A7RII initial thoughts and images

[eronald]

Bart,

 Yes, the sensor assembly oscillates, not the sensor itself - in the same way as the frame of the window bears the glass - I stand corrected.

 The issue about BSI is more complex. Because it allows connections to be made to the body of the sensor, not only to the edges, it becomes a major functional advantage. Also, it probably simplifies cell design, bus design etc, and reduces the cell shrinkage inefficencies. It is possible also that some design tricks might later arise that would increase DR or enable continuous readout, maybe by means of a "hidden" charge storage area. I'm not sure that the inter-line transfer idea is permanently dead.

 As for having in-lens stabilisers vs. in-body, the lens stabiliser is restricted to one technology, the body stabiliser gets updated. These days ONE long lens costs more than a body, maybe you hadn't noticed the way things have changed. Getting a body fixed/swapped under insurance is probably easier than a complex long lens repair with parts that are hard to source and need careful adjustment.

 I don't think they're using piezo, I think it's probably voice coil. But then why would I know? I also don't know how they control it.

 As for Sony's track record on aftersales - their video equipment is the pro standard, and used all over the world, so when they want to do things properly they certainly can.

 And yes, I agree they live on instant obsolescence. But is that not better than the arrested development of the rest of the industry? Phone cameras have been killing the big cams, it is time somebody fought back with  high-ISO, high frame rates etc.

Edmund

Hi Edmund,

I'm not so sure about the DR (in fact I'd be surprised if it was as good as its non-BSI predecessors), but it should help quantum efficiency because it has fewer obstructions / transfer gates, transistors, etc. blocking the light.

As far as I understand it, it's more the sensor assembly that can oscillate, not the sensor itself/by itself. That means that in principle, any sensor could use the stabilisation assembly.
 
While sensor stabilization is an interesting development with lots of potential, I'm wondering though which works better; lens stabilization or camera body stabilization. I have not seen any good comparisons, and I wonder where the camera shake is larger (and more accurately detected). IOW where is the center of rotation situated. Also, the sensor itself has a reduced size image to shift/rotate, which is a benefit as to the amount of displacement needed, but also requires much more accuracy (which I assume the piezo actuators are capable of).

For professional use, it's of course also much more vulnerable to have a single stabilizer (that has to operate for all different lenses and thus gets to do more work during its lifetime) that can fail (as all mechanical stuff does at some point) and become unavailable for all lenses at the same time (will the camera continue to work?), instead of a lens which can be replaced at relatively short notice, rather than having to wait for a body repair. Service repairs do not seem to be Sony's strong suit across the globe either.

Yes, certainly kudos to them for that. Yet, I still have a feeling that their technology becomes obsolete almost as fast as they introduce new stuff.

I therefore wonder about their track record for servicing 'older' camera models, what is their spare-parts and service training policy? Or is built-in obsolescence something we need to factor into the purchase price (as a new fact of life like non-replaceable phone batteries, battery dead after a thousand or so recharges, phone dead)?

Just some thoughts.

Cheers,
Bart

[Quentin]

Interesting fun shoot by Jason Lanier using the A7RII and FE lenses

https://www.youtube.com/watch?v=M3CwK21Hak4

[rainer_v]

I cant see as well why the 5ds is a better still camera, just because it might work better with canon lenses. The files from the sonys as from the hi-mp nikons are not even close to the canons and simple another class at lo isos. This counts if you use this cameras for architecture, landscape or still shootings.
Another big advantage  of the sony is  that you can mount virtually all lenses on it, excpet some rangefinders with very close exit pupils. If you put a leica wate on it, or a contax 45mm, or the 35 or 55mm fe lenses you really get a small camera, on which you can put as well the canonTS-e lenses or whatever you like or need.
Or you can put the a7r (1 or 2) under a drone, without exceeding with the whole package 5kilo, which in many countries is the limit to fly them more or less easy.

[Paul Roark]

Looking at the image stabilization -- "Steady shot" -- it looks like 1/15 with the 35mm can make top quality B&W prints.  I used a 10 stop ND filter to test outside, comparing a 200 iso 1/500 shot to a 5000 iso 1/15 shot.  100% comparison image slides are at http://www.paulroark.com/BW-Info/Sony-a7rii-35mm-SteadyShot-test.jpg .

The merged image is made from three 1/15th shots -- multi-sampling to increase the s/n ratio.  Obviously wherever there is movement only one frame can be used, but those areas (the trees) don't show much grain in the single shot.

Paul
www.PaulRoark.com

[D Fuller]

hi there, i kind of agree but understood that this was a thread to provide images and information about this new camera, if you'd like to see my photo prowess here is a link to some books i have published, most are out of print and command high prices, apologies if the "test shot" offended your asthetic sensibilities!:
http://www.adriantyler.net/alhambra.html

I certainly meant no ofense, Adrian. In fact I meant nothing personal at all, and certainly no aspersions on your photo prowess. (Very nice work, buy the way.)

My own aesthetic sensibilities don't really come into play either. These shots are obviously not intended to be artistic work.

My real point is that they are not very helpful tests. (Not meaning to single you out, the web is littered with "night for day" tests of the a7s, and will soon be with the a7r.) There is a lot to be learned about how the camera performs in shooting night scenes, astral photography, etc. but I don't see many samples that are helpful in that regard. I'd contribute some myself, but I don't own the camera. (And that's exactly why I'm interested.)

[rainer_v]

I received my a7rii today with the option to take a critical look on it till monday. First shots taken from tripod at base iso up to iso 6400 dont show an advantage for noise against the a7r. I would rate them as equallly good.
If the a7r2 wont be visible better in still mode, for me only remain two advantages ... stabilisation and 1. curtain.  It will make it hard for me to decide. I dont shoot videos so i dont care 4k.

[qwz]

the worse noise in Canon shadows, when downsampled to Sony, might reduce or even eliminate the DR advantage.

Nope. You have to sample at least 84 megapixels (twice than 42) to get visible effect of down-sampling. And around 168 to get 1 stop difference.
But With 5ds(r) you can use noise reduction a little bit effectively 'cause a more detail resolved.

[eronald]

Olympus and Sigma are going to make E mount lenses. I guess everybody is reading the writing on the wall.

Edmund

[SecondFocus]

I asked this in another thread but let me try here too...

Question for anyone shooting the A7 series cameras. Have any of you experienced photographing the really fast moving? Like an aircraft coming in to ground level and across you at 300+ mph? Would be interested in knowing how these cameras handle something like that.

Thanks!

[BernardLanguillier]

I asked this in another thread but let me try here too...

Question for anyone shooting the A7 series cameras. Have any of you experienced photographing the really fast moving? Like an aircraft coming in to ground level and across you at 300+ mph? Would be interested in knowing how these cameras handle something like that.

That, or a small kid moving randomly at short distance.

Thanks.


Cheers,
Bernard

[shadowblade]

Olympus and Sigma are going to make E mount lenses. I guess everybody is reading the writing on the wall.

Edmund

I'd be interested to know *which* E-mount lenses they will make.

It'd be fantastic if they made their 35/50 f/1.4 'Art' lenses in E-mount (and the longer 'Sport' lenses later on, once mirrorless AF catches up to fast action photography), so that we have some solid, fast, no-compromises lenses for the system.

What wouldn't be so great is if they simply bought into the 'mirrorless=small' philosophy and didn't release these lenses, but simply made new, slower, cheaper and smaller lenses to tap into budget buyers and those who want small size at the expense of everything else. We have enough small lenses for E-mount already - what we lack is capable, fast lenses that are the equal of SLR lenses, for pro users concerned about performance rather than size.

[eronald]

Re. Sigma, I saw this:

http://www.bhphotovideo.com/c/product/1121080-REG/sigma_401_205_24mm_f_1_4_dg_hsm.html

It is the 24mm 1.4 ART,  but it is -strangely- A mount. My apologies for a misread.

Never say never. I thought A mount was dead ...

Edmund

I'd be interested to know *which* E-mount lenses they will make.

It'd be fantastic if they made their 35/50 f/1.4 'Art' lenses in E-mount (and the longer 'Sport' lenses later on, once mirrorless AF catches up to fast action photography), so that we have some solid, fast, no-compromises lenses for the system.

What wouldn't be so great is if they simply bought into the 'mirrorless=small' philosophy and didn't release these lenses, but simply made new, slower, cheaper and smaller lenses to tap into budget buyers and those who want small size at the expense of everything else. We have enough small lenses for E-mount already - what we lack is capable, fast lenses that are the equal of SLR lenses, for pro users concerned about performance rather than size.

[shadowblade]

I kept the sony zeiss 24 to 70 a mount f2.8 lens in case I bought one of the A7 cameras, because I'll use it mostly for motion and f4 is too limiting for a lot of our work.

I haven't put in on an a7RII but did try it months ago on the A7 and the lens seems to weigh twice what the body did, (not actually, just feels that way).

I don't know how there going to get e mount lenses in f2 or f2.8 across the line and keep the weight down, unless they go to some type of composite material, but that would probably be very expensive.

Forget small size or light weight. You can only make a f/1.4 prime, f/2.8 zoom or f/4 supertele so small. For professional-use lenses, performance and capabilities trump size and weight almost every time. Small and cheap worked when mirrorless was targeting the beginner or 'step-up from point-and-shoot' crowd, but, with more and more professional non-action photographers using Sony (including quite a number of full-time wedding photographers around here) they need lenses capable of matching or surpassing SLR lenses in all respects, not just miniature budget lenses.

In regards of focus I have a series of different "mirrorless" or half mirrorless cameras that have autofocus by pdaf or contrast detection, or stuff they dreamed up and they're pretty good but all work subject dependant.  The gh3 is great at face tracking, the only em-1 good at faces and subjects, the 70 does great video autofocusing but lousy still image focusing and none of the Sony's I tested (haven't tried the RII yet) could do video autofocus without a lot of hunting.

For stills, I don't think anything get's close to standard mirror cameras.  The F5 was marvelous, the D3 as good though my Canon 1dx was better than my D3.

This was shot with a early production d3 in Japan for a gig and though Sanya is not a jet plane she's almost as fast as a dragster.

In this series the D3 hit 15 out of 20 consistently.

I don't think mirrorless can do this yet.

IMO

BC

Ever seen the AF on a professional video camera used to shoot sports, or a portable, pro-level camcorder used to shoot wildlife? Mirrorless, and the AF is nothing short of amazing - just as fast and accurate as the D4s or 1Dx. The technology is there. Thing is, running AF and an imperceptible-latency display at that speed requires fast processors and lots of battery power - something a small, mirrorless body doesn't have just yet. Mirrorless cameras are capable of being better than SLRs, or smaller - just, for the moment, not both at the same time.

[shadowblade]

Re. Sigma, I saw this:

http://www.bhphotovideo.com/c/product/1121080-REG/sigma_401_205_24mm_f_1_4_dg_hsm.html

It is the 24mm 1.4 ART,  but it is -strangely- A mount. My apologies for a misread.

Never say never. I thought A mount was dead ...

Edmund

I don't know why Sigma would try to push A-mount any further. There are far more people using E-mount cameras than A-mount cameras, and the momentum is also heading further in the direction of E-mount. The shorter flange distance has been a huge benefit for those looking to move to Sony for the better sensor (and better mirrorless technology in general).

If the A7r had been released with the A-mount instead of the E-mount, I doubt it would have sold even half as well.

The only viable future for A-mount would be if it was combined with a movable sensor plane, to adjust the flange distance while keeping the mount itself the same. That way, you could mount whatever lens you wanted to the mount and still use it, allowing Sony to continue poaching existing users of other systems, while allowing it retain the wider calibre of the A-mount and its advantages for certain lenses.

[eronald]

I don't know why Sigma would try to push A-mount any further. There are far more people using E-mount cameras than A-mount cameras, and the momentum is also heading further in the direction of E-mount. The shorter flange distance has been a huge benefit for those looking to move to Sony for the better sensor (and better mirrorless technology in general).

If the A7r had been released with the A-mount instead of the E-mount, I doubt it would have sold even half as well.

The only viable future for A-mount would be if it was combined with a movable sensor plane, to adjust the flange distance while keeping the mount itself the same. That way, you could mount whatever lens you wanted to the mount and still use it, allowing Sony to continue poaching existing users of other systems, while allowing it retain the wider calibre of the A-mount and its advantages for certain lenses.

backplane AF ...

Edmund

[shadowblade]

backplane AF ...

Edmund

That's certainly useful in some situations, although using backplane adjustment for AF rather than for correction of flange distance has the disadvantage of altering the magnification of the image, in a similar fashion to bellows factor in a LF camera. Of course, that can be used to advantage too, e.g. For macro photography.

[eronald]

i mean, the logical thing for Sony to try and kill next is the lens AF mechanics.
Maybe they need to keep the A mount alive for the extra flange distance it allows (requires), or some other physical parameter which is superior to E mount. It's dumb to kill a mount spec if you can reuse it, all that precision tooling and testing ...

Edmund

[shadowblade]

i mean, the logical thing for Sony to try and kill next is the lens AF mechanics.
Maybe they need to keep the A mount alive for the extra flange distance it allows (requires), or some other physical parameter which is superior to E mount. It's dumb to kill a mount spec if you can reuse it, all that precision tooling and testing ...

Edmund

The advantage of the E-mount isn't the mount itself, but the possibility of a short flange distance, allowing for the use of a much wider variety of lenses. But it's possible to keep the A-mount while shortening the flange distance to allow for the use of lenses with short flange distances, by using a movable sensor (and, with IBIS, the sensor is already movable). Keep the same socket and just move the sensor back and forth, to suit whatever lens is plugged into it - 46.5mm if you're using a Nikon F-mount lens, 44.5mm for A-mount lenses, 44mm for EF-mount lenses, 27.8mm for Leica M-mount lenses and 18mm for E-mount lenses.

Thing is, third-party lens compatibility is vital if Sony is to expand its market share - something that E-mount allows, but a non-movable A-mount would not. Being a latecomer to interchangeable-lens digital stills cameras, most people out there already have a collection of Canon or Nikon lenses they would be very hesitant to abandon in order to switch to Sony, no matter how good they make the cameras. Also, Sony itself does not have the expansive lens collection of either Nikon or Canon, making it much less attractive to newcomers without a pre-existing lens collection (sure, there are some ancient Minolta lenses, but, in the current age of high-resolution sensors, these are more relics of the past than working lenses that would be of interest to photographers investing in a new system). As such, they have to poach Nikon/Canon users and convince them to invest in Sony gear instead. But the need to abandon their existing gear is a huge barrier to switching. E-mount, allowing the use of third-party lenses, mostly removed this barrier. A movable-back A-mount would do much the same.

With regards to moving the sensor for AF as well as for flange distance adjustment, this has the disadvantage of altering the magnification of the image, sometimes to a large degree - for instance, when focusing close, a lens that is nominally 300mm could end up giving the same angle of view as a 240mm lens, or even shorter, if you had to move the sensor much closer to the lens to focus on a nearby object.

[Jack Hogan]

I received my a7rii today with the option to take a critical look on it till monday. First shots taken from tripod at base iso up to iso 6400 dont show an advantage for noise against the a7r. I would rate them as equallly good.

According to Bill Claff you should only notice an improvement at ISO640 and above (2/3 of a stop better DR for the A7RII).  Near base ISO the A7R is slightly (probably unnoticeably) better.

Jack

[adrian tyler]

I certainly meant no ofense, Adrian. In fact I meant nothing personal at all, and certainly no aspersions on your photo prowess. (Very nice work, buy the way.)

My own aesthetic sensibilities don't really come into play either. These shots are obviously not intended to be artistic work.

My real point is that they are not very helpful tests. (Not meaning to single you out, the web is littered with "night for day" tests of the a7s, and will soon be with the a7r.) There is a lot to be learned about how the camera performs in shooting night scenes, astral photography, etc. but I don't see many samples that are helpful in that regard. I'd contribute some myself, but I don't own the camera. (And that's exactly why I'm interested.)

no worries, thanks for clarifying, here you go, hope these help, all hand held, it took about 4-5 shots to get a sharp one at 0,5 seconds but from 6th - 8th second upwards, you can generally get tack sharp fotos hand held at least on the on the 55mm or 17mm (i don't have zooms or teles.) focusing took a while but it is do-able and taking into account that it was night time!


185 = 0,5 sec;   f/1,8;   ISO 1600; sony 55mm (really pushed in acr)
175 = 0,5 sec;   f/1,8;   ISO 1600; sony 55mm (really pushed in acr)
148 = 1/30 sec;   f/1,8;   ISO 1600; sony 55mm (not pushed in acr)