports and wildlife is the hot shiit than I’m in. Must be 36MP and be a serious upgrade over the A99 in terms of buffer speed and- well, that’s it actually- maybe ditch that bubble look and take on a more A7 styled design.
As far as the new sensor- awesome. Keep pushing the boundaries as much as possible. That’s why a lot of us are in this game. I mean, with good composition you can take a great photo with an IPHONE. Most of use are pro’s anyhow. However, I love the tech and welcome whatever Sony has up it’s sleeves next.
Most of us aren’t pros…is what I meant to say.
Wait for 5 axis ibis and focus shift sensor, so u can use any lens.
Dont complain about lack of lenses or high priced zeiss.
U don’t like it go buy canikon, don’t stick with alpha
uhh I hope it is high iso capable foveon style (I’d feel bad for sigma though, they have been really great lately, on the other hand I wonder how much profit they are making with their sensors and cameras, maybe it’s even better for them to just focus on the lenses, if they keep giving us more such lenses as the 18-35/1.8, 35/1.4, 24-70/2 and so on they should be doing quite well) not x-trans style bs 🙂
this means I will probably get a 24MP fullframe for now, since the price difference is greater than the resolution difference and I can always stitch if I want a huge resolution.
Mirrorless consolidation continues behind the scenes as now Fuji is working with Sony. Panny and Oly already are as well. Soon those brands will all get swallowed up by Sony imaging. Then they may actually have an imaging division that is profitable.
There is dimishing return on investment for customers too. What will a new sensor be able to do that much better than the excellent Sony 36 MP sensor? Sony needs to a lot of work to do to get the cameras right and a full system. The sensor is not what makes me hestitating to purchase a A7r. It’s the missing lenses, the AF, the lack of trust that Sony is already longterm committed to the system. (Sony sill drop on systen sooner or later: A or E mount.) But see are talking of different business units of course …
Btw, the F828 had non Bayer pattern color filter.
What about the possibility of a CYMK sensor. Wouldn’t that have the most accurate color rendition?
It got issues with artifacts and color accuracy.
Perhaps they solved it, who knows?
Oh, if I’m not mistaken – CMYK is the one CFA that requires really a lot of computing power to demosaic it into usable JPG. Which will be another large obstacle (though certainly not as huge as it was 5-10 years ago).
why CMYK?
LOL CMYK I think you mean CMY. k=black. Be kind of weird for the sensor to have a black pixel.
real film uses CMY dye couplers but with as many as 14-15 emulation layers.
I think Nikon had a point and shot that had a CMY sensor back in the early 2000’s. I can’t really find too much info on it…
CMYK? is K for black? xD
CMYG or CMYW produce greater luminosity sensitivity, but when you need to output it on screen on paper, chromatic noise will be much heavier. And, IQ on low ISOs will be worse than for RGB. And, RGB allows more accurate lateral chromatism supression in postprocessing.
He quite clearly means CMY or CMYW sensor that works in CMYK color space as it’s native. That in theory might have some advantages in print applications, but… I would say that disadvantages of that sensor overwhelm any advantages that it brings.
I have one request…please provide pixel binning to emulate larger pixel pitch(smaller MP) at raw level. Not everybody need full res raw, but pretty much everybody wants more than 8-bit jpg color data.
This is why I like Sony . They always come out with new technolegy and new innovation .No one know what is next and how it works.
I think if its 56MP it will be a new medium format camera. Lets not forget Sony has now tied up with HB
IF the 24MP & 36MP are also new , it should go with the next gen Sony Alpha – A8 & A8R coz it might have v.good AF 🙂 G8 news !
Maybe they try CMYK sensor? (maybe another color scheme..)
Well, Sony makes the sensor for the Nikon D810, maybe this sensor will make it in the D820 when released. Either way, Sony makes money!
well did this rumor ever happen?