ArtAlt: Thank you for your honesty (the "drawbacks" section). This is RARE in reviews of Sony cameras, which are often extremely misleading.
I've gotten badly burned twice by misleading reviews trumpeting the blazing fast AF of the a6000 (without mentioning the low light problems) and the amazing quality video (without mentioning overheating).
As a result, as much as I love my a6000 I don't trust Sony or most of the reviewers. When it comes to Sony I am in "show me" mode.
Under what conditions and settings did you find you had issues with the A6000 in low light? Unfortunately people just assume a DSLR approach to them and this doesn't always work.
nametaken: Are you going to test to see if the camera will focus wide open with A-mount and 3rd party lenses when shooting continuous bursts instead of focusing stopped down to whatever the aperture is set to like all E-mount cameras to date do with E-mount lenses?
Focussing at the taking aperture has an advantage with focus shift issues, obviously stopping down reduces focus shift, but you have to stop down a fair bit, it is in those middle ranges where focus shift becomes an issue... otherwise it was never an issue.
The other advantage is that the aperture doesn't have to reset after each shot, this is what causes very fast DSLRs to only operate at those speeds when standing on one leg while eating a ham during summer solstice to actually achieve those speeds.
This is also an advantage for video where AF works regardless of aperture setting, unlike say SLT where the aperture has to be fixed.
Lastly, from my understanding is if you put the camera into 'settings effect' mode into 'off' it should stay wide open until taking the shot.
njlarsen: I question the choice of lenses for this test. Short lenses are always going to be easier to track than long ones. When you get the actual samples for a real review, please make appropriate tests using a long lens (400 mm at least for a full frame). Don't do the easy part of telling us how it works with a standard lens -- all systems are perfect with that type of test.
Depends what kind of tracking really, but due to compression objects moving toward or away from the camera 'appear' to do so more slowly than a short FL lens at a much closer distance. Add that the DoF for this much shorter distance is less (especially with faster lenses available) and the range of being 'out' is very very small.
Yes tracking like this is often used at longer FL's but to suggest it is more taxing on the system is just wrong. Maintaining the subject is easier and maintaining an appropriate focus plane is easier at longer FL's. Also lens elements have to move much less in this scenario as well to maintain focus.
bongsogo: i hope this is at least as good as the a6000 focusing capabilities or even better...
It's better. It may not be able to claim 'world's fastest under xzy scenario blah blah', but it has far better control logic which was sorely lacking on my A6000.
helltormentor: @ Rishi SanyalSony's strategy on cameras has always been an enigma to me. It seems that they are more concerned about the number of sensors that they sell rather than that of cameras they manufacture. This camera has potential to persuade many enthusiasts to ditch their DSLRs for it but, strangely, Sony has crippled the camera in one way or another to make them think twice. When I was reading the article, I had a big grin on my face thinking the AF micro adjustment era was gone till I reached the section describing that the AF point couldn't be selected. Funny thing is that issues like this or lossy RAW can be addressed by just FW updates but Sony leaves much to be desired when it goes to FWs. Since you can directly talk to some Sony managers, have you ever asked why it is like this? I can accept that Sony doesn't pay attention to what people discuss in forums but I have difficulty believing they do nothing about a criticism that is mentioned on your reviews since they are well aware of the number of your readers. You mentioned this lossy RAW issue when you reviewed the A7/R but nothing changed and it continued its way to A7S, A7II and apparently A7RII. At present, a Sony manager says that they have heard the voice of photographers and they are going to address the issue by FW update. Seriously, how difficult could it be that it took them so long to start thinking about it?!! How long will it take them to let people choose the desired focus point on their adapted lenses?! Although it sounds silly, I cannot help but think that Sony has reached an agreement with Nikon not to cannibalize their sells since I believe this camera could do it.
Sorry no. Sony creates a set of features on the body, they don't make the adapters or how they 'translate'. It is up to the manufacturer of the adapter to make their adapter work properly. Rishi has already said those focus modes are available to the adapter, but the adapter doesn't work properly when in 'advanced' mode (which enables these focus modes). So it is up to Metabones not Sony to resolve this problem.
Yes because all other brands work so well with all other lenses? And Canon / Nikon are so very renowned for their firmware updates...
It isn't Sony's job nor intention to make their cameras 100% compatible with Canon or Nikon lenses.
neil holmes: I can not afford this and am happy with my A7s and A7 currently but Geez I like this new camera......and what is (I think) coming in future.
I think I will open a bank account today just to start saving for an A7sii
'The a7S did focus in lower light than anything we've ever tested, but it would take forever, slowly hunting before actually nailing it in very dim conditions.'
I have always wondered what settings you were actually using when testing this Rishi? When the A7S came out a number of users had similar issues, but with changes in mindset this is often not the case.
abortabort: Rishi, can you please test the Metabones in both 'green' mode and 'advanced' mode? They are functionally quite different as far as the camera is concerned and may offer different focus modes to be available.
As a side note, shouldn't that be in the text rather than a blanket 'those focus modes are not available to adapted lenses'?
I had the same issues with advanced mode on the older cameras, simply will not lock focus, but does enable other features. That's quite interesting at least those options open up, I wonder it this issue can be fixed? Other readers had the opposite results using advanced mode on their camera to me, so I wonder if there is some other setting that needs to be enabled? Also as advanced mode aperture works similar to native lenses whether that cause the problem in some way and maybe turning setting effect to off will change this behaviour.
Just thinking out loud. Thank you for the reply.
Searching: Sony is sure getting good coverage these days.
Every camera seems to get loads of coverage. Sony release more cameras than anyone else (not because they upgrade every two seconds like people seem to believe, but because they compete and bring out products in more segments than any other brand, often by a LONG margin). So it appears like there is always some new Sony out and to the casual viewer it 'appears' like Sony are releasing a new something every other day.
Dotes: Why not just shoot a video with continuous AF enabled instead of capturing the EVF with an Iphone?!
Sorry I meant a video captured in camera. An external video would either show or not show the overlays depending on settings.
PerL: So would this system enable Sony cameras to shot a low light indoor sports game with a 200 2.0 lens wide open at 10 fps with a higher number of keepers than a Canon 1Dx or a Nikon D4s, or is it just nerd stuff?
So I guess anything other than a 1DX or D4S is 'nerd' stuff?
And yes, this does seem to be more capable of subject tracking (not to be confused with AF-C) than a 1DX. Whether that means more keepers, who knows!
Rishi, can you please test the Metabones in both 'green' mode and 'advanced' mode? They are functionally quite different as far as the camera is concerned and may offer different focus modes to be available.
Infinal: I have a pretty dumb question: All this talk about mounting Canon EF lenses to the Sony a7R II while getting full PDAF; does that implicitly cover using other brands’ lenses (e.g., Nikon’s) in the same way, too? I only ever see people talk about how one can finally used phase-detection AF with Canon lenses on this Sony mirrorless, but never any other manufacturers’ lenses.Or is the big novelty simply that it’s finally possible with Canon glass as well, while it’s always been working with any other brand?
I completely agree that this 'may' have been possible with the A6000, however there may very well have been some usability limitations that Sony didn't want to have on a mass market consumer grade product. Maintaining a user experience, especially for novices, is often of a higher priority. While one may argue that having a limitation on third party lenses goes against this, we don't know what the side effects were / are. It may simply be that it is untested and they were erring on the side of caution. Remember when the very first PDAF enabled NEX models came out? They only worked with specific lenses and had to be updated to support new lenses. Then they released a FW update to 'enable PDAF with all and future lenses' which possibly just means they had more time to test the system, but they certainly 'lifted' that restriction. Fuji for example had to update all their lenses, so did Zeiss, to enable PDAF to function properly. What the 'cause' is, I don't know. I can only make guesstimates based on historical information. Sigma never did any kind of FW update for the DN series which predate PDAF mostly, so it is kind of a feat that these PDAF at all given the above.
Martinka: Why do you always stop for a while before taking the shot? This is not a typical usage of AF-C, so this test is more or less useless.
I think this test clearly shows how one can use the camera for focus-recompose and some additional movement from the operator and or model. It isn't a 'sports' test of someone continuously running. But it is a decent observation of how well the camera 'tracks' a subject across the frame, forward and aft, which is different to how fast continuous focus is. Managing to stay on the eye, regardless of where it moves is the thing being tested here, not the AF speed tracking it as it moves (you would need to shoot a series of continuous frames to get a better idea of that).
It definitely gives us an idea of how responsive the AF is and how well it sticks to a subject. Sticking to a person running should be considerably easier than sticking to someone's eye in a fairly dim lighting.
The A6000 was already pretty impressive in terms of AF and AF-C speed, but it lacked flexibility and control logic to work really well (it doesn't have expand flexible spot for example) and this let it down for me considerably. If the AF off sensor is as good or better than the A6000 but with its substantially improved control logic this should be an absolute winner.
nametaken: When you say it doesn't offer all focus modes for "3rd party lenses", do you also mean 3rd party lenses that were built for A-mount? Would the Sigma still have had the same problems of limited focus options if you had used one meant for A-mount instead of Canon?
@ Rishi - It 'might' simulate an LA-EA3 but there are other reasons to believe it actually doesn't, in that there are things the Metabones can do that the Sony cannot (such as auto VF magnification when changing focus distance and updating the on screen distance scale) so I suspect it isn't just pretending to be an LA-EA3.
Also the lens coding that the camera sees may be quite different to what it sees through the LA-EA3, the Sony shows the original A-Mount coding, the Metabones does it's own for Canon coding (pretty sure it shows the correct EF mount lens in the EXIF) which means the camera isn't seeing it as an A-Mount lens coding.
So there may very well be differences in the focus options available. I am particularly interested to find out as there may be advantages of going one mount over the other (EF already works with DMF where A-Mount doesn't, but if A-mount get more focus options it will go in their favour).
Metabones may be able to update their adapter to adjust this behaviour as well, as this is obviously a fairly new thing they haven't had time to try and take advantage of the features available.
Marcelobtp: This is almost like a revolution. This camera will be a Mark in history of photography and cameras in general. One thing, we need to see with objects moving too, in the use cases shown the IBIS is probably helping the af tracking compensating the movements of the camera, it can be completely diferent with moving objects and the camera steady, either way, i'm already Very impressed!
I would think that the faster readout speed and reducing rolling shutter would have a bigger impact on either camera moving or subject moving.
alextardif: I shoot a good amount of portraits with shallow DOF (1.8-2 on 85mm and 2-3.2 on 135mm). I've gotten quite good at catching the eye with A99 and CZ primes, but even my A7ii blows my A99 combos away. And... that's with a relatively slow AF on the new 90mm macro that just came out. I mean, it's almost magical! I can actually be in motion, walking along side my model, and let the camera do the work on catching the eye, it's flipping awesome. It'm actually changing my shooting a bit from spot focus to "wide" and just letting the camera hunt for the eye, which it does an admirable job of. In good light it's better than me! This is starting to have a real impact on my work - I focus more on composition and light and take less frames just because I have a higher rate of keepers that are in focus right there it needs to be.
New Batis 85mm and A7Rii can't arrive fast enough :)
@ Rishi - My understanding is that the focus assist lamps on flash units are 'near IR' (as you say IR beam) but as you know sensors have IR filtering on them, so the sensor cannot see that beam of light very well, which is why they need to use more of a visible spectrum light. The problem with these AF assist lamps is they actually reduce contrast significantly especially from the cameras PoV (light is completely front on), which makes the camera hunt. I find these actually work much better without the assist lamp (well existing Sony mirrorless that is).
There are other factors at play as well with mirrorless which means they must be treated differently to a DSLR when testing AF in low light. The first is that ISO makes a big difference: having them set to Auto ISO or a low ISO often means the sensor is operating at a low ISO and as such the 'AF module' isn't getting as much light (on a DSLR it makes zero difference what ISO you are using), or more correctly how it 'sees' the scene.
Aperture also affects focus because the camera will 'hold' the aperture at the taking size (by default) where a DSLR will keep the lens wide open when not shooting.
There are some remedies for this, you can up your ISO and keep your lens wide open. Or you can turn the 'settings effect' to off which means the camera optimises the setting for operation rather than preview of the resulting image.
Doing the above also speeds up the VF as the LV feed is different, I see lots of comments by people saying the EVF is slow etc etc, where I can guarantee that they are checking it out in a store with a stopped down aperture and / or low ISO and don't understand that WYSIWYG also applies to these things, if the EVF is 'slow' and laggy it's because that's how the resulting picture at those settings will turn out. This in turn affects the AF because as we know AF is done off sensor but the sensor is working at low refresh rates to 'display' the resulting image.
Also I find that on the A6000 at least, is that AF-C works better than the hybrid or more hybrid AF-S, which seems to be the way most people 'check' for AF in low light. The advantage with AF-C is that the lens is pretty much always at least close to the correct AF plane, so it is less distance to 'hunt' and it is being continually updated.
Testing the A6000 I found that the AF-C works perfectly fine throughout it's entire shooting envelope, which is to say that the AF-C only really conks out after a point in which you couldn't take a usable shot with it anyway. I expect this to be better again, given the spec and the larger pixel size that it too should be pretty well functional through the entire shooting envelope.
They couldn't put an SD card in. Besides a video won't show the focus points for obvious reasons.
ThePhilips: To make the AF in video really useful, they should start thinking about developing some cheap parafocal lenses.