p.35 #1 · Official: Sony A7 and A7R Fullframe Mirrorless
raul jarquin wrote:
... I read somewhere that the new A7r Sony sensor has gap-less micro prisms. This is not good news just for wide angle lenses but also for collecting more light. I am not sure the D800 has that gap-less technology that provides a larger surface area to collect photons and focus them on the photo cells. Of course the sensor is only part of the electronic pipeline... Nikon has been very good in this area so I will be very surprised (but delighted) if Sony can do better.
Nearly all of the current FF sensors use gapless microlenses, including the D800/E, 5DM3, 6D, 1DX, D3s, and D4. Not sure about the D600.
Yep, from the original China thread the 10-18 is good in the middle focal ranges. For those unfamiliar with the lens it's extremely sharp across the frame on the APS-C NEX's starting at f/5.6. It will be interesting to see how the resolution holds up at the FF edges.
p.35 #4 · Official: Sony A7 and A7R Fullframe Mirrorless
What there is, is Sony statements that the sensor is both improved and new, with no reference whatsoever to the previous 36Mp sensor they provided Nikon. There is also a new BIONX Z processor, already thought to be instrumental in the increase in colour fidelity and richness.
Brian Smith reports very good AF speed as well on the a7r, I am thinking that Sony might see people using legacy adapted glass as a stop gap measure until native lenses come along that suit their needs.
'In daylight focus is faster than NEX-7, but I didn't test it out in really low light. I only tried the A7R and it's not as fast as the A7. Right now all the firmware is beta (I used FW 0.01) and I know there was an update while I was away, so I'd wait to pass judgement until it's final as it often gets better with the release version of firmware.'
shutter:
'The sound really bugged me the first day, but by day two I honestly didn't notice it. I think it's more of a sound issue than vibration. Solid metal body with nothing to dampen the thud. It didn't seem to hurt the sharpness one bit.'
prints from wide angles:
'I just stopped by to see a 60"x46" test print of a shot from Haiti done with A7R +Zeiss 18/4 ZM that was incredibly sharp. So yea, it hold up. It could be that the micro-lenses on the A7R sensor improve performance with extreme wide M-mount lenses. That certainly makes sense to me...'
p.35 #5 · Official: Sony A7 and A7R Fullframe Mirrorless
Not sure if this has been posted yet on FM but the Camera Store has a good 15 min hands one review of A7/7R and 35mm f/2.8. The interesting thing of this review is that both cameras are used with one doing the video recording while the other one is being reviewed. Dave
p.35 #8 · Official: Sony A7 and A7R Fullframe Mirrorless
HopeIsEternal wrote:
There is no evidence presented so far that the A7R processor is anything but a D800 sensor with customized microlenses and no AA filter. You are really speculating about better noise performance than D800/D800E.
Yes there is, Canon have stated in the past that any new camera's sensor will outperform a previous camera model's sensor. Common sense would suggest that an 18-month old sensor is not going to be used for a new product, it is most likely a second generation sensor and I wouldn't be surprised if it out DXOMark's the Nikons.
p.35 #9 · Official: Sony A7 and A7R Fullframe Mirrorless
It will be interesting to mount Tokina 11-16 F2.8 (aps-c) on A7r. Many use it successfully on full frame especially as a 16mm wide.
And why not, even zeiss E 24 1.8 looks like it will be perfect..
p.35 #10 · Official: Sony A7 and A7R Fullframe Mirrorless
efgm wrote:
Yes there is, Canon have stated in the past that any new camera's sensor will outperform a previous camera model's sensor. Common sense would suggest that an 18-month old sensor is not going to be used for a new product, it is most likely a second generation sensor and I wouldn't be surprised if it out DXOMark's the Nikons.
Agree... it is more logical to assume that the sensor is different, due to a) intervening time span b) competitive agenda and c) Sony owns the #$*&# sensor fab! than it is to assume that it'll be the same ol' sensor and microlens package as he D800 has.
Here's hoping (pun intended) it is another leg up the ladder.
This is quite a surprise! It's not too unusual for longer lenses to project a larger-than-necessary image circle, but for an UWA it's pretty big news. This will be a hell of a useful lens on FF35.
Craig Gillette wrote:
Won't the 10-18 be 10-18 on the ff cameras and like 15-24 on the aps-cs?
p.35 #14 · Official: Sony A7 and A7R Fullframe Mirrorless
Craig Gillette wrote:
Won't the 10-18 be 10-18 on the ff cameras and like 15-24 on the aps-cs?
exactly, I was just going in reverse, since I've only shot APS-c, never full frame. You would need a 6.8mm +/- UAW APSC lens to achieve FOV of 10mm on FF.
p.35 #15 · Official: Sony A7 and A7R Fullframe Mirrorless
redisburning wrote:
that is absolutely staggering.
I didnt think it was possible for the A7r to be any more hype but this + the positive report with the ZM18 has me outright drooling over this camera.
The shots here look great, almost too sharp, as they are Jpegs with increased sharpness and contrast, as mentioned by Gustav. I will be fascinated to see the eventual RAW files, as I have also noticed with the RX1R, there is vast difference between the RAW and Jpegs. I do feel this camera, for now with the 35/2.8 and 50/1.8 will be great, and we will be comparing to MF. I still have my A7r and 50/1.8 on pre-order. I have been informed by Sony, the first lens that will be available is the 35/2.8 being released at the same time as the A7's, and 50/1.8 will not available until Jan! It does concern me, with the uncertainty with third party lenses, that Sony will initially only be releasing one lens, not even at the same level as the RX1/R.
p.35 #16 · Official: Sony A7 and A7R Fullframe Mirrorless
nandadevieast wrote:
even zeiss E 24 1.8 looks like it will be perfect..
Only in crop mode, the image circle doesn't come close to covering the FF sensor. You may be able to remove the plastic shield at the rear of the lens to improve things a little, otherwise count on cropping.
It depends on the definition of "works". The Canon EF-S 10-22mm "works" on a 5D from 15-22mm, in the sense that there is no vignetting (and the mirror doesn't hit the back of the lens!)
However the corners of the image are complete mush. Which kind of makes sense - why would designers work to improve areas that would fall outside the intended APS-C sensor?
p.35 #18 · Official: Sony A7 and A7R Fullframe Mirrorless
Matt Grum wrote:
It looks and sounds like a form of deconvolution sharpening, the sort that most decent sharpening plugins can do nowadays. The only novelty is that it can be done by the in camera JPEG engine. I'd be very surprised if they were applying it to un-demosaiced RAW files!
Tariq Gibran wrote:
Looking at the image processing chain that Sony shows in the brochure, the Bionz X processing (which includes this sharpening/ detail enhancement) occurs before raw or jpeg output.
I'm not sure how you got "sharpening is applied to RAW files" from that diagram! All it means is the same microchip creates RAW and JPEG files. Which is incidentally the case with any DSLR. It doesn't say anything about processing applied to RAW files, and in any case I'm still convinced that you can't apply any sort of meaningful sharpening to RAW files without demosaicing as part of the process!
edwardkaraa wrote:
I think we will need like 80 mp FF sensor to get rid of moire (almost) completely. Of course, you can also reduce the lens sharpness by shooting stopped down to f/16, or opening up below the lens optimal aperture.
Tariq Gibran wrote:
Sony is using some sort of Diffraction-reducing technology as well that takes into account the lens aperture so I'm not sure if that trick will work. That will be interesting to test on these cameras. Some of the samples shot at F16 and F22 look remarkably sharp from the A7's...at 100%!
I imagine all they're doing is using a different point spread function for each aperture to drive a deconvolution algorithm. There really isn't much else you can do to reverse the effects of diffraction. But provided noise is low, deconvolution can work very well.
In any case using diffraction as an AA filter will still work, as it will still spread light over multiple pixels. The diffraction reduction system will be applied later in the signal processing chain, after demosaicing. But really it's not some incredible new technology, it's just software.
p.35 #19 · Official: Sony A7 and A7R Fullframe Mirrorless
Matt Grum wrote:
I'm not sure how you got "sharpening is applied to RAW files" from that diagram! All it means is the same microchip creates RAW and JPEG files. Which is incidentally the case with any DSLR. It doesn't say anything about processing applied to RAW files, and in any case I'm still convinced that you can't apply any sort of meaningful sharpening to RAW files without demosaicing as part of the process!
That would suit me fine. I got it from Sony's explanation of their Bionz X processor where they attribute it to all these features (and the diagram showing everything going through that processor). Sony does not say raw or jpeg in that description. While it makes perfect sense that raws may very well not be affected, it is also the case that Sony and others have "cooked" raw files in the past, which introduces some doubt as to whether this is occurring with jpegs only. I am inclined to believe (and hopeful) that the raw files are not "cooked".
p.35 #20 · Official: Sony A7 and A7R Fullframe Mirrorless
Interesting, it looks like Voigtlander is releasing their own helicoid m-mount-e-mount adapter, with close focusing ability. I have no real issue with the build on my Hawk's adapter, but perhaps this will be even more solid.
