A source told me that Sony is testing a High Resolution mode that works just like on the Olympus E-M5II. The sensor takes multiple shots in slightly different positions moved by the stabilization system and then combines time in one High Res Picture. The source told me he saw it creating 48.9 MP images on the fly. Of course that features works fine when shooting still objects. The source doesn’t know yet if the tested feature will work as hoped by Sony. If yes than it could be introduced in a future firmware upgrade. If you want to check out how good the Olympus solution works check out the test at Imaging Resource.
[poll id=”221″]
A7II links:
In USA at Amazon, BHphoto, Adorama, SonyStore and Sony Store Canada.
In Europe at Sony Germany, Sony UK, Jessops UK, Sony France, Sony Italy, Sony Spain, Sony Austria, Sony Switzerland, Sony Netherland, Sony Belgium, Sony Portugal, Sony Sweden, Sony Finland, Sony Norway.
In Asia at Sony Store Australia, Digitalrev and Amazon Japan.
A7II lleather case in [shoplink 32354 ebay]Black (Click here)[/shoplink], [shoplink 32396 ebay]Brown (Click here)[/shoplink], [shoplink 32355 ebay]Orange (Click here)[/shoplink] and [shoplink 32356 ebay]Camel (Click here)[/shoplink].
–
To not miss any rumor quickly subscribe our social network:
RSS feed: http://www.sonyalpharumors.com/feed/
Facebook: https://www.facebook.com/sonyalpharumors
Twitter: https://twitter.com/SonyAlphaRumors
GooglePlus: https://plus.google.com/104216500156717934391/posts
And if you have some “rumor” to share just send me an anonymous messages here:
[contact-form-7 id=”14822″ title=”Share”]
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!
101 Responses
I said it with the OLY implementation of this feature, and I say it again: This is very limmited. Best use in studio, on a very good tripod, with absolutely still objects. Any movement in the object or camera will distroy the high resolution and will add artifacts that are not nice. Many pictures on the net shows this. For me it is just a gimmick people will use once or twice, show the world how good it is and go on in normal shooting mode.
You know…some people do actually shoot on a tripod. Not everyone shoots exactly the same things exactly the same way.
You are absolutely right about that! But lets look at the cases where people want and/or need the higest resolution:
Wildlife: So you can crop your picture more, or get a higher resolution shot after cropping.
Landscape: to capture the most details in your picture
Architecture to capture the most details of a building, or to be able to crop for details
Maybe for portraits or fashion for showing fine details.
high res product shots, for the fine details in the products.
From al these options only architecture (when nobody is moving in front of the building) and high res product shots, this option is realy capable of delivering a good shot. That is why I call this option (very) limmited.
That pretty much sums up my concerns. Even for macro, there’s too many variables that could potentially ruin a shot. you really need absolute control over the subject and lighting for it to be effective. Also bear in mind that e-m 5ii owners don’t have a higher res option, whereas the above are all circumstances where an a7r would be the obvious solution.
Although it is a smaller part of my work, I do shoot architecture and interiors. If I owned an a7-2 this feature (if it works well – and I understand your concern that it may not) could render an a7r rental unnecessary for such shoots. No it’s not a mass market feature but if it works then it could broaden the usefulness of the a7.
Tried em5m2. Too many restrictions (tripod, still object), and the lens (7-14) is not good enough for high resolution, JPG looks good, RAW looks horrible, heavy sharpening has to be applied. If this new feature is for free then it’s good. Otherwise I’ll skip it.
This news increases the chance that no new 50 mpx sensor this year for Sony! Maybe Sony thinks it isn’t worth the investment for now, Canon is too far ahead in the market anyway and Sony is selling enough for now.. If Sony is not producing the Canon sensor, then Sony by itself might have not high enough production numbers to scale up that type of sensor. Hope I am wrong.
I’d prefer it if they updated the A7 firmware to match the focus of the A7II, rather than look at gimmicks like this.
The percentage of shots where this might prove useful is likely tiny (yes, I know there will be some people who could find it useful). Just look around at all the pictures you see on a daily basis, in various publications and on walls, and think whether this would have made the slightest noticable difference. If you see one in 10,000 I’d be surprised.
With that in mind, the survey is too limited with no ‘It’s a pointless exercise’ option!
Agreed, they should update the firmware so that the a7 has ibis and internal 4k.
Troll detected! He is asking for an implementation of the better algorithm that A7 II has over A7. But Sony won’t do it to keep claiming that A7II is faster than A7 for marketing purposes. But hey, they could just say, “30% faster than original A7*” and everyone would be happy.
Guess who is an A7 owner too.
*at its launch time
I’m not entirely sure whether you’re being serious here, but those are features which firmware can’t address.
Given that the sensor and processor are identical between the A7 and A7II, it should be straightforward to update the A7 focussing to match the A7II, although I realise the likelihood that it might affect sales. However, I think that would be a very slight effect, whereas the good PR such a move would generate could be of great worth.
An af upgrade should be perfectly doable. I can’t imagine many people are considering upgrading from the a7 to the a7ii, so it’s unlikely to impact on sales.
I think you can safely assume he was being ironic.
Also, do we know the A7II processor is identical? Even if it’s basically the same, it’s always possible it’s clocked at a faster speed and has more CPU power available. Also, camera CPUs aren’t general purpose in the sense of one in a computer. For speed and power efficiency reasons much of the logic is affectively “burned in” to provide hardware assists.
So it might be that AF can be improved in firmware alone, but it’s not guaranteed.
You may be correct, although it is almost universally quoted as being the same sensor and processor.
The reason even Sony gave for improving the AF by ‘30%’ was an ‘improved algorithm’. To me this screams pure firmware and the only reason why they haven’t implemented it for the Mk1 is potential damage to sales. After all, people still complained about AF speed compared to the competition, so I expect the thinking is that those seeking a little better would upgrade.
Personally, I believe this is a misjudgement and that improving existing cameras would nurture brand loyalty. Unfortunately, this has never been a strength of Sony; they seem to prefer a ‘take it or leave it’ approach!
Hoping Sony offers the option of either extra resolution (Olympus), or uninterpolated color (Pentax).
Agree – use it to lose moire
this feature would be a killer feature on a7sII. short exposures due to great sensor performance and mid iso + high resolution would make it a universal camera. great colors and low noise – a7sII, high resolution yet low noise – yet again a7sII.
Brilliant for studio still life shots and in the landscape on windless days. Image the quality if used with a 50MP sensor 🙂 Pentax will be implimenting this is their FF camera when its launched later this year
That would make me upgrade to A7m2 from A7.
Just about any camera can do this in PP – Google for super resolution or a bit of software called “photo acute”
it’s a feature I am thinking about since the Konica Minolta 7D introduced IBIS. Back in these days, CPU power was probably simply too low. But the a77/850/900 could had it certainly as a feature… hopefully we see it now!
another feature which would have even more impact is on-chip binning; if you have a 50MP sensor, than bining 4 pixels to one would result in an a7s like resolution and pixels with the same size (and probably a comparable sensitivity). But you can chose whether you like ultimate resolution or performance, having just one body. And with a 50MP sensor, you could even do that stiching thing resulting in approx. 75MP pictures….
please sony, I will kiss all of you for these both features, not only the first engineer I will meet on street 😉
Binning on Bayer pattern sensor presents all sorts of problems as the R & B channel sites are not adjacent, and the G sites only adjacent on the diagonal. So in each 4 x 4 block, you will get 1 x R channel, 1 x B channel (each of 4 photosites binned) and one G channel pixel made of 8 photosites.
In effect, with a 50MP sensor, you’d end up with something like a 3MP image, each with 3 full colour channels. Imagine it something like what you’d get from a 3MP Foveon-type sensor, albeit with much better colour separation. It’s going to be pretty well impossible to derive anything better than that 3MP as the binned photosites overlap (whilst they don’t on the native Bayer setup).
The sole advantage of using “binning” (at the charge level) rather than algorithmic down-sizing is that, in principle the charge-to-voltage-to-digital conversion change provides for a lower read-noise than the alternative of doing the same c-to-v-to-d conversion independently and then combining the results algorithmically. However, the algorithmic approach is far more versatile as it can generate a continuous balance between spatial resolution and SNR performance, and modern sensors have relatively low read noise.
There are some special filter layouts around, like Fujifilms “EXR” format which seeks to ameliorate some of the binning issues of the Bayer filter by placing pairs of R & B channel photosites on adjacent diagonals, but this involves other compromises.
Note that I’m being careful with terms here. Binning involves combining charges before the charge-to-voltage conversion point. Some people use the term far more loosely, including combing results after the ADC phase. That, to my mind, is not binning as, whilst it provides for a statistical reduction in noise, it still means that there is the combined noise from the separate read operations.
If binning made sense on Bayer-matris sensors, we would surely have seen it widely used as it’s hardly a new idea. As it is, it is largely confined to monochrome video cameras
There are design reasons why the A7S is better at ultra low-light which go beyond the “big pixel” factor. This can be seen in that the A7S is not a great performer at base ISO. There have been design compromises made which gives is stellar read performance at high ISO, but not at base ISO. (The Nikon D4s sensor has similar optimisations and characteristics).
How about if Sony moves the sensor so that each image pixel is build from 4 sensor pixels: R+G+G+B? This would substitute for Bayer color interpolation, yielding a resolution equal to the stated sensor resolution.
That’s not binning as that happens in the charge domain. What you are describing is more like stacking.
In any event, if the image is still enough that four exposures will work then you can surely just drop the ISO 2 EV and run with four times the exposure time.
What I tried to describe is 4 exposure stacking with 2×2 sensor shifting. Each pixel in the stacked image is then built from a red, 2 greens and a blue. No need for color interpolation or de-mosaicing. Equivalent to sequentially changing the color filters in front of each pixel.
Given Sony’s history, this is destined to be a feature of the A7mk3 released later this year. No surprise if they’re copying the high rez sensor shift, if the processor can handle it why not.
A7SII with that as an additional feature would be nice.
Looks limited in use, but does it mean there is definitely an electronic shutter that can avoid shutter slap?
Maybe while they’re at it, they can give us a custom minimum shutter speed in auto ISO.
Although at least the a7ii doesn’t seem to have the 1/60 love affair going on… An option to have it follow FL would be hugely appreciated!
If you have a LA-EA4 adapter, it’ll actually follow the focal length until 1/60. Somehow they forgot to enable that with native lenses.
Wouldn’t that be 96MP though?
The pixel dimension would depend entirely upon how the multiple images were processed. It’s not necessarily just stacking pixels from multiple exposures.
I am sure this will not be in a firmware update. If I am sony of course I would put it in other models or the A7III as the main selling point. Cameras are like cell phone now, new model will be out every year.
This might be true. Like all the developments of Sony cameras that many expected to be realeased allready kinda vanished right around the time that Olympus released it’s OMD-5II.. who knows. Now bring that to the upcoming 46-50mpix E-mount FF acmera say 60-80 mpix stills.
The poll misses the option “I don’t give a shit about it”, admin please add.
You forgot the poll option – MEH
It’s the same thing as “Nothing is ever good enough fo me.”
Well I see they left an option out especially for you WML, that is: I am an idiot and will just go with what ever i’m given.
Hate to break it to you buddy, you don’t have to be a sheep. You’re allowed to have an opinion.
I have an opinion and people love to hear it. LOVE IT!
I go to parties and tell people everything I think about all things. War in Sudan? I have an opinion. LHC and Higgs Boson?! I have an opinion! Brangelina adopting a new baby? Guess what, OPINION!
You should see the faces of the people when I am being so smart.
Wait a second because I have to smell my own fart.
I have an opinion and people love to hear it. LOVE IT!
I go to parties and tell people everything I think about all things. War in Sudan? I have an opinion. LHC and Higgs Boson?! I have an opinion! Brangelina adopting a new baby? Guess what, OPINION!
You should see the faces of the people when I am being so smart. I mean, the other day I was in the bus and I started telling this guy my opinion about why he shouldn’t use Android. He was just as stubborn as you.
This would only be of use for tripod mounted still objects.
I dont do that so this feature is of no use to me.
What I currently do is to , handheld, is take multiple close up pictures of each portion of the subject then stitch it in photoshop as a panorama.
In camera panorama modes are similarly useless because the camera produces pictures that are no more high res than a single picture.
High res panoramas must be done outside the camera.
How do you get the animals to hold still long enough to get multiple shots of each part of their bodies? Are they dead already?
Most animals are not constantly in motion.
Clearly this technique does not work for animals in motion. I make many exposures, when possible, because it sometimes works and sometimes does not
Because my photos are used for refferrence I can often combine photos that will not stich claenly but prpvide missing details for my paintings.
Unlike most photographers I fo not vare about ugly distractions in my images because I do not paint those.
The painting is the finished prodict andmy photos are refferrence tools for me to use in producing them.
When combinig photos that cannot be stiched I try to photograph the same animal from tje same angle with the same lighing. Even though the combined photo will show harsh edgesand missmathed background the painting will look consistent because the lighting matches.
The ideal resolution for me would be 78mp or more because this would allow me to make 24″x36″ refferrence prints at 300dpi.
This is the smallest size refferrence print I work with.
It would be great pleasure for me to be able to take a single ready to use photo without resorting to elaborate, time consuming workarounds.
On one occasion I spent 3 entire days comining 20 photos into a single high resolution refferrence photo.
Clearly the percent of keepers for me is low and often I have unusable photos of great subjects.
I am often accused of being obsessed with high resolution. ..but I have good reason for that obsesion.
As for mf…clerly it would be unsuitable for what I shoot.
“Most animals are not constantly in motion”- You must be referring to the dead ones.
You must of course know that the statement you reffer to cannot be disputed.
They are always in motion. Organs are always moving until dead
Since you seem to be using these images as source material for painting composition and detail, great. I question what your art is like, and its value as art. I’m not disparaging, yet–I’m genuinely asking.
Ok Barry, I tried, I really tried. I read your entire post After reading this drivel I have come to the conclusion that you need help.
First. You go on about resolution and sharpness and long lenses like a broken record. Then you have the balls to say you are doing all of this hand held. You need a tripod! If the animals sit still as long as you say then you have time to do it. A mono pod at absolute minimum. You are probably pushing the daylights out of your iso to get a hand holdable shutter speed on your bigma. To top it all off you don’t care about the quality of your hand held panos because you only need the pictures to look at to make paintings. WTF? For all the complaining you do and now we get this piece of info. I can’t throw an insult at you because I’d need too many to cover all the bases at this point. Knucklehead.
I utilize a simple camera support rig that provides stability midway betwwen a monopod and tripod for handheld photos…and makes carrying the equipment much easier.
You are too intent on finding fault in whatever I say to realize that there are lots of techniques I have developed to get results that you are ignorant of.
My need is primarily for detail and resolution, and many things that may be more important to you or others are not as important to me.
I do not do the same thing with my photos as you do but you are intent on judjing all my comments by your standards.
This is why you are incapable of finding value in my comments, only fault.
You and I clearly have differrent standards. What you value would for that reason fall far short of my standards.
That does not make your values wrong, but it does make you very close minded.
Barry, if a painting is your finished product why do you need such high resolution.
I incorporate a huge amount of detail into my paintings .
I need to see not only the details themselves but the structure and pattern in the detail in order to create it in the paintings.
I can guess at what is missing but take great pride in my anotomical accuracy, so that is very important to me.
It sounds to me like you aren’t much of a painter. Why even paint at all if you aren’t going to understand light, colour theory etc? You’re simply copying a photograph, that’s not very creative from a painting standpoint
Where it might help me most would be in shooting in APS-C mode. If one could get an A7m2 to have 21 MP equivalent in APS-C mode it would make small lenses a more attractive proposition. Something like the Sony 10-18mm lens would become a more attractive proposition, as would a 55-210. Or if you wanted greater reach on the 70-200 (for wildlife or leisurely paced sports), you could now go into APS-C mode and still get a lot of pixels.
You can not do wildlife and sports in High Resolution Mode …
No way you’re going to use this kind of functionality on wildlife or sports. Has to be a completely immobile object you’re shooting.
What’s with all the options? We have to say it’s a nice feature? Where is the ‘It’s a poor gimmick. They should just make it native 48MP to begin with.” option?
Where is the ‘It’s a poor gimmick. They should just make it native 48MP to begin with.” option?
That would be the unsubstantiated-opinion-internet-troll option. And we’ve all learned not to feed the trolls.
Right. Because all the polls have three approving options only.
Even if my camera was 48 Mp, I would still like this feature and have it boost the resolution even higher. I think it’s an addition, not a subtraction.
If people who already purchased the A7ii gets a free feature, then more power to them.
Except in this scenario Sony wants to use this 48mp gimmick as a replacement for a ‘real’ 48mp sensor. If Sony want to include this in all their cameras from now on, that’s fine. But don’t include it in a high resolution specific model as a cheap substitute for actual resolution.
This is just a free service for pre-existing users of the A7ii. It’s something engineered from software, it’s not going to take any development time from their hardware department. So you’re pretty much pissing on the breakfast of people who own the A7ii just because you have an irrational fear that Sony will never release a new camera body.
Shame on you Sony, the Nikon D7200 has better DR than any of your FF cameras…
http://www.dxomark.com/Cameras/Compare/Side-by-side/Nikon-D7200-versus-Sony-A7-II-versus-Sony-A7R___1020_996_917
Great News!
This technique is also excellent for true noise reduction in PS. PSs stacking mode can average out sensor noise by averaging pixel data. Also great for higher ISO shots. There are many useses for combining stacked images, resolution, noise reduction, focus stacking. panos.
Sony already does Stacking for noise reduction on the A77s I would think it would be in e-mount too. Multi-shot noise reduction
I just reviewed the Imagine Resource review of this feature on the Olympus, and it’s pretty impressive. If Sony could implement it with the same quality, why not? Granted, it’s only usable for still work on tripods, but still, would be great for those opportunities.
I’d welcome this for the way it gets rid of colour artifacts as much as for the extra resolution (assuming they implement it similarly to Olympus)
In fact, a 4 shot mode just for the colour, without the resolution gain, would also be great.
Horrible idea. Who would want to buy a camera that can only shoot vegatables.
JPG. It can produce RAW files, but Imaging Resource reports that:
– the RAW converter one must use is very limited compared to what we currently use, and
– the RAW file after conversion was no more detailed than the lower-res JPG.
I don’t want JPG. I don’t want limited RAW conversion.
-No need for crappy RAW converter since ACR 8.8 rolled out.
-RAW files are very robust when opened in ACR 8.8. IR’s lower-res jpg observation was due to the crappy Oly plug-in. It’s a whole new ball game with 8.8.
That would be a deal breaker for most people I’d guess. I’d you need the res, you almost certainly need raw…
I would like to have at least a dng or tiff.
Capture One just added support for the Olympus high-res shots.
Would be a nice touch for the A7rii, considering that it’s the “resolution” camera or whatever and probably the choice for people doing a lot of tripod and studio work.
If it’s on A7ii though, I certainly wouldn’t complain.
Why not try this to begin with?
http://www.tomen.de/enhance-superresolution-tutorial-in-adobe-photoshop-ian-norman/
Why would you not prefer that the camera takes the burst of photos and stitches them together for you rather than what this is showing where you take a burst of photos and then try to use photoshop to pull them all together?
I don’t mind either way.
It’s possible now…
I”m sorry. Until we have a native 135mm Zeiss 1.8 FE I can’t image a single picture worth using such high resolution on.
that’s the trick; the lens does not need to resolve 50 or more MPix… “just the 24 or what ever because “upscaling” ist is just combination of single frames. so, with a Sony 70-200/F4 or a Sigma A 50/F^1.4 or the new Zeiss lenses as well ass ome others you will be fine on a7 and a7r
A supper news
Sure, why not. Not going mad over it though, since most I shoot is out of hand.
just watched the Northrups’ review of this feature; apparently the most challenging part of it is to find something to actually use it for….even landscapes are tricky: if anything moves the slightest bit, it screws up the image. might be interesting for taking really weird photos tho..
Actually I would use it for just that 🙂 It is useful for product work, including plated food.
Rather than using the technique to up resolution, I hope they use it to take enough shots to get true colour images.
Olympus said that they were working on a version of this technique that could work hand held, I wonder if Sony’s will be able to
AFAIK Olympus has global shutter with 1 / 16.000 sec (the electronic one).
This allows to squeeze into 1/2000 sec 8 frames in (theory).
It does not matter if the sensor is moving half a pixel while recording a single frame, it results in a kind of AA-filter simulation like Pentax does.
I doubt Sony can do something like this without global shutter sensors 🙁
shutterspeed is not the limmiting factor here, data transfer speed is. When you have a sensor that can do 24 FPS then it is a very fast one, you can not have a faster pictjer taken then 1/3 of a second. To get a good shake free picture you need 1/50 of a second (or so, remember IBIS is of when using this option).
Will buy one tomorrow, 90% of the time I don’t need 48mp, but when I want it why not????
they could do this with the normal ibis of a mount, to atleast get foveonish pixel stacking
not interested. if i need higher resolution, i will buy a7r and its sequels or for still scenes use stitching. these features don’t give a stable results.
Animals are niether static nor constantly moving.
I photograph animals moving as well as capturing them between movements.
What I do …does not mean it is all I do or that I do it all the time.
I often work with refferrence that is far inferior to what I would preffer to work with.
The more detailed my refferrence the more certain I am of anatomical accuracy. The less detailed my refferrence the more I need to create my own detail and/or find alternate reference sources.
I shoot lots of shots and some of them are stitchable.
I also often work from multiple djifferrent pictures thati are totall dissimilar and combine the details.
My paintpings are hyperealistic and each painting goes beyond what the photo can capture in some way, otherwise there would be no point in painting it.
I require the maximum possible resolution because this detail is what I use to determine the structure of each hair, feather , scale etc and incorporate that into my paintin.
I do not copy each element but determine its structure and pattern and recreate that structure and pattern in the paintings.
If you can even ask why medium formgat is poorly suited for wildlife photograpihy than you clearly do not do ot yourself.
Where do you come up with this nonsense about dead animals…certainly not from any commenti that I have made.
To be honest – I am more interested of getting rid of AA-filter and use X/Y shift for AA-filter simulator like Pentax does. But I think there are patents 🙁
I’ll take anything they are offering for free! Who doesn’t like free? I hope it’s a free upgrade. 😀
I can’t be the only one who would be stunned if they added a feature like this in firmware… hardly Sony’s style.
I’m guessing this is similar to the Hasselblad MS system. This moves sensor one pixel over each frame so that you get full values of each RGB (someone please correct me if wrong). It does produce a noticeably sharper image but is an intensive process and needs tripod and still objects. A good marketing feature but with limited use
Yes, I will send Sony a bottle of good red wine,
IF Sony put it in the A99II, 🙂
ohh fine…so A7I goes down to 699,-$..yep:)
Food is so 2005 – it’s all about selfies now (apparently)