Sony to unveil a 50 MP new A7 body at Photokina?

[mrsfotografie]

The sensor is real...

Oh, here it is:

http://www.sonyalpharumors.com/two-new-medium-format-sony-sensor-camerasbacks/

 

[jrista]

The sensor is real...

Oh, here it is:

http://www.sonyalpharumors.com/two-new-medium-format-sony-sensor-camerasbacks/

That's the medium format sensor. I believe this rumor was about a 35mm FF sized 50mp sensor, for the A7 series of bodies (which don't have a mount large enough to handle a medium format sensor).

 

[mkabi]

As per recent rumor: http://www.sonyalpharumors.com/sr4-no-other-new-camera-from-sony-at-photokina-sr5-new-generation-sensor-and-cameras-release-in-january/

There is no 50MP A7 at photokina 2014.

 

[sdsr]

The benefit between 22 and 36mp bodies is negligible when using let's say the Tamron 24-70mm. Once you get the zeiss 135mm APO, the difference is stark.

However, doesn't change my point that even at 36mp it's hard to get the detail out of it and let alone a 50mp sensor in 35mm format.

Funny you should mention that lens, inasmuch as Roger Cicala reports that the Tamron 24-70 on a D800 outperforms in terms of resolution the (superior) Canon 24-70II on a 5DIII, a combination that outperforms the Tamron 24-70 + 5DIII. You can't attach the Canon to a D800, but you can (and people do so) to a Sony A7r, so....

I'm not sure what you really mean by "get the detail out of it". You seem here, and in other posts, to suggest that unless you own lenses that can make the most of (whatever that means) a higher resolution sensor, there's no point in using a camera with a higher resolution sensor (other things - such as noise, dynamic range, etc. - being equal, presumably). I'll leave the science to others, but have you tried this yourself or is your argument based on speculation? I've attached a fairly wide array of lenses to my A7r, mostly Canon EF (including some rather inexpensive ones) but a few others as well, including some fairly elderly inexpensive manual lenses even older than those you sneeringly (or so it seems) invoked in another post, and with only one exception so far they are capable of sharp detailed images when viewed at 100%, probably more so than when I use them on my 5DIII (which is not to say the differences are significant). Is your experience different?

But of course the appeal of a sensor, let alone a camera, isn't just its resolution - it's other aspects of its performance; when Sony (or whoever else) releases a 50mm FF camera the other things mentioned above won't be equal. E.g. there are reasons to like the Sony A7 line independent of sensor resolution - noise, dynamic range, that fact that the sensors are housed in mirrorless bodies to which you can attach just about any lens and which have good EVFs, and so on. Given all the other benefits (they're not for everyone, of course), it's nice to be able to attach EF lenses and obtain images that look at least as good as they do via their native bodies. Whether those lenses "gets the most out of" the sensor doesn't matter much to me, and I suspect it won't to lots of others too; you can always supplement them with lenses that perform better. It would be disappointing if the images looked worse, but no-one has provided any reasons to suppose that they would.

 

[RLPhoto]

The benefit between 22 and 36mp bodies is negligible when using let's say the Tamron 24-70mm. Once you get the zeiss 135mm APO, the difference is stark.

However, doesn't change my point that even at 36mp it's hard to get the detail out of it and let alone a 50mp sensor in 35mm format.

Funny you should mention that lens, inasmuch as Roger Cicala reports that the Tamron 24-70 on a D800 outperforms in terms of resolution the (superior) Canon 24-70II on a 5DIII, a combination that outperforms the Tamron 24-70 + 5DIII. You can't attach the Canon to a D800, but you can (and people do so) to a Sony A7r, so....

I'm not sure what you really mean by "get the detail out of it". You seem here, and in other posts, to suggest that unless you own lenses that can make the most of (whatever that means) a higher resolution sensor, there's no point in using a camera with a higher resolution sensor (other things - such as noise, dynamic range, etc. - being equal, presumably). I'll leave the science to others, but have you tried this yourself or is your argument based on speculation? I've attached a fairly wide array of lenses to my A7r, mostly Canon EF (including some rather inexpensive ones) but a few others as well, including some fairly elderly inexpensive manual lenses even older than those you sneeringly (or so it seems) invoked in another post, and with only one exception so far they are capable of sharp detailed images when viewed at 100%, probably more so than when I use them on my 5DIII (which is not to say the differences are significant). Is your experience different?

But of course the appeal of a sensor, let alone a camera, isn't just its resolution - it's other aspects of its performance; when Sony (or whoever else) releases a 50mm FF camera the other things mentioned above won't be equal. E.g. there are reasons to like the Sony A7 line independent of sensor resolution - noise, dynamic range, that fact that the sensors are housed in mirrorless bodies to which you can attach just about any lens and which have good EVFs, and so on. Given all the other benefits (they're not for everyone, of course), it's nice to be able to attach EF lenses and obtain images that look at least as good as they do via their native bodies. Whether those lenses "gets the most out of" the sensor doesn't matter much to me, and I suspect it won't to lots of others too; you can always supplement them with lenses that perform better. It would be disappointing if the images looked worse, but no-one has provided any reasons to suppose that they would.

In short, yes. My 50L in particular is a great example, It looked really good on my 5Dc but on the MK3 it was less than stellar wide open and stopped down it was sometimes hard to tell the difference between them. The 135L saw a marginal improvement and the 24L II saw it alittle bit wide open. Stopped down performance 5Dc vs 5D3(ISO 100), There is a gain but I sometimes would edit a 5Dc file that I thought was 5D3 file. In practice, then going to print, it was a wash 90% of the time.

The biggest differences were with the 135L between the two cameras. That lens get much much closer to resolving the full 22mp of the 5D3. (which it never will but it gets closer)

My original point here is sony will sell a 50MP 35mm sensor, now will you ever get close to getting the full 50MP out of it with most of your lenses? Probably not, and the differences in practice will be similar to the current A7R. The only way you'll be able to resolve the majority of those 50MP to Definitely SEE the differences, is with the Otus or other lenses like it. Also don't forget we don't always shoot our lenses at F/8 and have to use wider apertures which makes the difference even less so.

Can sonys native A7 Lenses do the job? I don't believe they will but they'll sell alot of cameras advertising those 50MP.

 

[jrista]

My original point here is sony will sell a 50MP 35mm sensor, now will you ever get close to getting the full 50MP out of it with most of your lenses? Probably not, and the differences in practice will be similar to the current A7R. The only way you'll be able to resolve the majority of those 50MP to Definitely SEE the differences, is with the Otus or other lenses like it. Also don't forget we don't always shoot our lenses at F/8 and have to use wider apertures which makes the difference even less so.

You won't ever realize the full 50mp...however, you could indeed realize more resolution than the 22.3mp or 36.3mp of a lesser camera. That's the point, right? Look at my numbers above. Theoretically, a 50mp at just f/4 could at best resolve over 37% more than a 22.3mp sensor. That's pretty much guaranteed if you slap something like an Otus on, but any other diffraction limited f/4 lens, of which many come very close to, will also perform better. Maybe not 40%...but a 10% difference is enough to be meaningful, and a 20% difference would be excellent.

It's not about resolving the maximum that any given sensor is capable of. It's about resolving more (in the output image) than a lesser sensor is capable of. You won't ever achieve the maximum sensor resolution so long as the sensor is the limiting factor...but you can and will realize useful gains by moving from a lower-resolution to a higher-resolution sensor. If that was not the case, we wouldn't clearly see such a HUGE difference between a 5D III and a D800. The difference in the smallest resolvable details between those two cameras is quite stark...even with lenses less than a Zeiss 135 or Otus. I have seen marked differences in the sharp in-focus parts with fast 85mm optics wide open...the periphery of the photos may be soft or totally OOF boke...but for the parts that are sharp...i.e. the center of the lens (which pretty much performs ideally on the majority of lenses)...the differences between a 5D III and D800 are massive. (And no, I don't mean in terms of DR...I just mean in terms of raw spatial resolution.)

 

[RLPhoto]

My original point here is sony will sell a 50MP 35mm sensor, now will you ever get close to getting the full 50MP out of it with most of your lenses? Probably not, and the differences in practice will be similar to the current A7R. The only way you'll be able to resolve the majority of those 50MP to Definitely SEE the differences, is with the Otus or other lenses like it. Also don't forget we don't always shoot our lenses at F/8 and have to use wider apertures which makes the difference even less so.

You won't ever realize the full 50mp...however, you could indeed realize more resolution than the 22.3mp or 36.3mp of a lesser camera. That's the point, right? Look at my numbers above. Theoretically, a 50mp at just f/4 could at best resolve over 37% more than a 22.3mp sensor. That's pretty much guaranteed if you slap something like an Otus on, but any other diffraction limited f/4 lens, of which many come very close to, will also perform better. Maybe not 40%...but a 10% difference is enough to be meaningful, and a 20% difference would be excellent.

It's not about resolving the maximum that any given sensor is capable of. It's about resolving more (in the output image) than a lesser sensor is capable of. You won't ever achieve the maximum sensor resolution so long as the sensor is the limiting factor...but you can and will realize useful gains by moving from a lower-resolution to a higher-resolution sensor. If that was not the case, we wouldn't clearly see such a HUGE difference between a 5D III and a D800. The difference in the smallest resolvable details between those two cameras is quite stark...even with lenses less than a Zeiss 135 or Otus. I have seen marked differences in the sharp in-focus parts with fast 85mm optics wide open...the periphery of the photos may be soft or totally OOF boke...but for the parts that are sharp...i.e. the center of the lens (which pretty much performs ideally on the majority of lenses)...the differences between a 5D III and D800 are massive. (And no, I don't mean in terms of DR...I just mean in terms of raw spatial resolution.)

The d800 vs 5d3 resolution wise one shot with a 600LII and the equivalent Nikkor equal about the same total resolution in the image. That reflects my weekend with the d800 vs 5d3 with my L primes. The difference isn't as great as I thought it was going to be. The shots that were at f/8, the difference is more noticeable but at best marginal, like you said 10-15%. (Hardly massive.) To see the stark difference between the two, otus or better. (Then it's obvious, like 30% or more.)

And you cut out my biggest original point here, the A7 native lenses are not up to getting the majority of those 50mp yet. Sony should address their lens lineup before jumping ahead with more MP.

 

[jrista]

My original point here is sony will sell a 50MP 35mm sensor, now will you ever get close to getting the full 50MP out of it with most of your lenses? Probably not, and the differences in practice will be similar to the current A7R. The only way you'll be able to resolve the majority of those 50MP to Definitely SEE the differences, is with the Otus or other lenses like it. Also don't forget we don't always shoot our lenses at F/8 and have to use wider apertures which makes the difference even less so.

You won't ever realize the full 50mp...however, you could indeed realize more resolution than the 22.3mp or 36.3mp of a lesser camera. That's the point, right? Look at my numbers above. Theoretically, a 50mp at just f/4 could at best resolve over 37% more than a 22.3mp sensor. That's pretty much guaranteed if you slap something like an Otus on, but any other diffraction limited f/4 lens, of which many come very close to, will also perform better. Maybe not 40%...but a 10% difference is enough to be meaningful, and a 20% difference would be excellent.

It's not about resolving the maximum that any given sensor is capable of. It's about resolving more (in the output image) than a lesser sensor is capable of. You won't ever achieve the maximum sensor resolution so long as the sensor is the limiting factor...but you can and will realize useful gains by moving from a lower-resolution to a higher-resolution sensor. If that was not the case, we wouldn't clearly see such a HUGE difference between a 5D III and a D800. The difference in the smallest resolvable details between those two cameras is quite stark...even with lenses less than a Zeiss 135 or Otus. I have seen marked differences in the sharp in-focus parts with fast 85mm optics wide open...the periphery of the photos may be soft or totally OOF boke...but for the parts that are sharp...i.e. the center of the lens (which pretty much performs ideally on the majority of lenses)...the differences between a 5D III and D800 are massive. (And no, I don't mean in terms of DR...I just mean in terms of raw spatial resolution.)

The d800 vs 5d3 resolution wise one shot with a 600LII and the equivalent Nikkor equal about the same total resolution in the image. That reflects my weekend with the d800 vs 5d3 with my L primes. The difference isn't as great as I thought it was going to be. The shots that were at f/8, the difference is more noticeable but at best marginal, like you said 10-15%. (Hardly massive.) To see the stark difference between the two, otus or better. (Then it's obvious, like 30% or more.)

And you cut out my biggest original point here, the A7 native lenses are not up to getting the majority of those 50mp yet. Sony should address there lens lineup before jumping ahead with more MP.

Maybe it's just a difference in perception. I'd need to see the photos you took with the D800 and 5D III to see if I personally could see a difference. I never said anything was massive...just that 10% is meaningful, 20% is excellent. If anyone could actually realize a 40% difference, THAT would be massive...but you would really need a diffraction limited lens. I figure the Otus could do it. I think there are a few Canon normals and short teles that could do it. I think that many lenses, in the center of the frame, can do it as well. If you want that kind of performance corner to corner...then yeah...Otus and maybe some of the Canon great whites are probably some of the very few lenses that can do that.

 

[RLPhoto]

My original point here is sony will sell a 50MP 35mm sensor, now will you ever get close to getting the full 50MP out of it with most of your lenses? Probably not, and the differences in practice will be similar to the current A7R. The only way you'll be able to resolve the majority of those 50MP to Definitely SEE the differences, is with the Otus or other lenses like it. Also don't forget we don't always shoot our lenses at F/8 and have to use wider apertures which makes the difference even less so.

You won't ever realize the full 50mp...however, you could indeed realize more resolution than the 22.3mp or 36.3mp of a lesser camera. That's the point, right? Look at my numbers above. Theoretically, a 50mp at just f/4 could at best resolve over 37% more than a 22.3mp sensor. That's pretty much guaranteed if you slap something like an Otus on, but any other diffraction limited f/4 lens, of which many come very close to, will also perform better. Maybe not 40%...but a 10% difference is enough to be meaningful, and a 20% difference would be excellent.

It's not about resolving the maximum that any given sensor is capable of. It's about resolving more (in the output image) than a lesser sensor is capable of. You won't ever achieve the maximum sensor resolution so long as the sensor is the limiting factor...but you can and will realize useful gains by moving from a lower-resolution to a higher-resolution sensor. If that was not the case, we wouldn't clearly see such a HUGE difference between a 5D III and a D800. The difference in the smallest resolvable details between those two cameras is quite stark...even with lenses less than a Zeiss 135 or Otus. I have seen marked differences in the sharp in-focus parts with fast 85mm optics wide open...the periphery of the photos may be soft or totally OOF boke...but for the parts that are sharp...i.e. the center of the lens (which pretty much performs ideally on the majority of lenses)...the differences between a 5D III and D800 are massive. (And no, I don't mean in terms of DR...I just mean in terms of raw spatial resolution.)

The d800 vs 5d3 resolution wise one shot with a 600LII and the equivalent Nikkor equal about the same total resolution in the image. That reflects my weekend with the d800 vs 5d3 with my L primes. The difference isn't as great as I thought it was going to be. The shots that were at f/8, the difference is more noticeable but at best marginal, like you said 10-15%. (Hardly massive.) To see the stark difference between the two, otus or better. (Then it's obvious, like 30% or more.)

And you cut out my biggest original point here, the A7 native lenses are not up to getting the majority of those 50mp yet. Sony should address there lens lineup before jumping ahead with more MP.

Maybe it's just a difference in perception. I'd need to see the photos you took with the D800 and 5D III to see if I personally could see a difference. I never said anything was massive...just that 10% is meaningful, 20% is excellent. If anyone could actually realize a 40% difference, THAT would be massive...but you would really need a diffraction limited lens. I figure the Otus could do it. I think there are a few Canon normals and short teles that could do it. I think that many lenses, in the center of the frame, can do it as well. If you want that kind of performance corner to corner...then yeah...Otus and maybe some of the Canon great whites are probably some of the very few lenses that can do that.

I wish I had them still. It was work for hire as a 2nd photog and we switched each others equipment for the wedding. We reviewed the files on his mac and we didn't notice the difference until the formals @ f/5.6-f/8 and it wasn't a huge difference either. He expected more from the d800 but it wasn't the E version either.

In short, he kept his gear/cards and gave me back mine. I don't have the files but I remember it wasn't that impressive with our tamrons and L&G primes at most of the working apertures we used.

 

[jrista]

My original point here is sony will sell a 50MP 35mm sensor, now will you ever get close to getting the full 50MP out of it with most of your lenses? Probably not, and the differences in practice will be similar to the current A7R. The only way you'll be able to resolve the majority of those 50MP to Definitely SEE the differences, is with the Otus or other lenses like it. Also don't forget we don't always shoot our lenses at F/8 and have to use wider apertures which makes the difference even less so.

You won't ever realize the full 50mp...however, you could indeed realize more resolution than the 22.3mp or 36.3mp of a lesser camera. That's the point, right? Look at my numbers above. Theoretically, a 50mp at just f/4 could at best resolve over 37% more than a 22.3mp sensor. That's pretty much guaranteed if you slap something like an Otus on, but any other diffraction limited f/4 lens, of which many come very close to, will also perform better. Maybe not 40%...but a 10% difference is enough to be meaningful, and a 20% difference would be excellent.

It's not about resolving the maximum that any given sensor is capable of. It's about resolving more (in the output image) than a lesser sensor is capable of. You won't ever achieve the maximum sensor resolution so long as the sensor is the limiting factor...but you can and will realize useful gains by moving from a lower-resolution to a higher-resolution sensor. If that was not the case, we wouldn't clearly see such a HUGE difference between a 5D III and a D800. The difference in the smallest resolvable details between those two cameras is quite stark...even with lenses less than a Zeiss 135 or Otus. I have seen marked differences in the sharp in-focus parts with fast 85mm optics wide open...the periphery of the photos may be soft or totally OOF boke...but for the parts that are sharp...i.e. the center of the lens (which pretty much performs ideally on the majority of lenses)...the differences between a 5D III and D800 are massive. (And no, I don't mean in terms of DR...I just mean in terms of raw spatial resolution.)

The d800 vs 5d3 resolution wise one shot with a 600LII and the equivalent Nikkor equal about the same total resolution in the image. That reflects my weekend with the d800 vs 5d3 with my L primes. The difference isn't as great as I thought it was going to be. The shots that were at f/8, the difference is more noticeable but at best marginal, like you said 10-15%. (Hardly massive.) To see the stark difference between the two, otus or better. (Then it's obvious, like 30% or more.)

And you cut out my biggest original point here, the A7 native lenses are not up to getting the majority of those 50mp yet. Sony should address there lens lineup before jumping ahead with more MP.

Maybe it's just a difference in perception. I'd need to see the photos you took with the D800 and 5D III to see if I personally could see a difference. I never said anything was massive...just that 10% is meaningful, 20% is excellent. If anyone could actually realize a 40% difference, THAT would be massive...but you would really need a diffraction limited lens. I figure the Otus could do it. I think there are a few Canon normals and short teles that could do it. I think that many lenses, in the center of the frame, can do it as well. If you want that kind of performance corner to corner...then yeah...Otus and maybe some of the Canon great whites are probably some of the very few lenses that can do that.

I wish I had them still. It was work for hire as a 2nd photog and we switched each others equipment for the wedding. We reviewed the files on his mac and we didn't notice the difference until the formals @ f/5.6-f/8 and it wasn't a huge difference either. He expected more from the d800 but it wasn't the E version either.

In short, he kept his gear/cards and gave me back mine. I don't have the files but I remember it wasn't that impressive with our tamrons and L&G primes at most of the working apertures we used.

If you were working at around f/5.6 to f/8, the differences are going to be much smaller, for sure. Your more diffraction limited at those apertures, particularly f/8. There are some differences, but they won't be huge. Your going to notice more of a difference at f/4 and faster in the center of the lens (or across the frame with something like an Otus), since the diffraction-limited resolving power is much higher at those apertures. At f/8 and below, sensors are starting to oversample the lens...so, as you saw, differences were more noticeable. However, because you are diffraction limited, the differences cannot be large.

I'd be willing to bet with either companies 135, or the Zeiss 135, or even similar short tele Sigmas, that the differences in the center and center midframe would be eminently noticeable.

Diffraction is a fickle friend. By stopping down, you eliminate optical aberrations, which can present in a thousand ways and affect IQ in just as many ways. However, at the same time, by stopping down, your implicitly limiting your maximum resolution.