(Updated June 15th after press conference)
The new Sony A7R II is the camera I’ve been waiting for, which everyone has predicted, and which seems to tick every box without having a huge price label on its own. I find the $3,200 (UK coinfirmed £2,600) matches its stated specifications well. Others may disagree, but they’re probably influenced by the price collapse of the original A7R, now occasionally found for under £1k.
Even so, at $3,200 the A7R II commands a $1,500 premium over the A7 II and much of that must be what you pay the new sensor – which does not seem to be licensed or sold to any other brand. Not even to Nikon, yet. The A7S remains the most expensive model despite the minimal 12 megapixel capture and lack of in-body stabilisation (SS in Sony terms, or IBIS generically).
On Monday June 15th I flew to London to have a look at the A7R II and the new RX10 II (£1,200) and RX100 IV (£850). This was a bit like a motoring journalist going to a car launch and being told, you can sit in the seat, waggle the gearstick but don’t start the engine as no photography was allowed with any of the demonstraton cameras. I was surprised to find it was a European conference, as this normally means journalists from across the Channel have a facility trip to be present, and that seems very extravagant just to look at cameras which can not be tried out. I wish I lived in France not Scotland – it might not have cost me almost £300 to be there, eight miles from Heathrow (but an eight miles which might as well be a fifty Scots miles!).
Don’t expect to get one on June 17th, as B&H’s information and too many bloggers have repeated. We are told by B&H it won’t arrive until August even though pre-orders open on June 17th in the USA. It may be later arriving in some regions. Demand is going to be so high that if you want one, you’ll need to crash into that queue…
But you can snag a Canon EOS 5DS – 50 megapixels – for only $3,699 right now
A7R II – or A7 II R?
In brief, the A7R II consists of an A7 II body with a new 42.4 megapixel backside-illuminated CMOS sensor, same Bionz X processor allowing 5fps at full resolution, new 399-point Phase Detection AF on the sensor covering most of the field (up from 117 points), a similar EVF with improved eyepiece giving a genuinely impressive 0.78X instead of 0.71X virtual magnification, the same rearRGBW bright LCD, plus silent shutter and HD 4K movie functions improving on the offering of the A7S. The new shutter mechanism is claimed to have a 500,000 actuation life expectancy which puts it ahead of almost every pro DSLR yet made. The back of the camera body is magnesium, where it’s solid composite plastic in the A7II. And it has, unlike the A7R, five-axis sensor stabilisation which talks to Sony OSS lenses for the best blend of anti-shake methods ever devised.
The new EVF size, to the eye – compared with the old (A7II, A7R, A7) 0.71X view below (A7R, Sigma 12-24mm at 12mm, Canon EF fit, on Commlite EF-FE adaptor).
You will read in the specifications and promo blurb that it has a new LCD double the brightness, new tough body and strengthened mount, new shutter release and controls but all these ‘improvements’ are listed by Sony over the A7R and already existed in the A7 II. Instead of making comparisons with the A7 II – which this is really a development from – Sony has listed many advances made relative to the A7R. It is not an A7R II. It’s really an A7 II R.
The eyepiece surround is much improved, wider and softer still than the A7II which in turn is softer on specs then the earlier models. Eyepoint and position flexibility both improve and there are no unsharp zones at all even if you shift your eye around.
It’s important to understand that many of the improvements already exist in the A7 II partly as a result of criticisms of the original A7R made by objective reviewers, not Sony artisans or staff or sponsored bloggers. You don’t owe this camera to the success of its predecessors or the daily Facebook sermons of awestruck evangelists – you owe its features to corrections made to the shortcomings of the models so far. And to those who have had no vested interest (other than ownership) persuading them to weaken critical appraisal. The further improvements in the A7R II are either extremely technical – serious core improvements in the sensor and focusing – or minor refinements and carries-over from the A7II.
42.4 versus 36 point anything
If you really think 42.4 megapixels is going to take you to realms far beyond your 36 megapixel sensor, think again. It is the same step up as from 18 megapixels to 21 megapixels, a move Canon made without absolutely transforming the images created, or about the same as from 10mp to 12mp. There’s one big difference – it does not make the jump to any larger common print or repro size. Remember going from 6 to 8? That was from sub-full-page to a decent full page resolution, for US or A-size documents at a touch under 300dpi. 24 megapixels took us to a really sharp A2, 36 megapixels takes us to a acceptable A1, and all that 42 does is to make a slightly better A1 but not 300dpi.
Above you can see the actual, real size difference (in proportion) between a 36 megapixel shot and a 42 megapixel shot. If you click on the bigger version, it will take you to my pBase page with a full A7R II sized version of this A7R shot. Zeiss? No – a 45 year old Asahi Pentax Super-Multi-Coated Macro Takumar 50mm f/4, used at f/11, and a 30 second exposure at ISO 50 lit using the ICE Light 2 moved round the subject in horseshoe shaped path for 15 seconds, laid flat, and then moved under the perspex for the remaining 15.
In practical terms, it’s 7980 x 5320 pixels (or very close – Sony has been extremely coy about releasing full specifications, even at the conference I could not find this out) versus 7360 x 4912 for the A7R. In perspective, make a big print from the A7R and it’s 24.5 inches long at optimum resolution; use the A7R II and you get one inch extra each end on the long side, 2/3rds of a inch extra top and bottom. The A7R makes a 16.3 x 24.5 inch print to perfection; the A7R II makes a 17.7 x 26.6 inch print.
Anything smaller than A4 printed, it’s got no great advantage over the 12 megapixel A7S – but you are getting close to enabling a 2X crop (one quarter of the frame) to look as good as the A7S full frame. Sony showed A3 prints. They could, honestly, all have been shot on the Sony A100 from 2006 and no-one would have been any the wiser. One enlarged section was the only real test of the camera. I’m sure the model’s dermatologist loves it.
Where it does count most is when using crop frame mode. In APS-C crop mode, the A7R II file is large enough for a 300dpi double page fine art magazine spread, just under 18 megapixels. I’d say that where 42.4mp is not a critical size, 18mp actually is. You can get away with 16, and for Nikon, Panasonic, and Olympus this had been an important baseline. Cropped frame FF from Sony now rises above that baseline instead of sitting just below it.
What I’d like to see would be 1:1, 4:3, 5:4 ratios implemented with the EVF and LCD screens cropped to match – and ideally the raw files reduced in size the same way. A square 1:1 would be 28 megapixels and that crop allows so many APS-C lenses (like the Zeiss Touit 12mm) to be used without vignetting or limits of coverage distortion issues. The example above is from the A7R and it’s a square crop 24 x 24mm from a frame taken with the 10-18mm f/4 Sony OSS, at 11mm; the lens would have allowed a 4:5 crop equally well.
Important edit: just read another ‘Sony artisan’ blog post asking the (redundant) question as to whether Sony lenses will be up to this new resolution. Anyone who owns an A6000, NEX-7, or A77 is already shooting at well above this resolution (full frame will need to match the Canon 5DS 50 megapixels to beat them). The resolution of the A7R II is slightly lower than that of the base level entry A3000. Don’t panic. Plenty of old legacy lenses will match it well, let alone any new Sony FE and A-mount designs.
I checked out the 20mm f/2.8 SEL lens with the new version 2 wide and fisheye black converters on full frame at the Sony event. Really, this lens comes so close to doing a good full frame and the converters even leave much of the area intact for a much bigger crop than APS-C.
And that’s all without removing rear baffles or doctoring the built-in lens hoods of the converters!
When we get a chance to use the camera, the following points will be of interest:
Has the mount been upgraded again? It still has only four attachment screws, compared to Fujfilm X system’s six screws (and the A-mount uses six too). My two camera bodies and two changes of mount on the A7R, to Tough E mount and then 2nd generation Tough E mount, all produce unpredictable degrees of slop, smoothness or jam-on tightness from various adaptors showing that no matter what, tolerances are broad. Comment: can’t tell from changing lenses at the event, it feels much the same as the A7 II.
Has the Memory position, 1 and 2 on the mode dial, been improved to remember MORE of the important settings – notable, Setting Effect ON and OFF, for saving a studio flash preset mode with the EVF/LCD setting effect disabled? Answer: No.
Is the hot shoe part of the Multi Function Accessory Shoe hampered by paint, or tolerances in fit, or does it readily accept all standard ISO hot shoe simple flash devices and triggers? Looks clear.
Canon 85mm f/1.8 USM on Focus EF-FE adaptor (also works perfectly with Commlite) on A7R. The 40mm f/2.8, and Sigma 12-24mm in EF mount work well on my A7R with these two sub-Metabones price adaptors. At the press event we found the 85mm just didn’t focus at all with any adaptor on any of the pre-production A7R II bodies, but the 40mm was fine.
Will the promised ability to use PD-on-sensor AF with Canon and other lenses rely on Metabones as the only adaptor, or is it generic? The microlenses on a backside illuminated sensor have a large effective aperture than traditional design, and this means the PD-lenses (a special variant of the microlenses used on sensel pairs) will be similarly improved. This may make some difference, but it’s actually the focus motor control via lens to body data communication which will enable fast and sure operation with Sony SSM on LA-EA3, Canon USM on EF-adaptor, and so on. Remember, this does not make screw drive or SAM, or micromotor Canon AF pre-USM lenses, function any better. It will only apply to ultrasonic, piezo, linear motor and similar finely controllable AF mechanisms with close to zero play and accurate (8 contacts, not 5) distance and ‘state’ reporting. Note, too, that Sony’s revised lenses (SSM II) are not just optical and weatherproofing reworks – the new SSM is designed to work with contrast detection, as found on the A7R, much better.
Comment: we found that the Canon 85mm f/1.8 USMdidn’t work on any adaptor on the A7R II, while the 40mm f/2.8 activated the PDAF points and focused very rapidly, and a 24mm f/2.8 USMf/2.8 focused fast – and that various different demo A7R II bodies responded differently and one malfunctioned a lot of the time even with Metabones. Sony said this was known and the final retail stock should at least work OK with Metabones IV and probable firmware updates, but other cheaper adaptors will not be tested.
The new camera’s mode dial has a central lock button, and a slightly lighter click action without risk of being turned by mistake. We’d had liked to have seen a lock on the +/- EV compensation dial too, but this just has slightly strengthened clicks.
The same small battery has been used yet again despite the II body design having what looks like enough room for a full sized Alpha battery (see below – carefully positioned batteries with A7 II body). Let’s hope for upgraded batteries from Sony.
Please, Sony, you provided a GPS pinout in the new shoe – you have never rolled out a GPS module or firmware. It’s three years now and no news. Hell, I nearly bought a brand new boxed A99 at Dixons Heathrow Terminal 2 shop for £1075 inc VAT maanger’s special, I miss GPS so much!
Please let the Lens Data entered into the menu for SS of manual lenses, without data communication, be embedded into EXIF so if I enter 50mm, my files say so. And ideally, please make it possible to enter the focused distance (this would improve stabilisation) and the aperture in use (just to complete the EXIF data).
Sony pointed out that the latest version of the lens correction App will record the focal length and aperture as you enter them, in EXIF. It has its own SS on/off setting and automatically recognises whatever focal length you have entered. You can name and recall each different lens, and if for example you normally use your 24mm f/3.5 Samyang shift lens at f/16 for architecture, you can enter f/16 as the lens’s aperture and that will be corrected embedded in your EXIF. But to get this you must run the app, not just shoot with a manually set focal length for SS.
Please change the Memory 1 and 2 registers to save and recall ALL the camera settings and not just those in the first bank of the menu system (but see the vital point above about Setting Effect On/Off). Until I test the camera, no more to say – but Sony does not usually keep quiet about changes, and has not mentioned this aspect.
The existing rear screen – the II design, left, improves on the original A7R but this is still a basic, amateur level screen to be working with and a fully articulated design would be better.
Though you’ve missed the boat with this camera, the crudely hinged and angled rear screen needs to be replaced with a fully articulated screen that can be reversed to the camera for protection and to prevent distracting light when working in the dark.
Out of the loop
I’ve been out of reviewing new Sony gear for some time, as it has not proved possible to get hold of it early enough or for long enough to give any meaningful assessment which Joe Photographer anywhere in the world couldn’t appear to do themselves. For six or seven years I have bought and sold new Alpha gear to fill the gaps between the occasional availability of review kit, but recently that has become so expensive it exceeds any margins available from the three magazines I publish, or any fees I can obtain from other media. Like politicians, people who write about gear either need an independent mind or independent means – without one of these, you’re always in someone’s corporate pocket or feeding from crumbs under the main table.
The result, as we see all the time, is that many early users or reviewers of Sony kit are no longer all that independent and much of the first wave of information now comes through the channel of ‘artisans’ (as it does with ‘ambassadors’ for all makes). And we see plenty of others who are clearly of independent means, whose main purpose in life is to be the first to post pictures taken with new item X regardless of the cost.
So maybe I don’t need to push to get hold of an A7R II for the too-short two week period of any review loan, after a six month wait while other consumer-orientated magazines and blogs take priority – or indeed rush to buy one.
But… like the RX10 which I use all the time… like the A6000 kit which is co affordable and compact it’s essential… like the RX100 MkIII which goes where even the RX10 is not welcome… like my A7 II with stabilisation which has transformed a box of assorted lenses into a solid outfit… this one’s possibly something to buy because I actually need it and will use it.
I may not even cosy anything as it will make both the A7R and A7II redundant, because it does both jobs and also covers the A7S I did manage to borrow but never bought. And it does more.
So, thinking whether or not to bother with this upgrade is a bit irrelevant. Even if it was still ‘just’ 36mp the other improvements would mean it still replaced the need for a handful of A7 models, all in one.
My one doubt is that the A7R II may be beaten in practical terms by the RX10 II. Please note that so many incorrect snippets of info have gone around about the ‘stacked’ sensor design, I thought it referred to RGB stacking. It does not, the sensor is a conventional Bayer pattern, and what is stacked is the electronic substructure. This does not affect the top side of the sensor and the performance in image quality should be similar to the existing models. What it does is greatly speed data transfer and enables over 1000 (lower resolution) frames per second to be clocked through from photon received to movie frame recorded.
The RX10 and 100 new versions offer ridiculous levels of high speed slow motion capture, clean 4K video and other technical benefits which come with a very small chance of dust on sensor, unlike the A7R II which is almost guaranteed to be a dust devil. Why do I say that? Because a backside illuminated sensor renders dust on its cover glass even more sharply than a conventional one! We know the RX models are not dustproof and if you are unlucky enough to get a spot on the sensor it’s a service visit to get it removed, but in my experience with five or them so far I have never had a single dust spot.
So what? Just retouch? Not when making movies! Admittedly most movie makers will open up the lenses to max or only a stop down on these 1″ sensor cameras, and would open up lenses just the same on the A7R II and never see dust even if it was there. But what about the time you want that ‘American take’ – f/22 at 20mm? Traditionally they were taken in dusty settings for the spaghetti westerns!
All I can say is that the RX10 has come very close indeed to removing the need for any other camera and it’s been a pleasure to work with the raw files. The RX10 MkII might be so much better that I forget about DSLRs or mirrorless systems and just get on with capturing great images. Or then again…
– David Kilpatrick