I checked the RAW files in Adobe Camera Raw, and yes, they are overexposed.

I did the same. I think Graham is right, there aren't too many. Most are OK, the ones exposed for the players with windows behind are obviously over exposed and I don't count the light bulbs as 'clipped highlights'.

I think that Graham is sitting a dark, windowless room with the brightness on his monitor turned down to 50 nits. Overexposed is overexposed. Looking at the RAW files, there are lots of blown highlight. I could fix most of the exposure issues in post, although I can't do much about the noise in the high ISO images or the general softness.

We must have different versions of ACR. I don't see 'lots' of blown highlights. I see a 'few' yes. I also see lights that are blown, yes as in light bulbs - but would you seriously expose for them not to be blown??. I also see shots where the photographer has exposed for the players but the windows behind are blown.

My assumption is that DPR have done this as they've uploaded the RAW's untouched. Would be a strange looking gallery if they'd exposed for highlights to lift shadows in ACR - would be a pretty dark gallery.

They are on the bright side but correct exposure is somewhat objective, it seems like the subjects are mostly properly exposed which has caused some blowing out in other less important areas. If all the highlights had been preserved the subjects would probably be underexposed, you can't have it all, especially with challenging light.

The "correct" exposure has as much to do with artistic intent as it does technicalities, hence high and low key lighting, silhouettes etc.....

ISO free sensor, tell me more.

And what color is the color cast you pretend to see ?

What I find is the colour cast is less problem with 24-70mm F4 compare to 70-200mm has more colour cast. Mostly toward cyan and green cast. Like what I see in those sample, same story of that. That test photos are all AWB I believed it should be as I know Dpreview does that to test it using AWB to see how it goes with the new lens. My best guess is that, more glasses are needed for telephoto to create a telephoto lens, so probably more colour cast than usual is to be expected. 10/12 lens groups/elements for 24-70 compare to 70-200 is 15/21 is a lot more.

So exactly how bad is this? And do you have something more than "my best guess"?

When you say "AWB" do you mean Auto white balance or is that Sony's name for custom? Because I see a color cast to these photos as well.
The only way to really get rid of a cast is to set your white balance before you start shooting using a calibrated gray card. I prefer the WhiBal card from Michael Tapes, the guy who does the LensAlign tool.

@naththo
No other way to say this but you don't really know what you're talking about. Lenses do not have colour casts in the sense you mean. If the pictures look too green or two blue (check white) it's white balance.

A lens of this quality is likely to be more colour neutral than a cheaper lens, not less.

If there is too low light for the Sony camera, there will not be more light for a Canon camera. I bet the results under the actual conditions would be pretty much the same with any of the best FF cameras.

@AngularJS
You just suggested that the test photos for a Sony 70-200mm lens should have been taken with a 200mm Nikon or Canon lens and 4 people liked it. They must genuinely be complete idiots.

RubberDials: Sure, something sounds wrong when one want to compare a telezoom lens with a one stop brighter prime lens, to get one full stop faster shutter speed. Why? First: This is a test of the Sony GM 70-200 mm zoom lens. Second: The roller pictures are action photos where medium fast shutter speed (1/500 sec) plays a significant role; we can see that some subject motion is causing blur, and for some of the samples might camera shake also add some blur. Third: The lens is used with a native full frame E mount mirrorless camera, to test how this spesific lens performs when tracking fast moving subjects.

Sure, one could compare cameras and AF performance, or sharpness, or noise, or whatever. But that would be something completely different than what is described by the Dpreview staff here!

@RubberDials - be polite, no need to call people idiots.

All I was trying to say is the test is not good to demonstrate the lens performance, unless getting noisy and blurry images was the goal. An outdoor sports event in a good light would yield much better results.

A 200/2 + a sports body would be my personal preference for an indoor event like that one. It's quite hard to catch a "peak moment" with a 5fps camera. Each camera / lens has own purpose - you don't bring a knife to a gunfight.

AngularJS wrote: "It's quite hard to catch a "peak moment" with a 5fps camera".

Unless you practise to push the shutter release button slightly before the peak moment - timing. Not that hard, and a technique that was widely used by photographers for more than a hundred year before high frame per second cameras were available, and still is pretty widely used. With 5 fps "spray and pray" often comes much worse out than shooting single frame with good timing.

Sometimes, no, often, the Zeiss produc demand is artificially high.
It's less of a demand and more of a lack of supply to create the impression of high demand.

Canon and Nikon somehow don't have the issues Zeiss do and sell MANY more products. I was waiting months for a Batis, then the 85GM was announced and I cancelled my order and bought the Sony, which was in stock only 4 weeks after launch.

A little earthquake you may have heard about has delayed a lot of things from Sony and Nikon.

@DamianFl
Got any figures to back up your assertion? I bet more people were trying to buy the Batis 85 than any Canon or Nikon 85. Canon may have sold more 85mm lenses over a ten year period but in the same time frame - I doubt it.

@RubberDials

it is a tad silly to think a lens made for a smaller system that is at least double the price of equivalent lenses from Canon and Nikon will sell more looking at B&H the Batis 85mm F/1.8 is nineteenth on the best seller list for mirrorless lenses . The Canon 85mm is 7th in the bestselling DSLR lens list and the Nikon 85mm is 12th

In Amazon UK the Batis lens does not even feature in the top 100 selling lenses. The Canon and Nikon 85mm both appear

Oh please.... as if you could tell which are "Sony colors" from a gallery of mixed photographs...

Lens is the problem can introduce colour cast like that and 70-200mm F4 previously with A7, green colour cast can be seen from this lens I own with A7. Thats a major issue problem. So far with 24-70mm F4 with A7, haven't had much issue problem so far. So its the lens, not the camera AWB.

lol, Sony colours.

Almost every broadcast cam in studios I go into is Sony, no one complains about the colours.

Canikon guys (read: Ken Rockwell readers) use 'Sony colours' as an excuse to bag a product, Sony doesn't have bad colour just as Zeiss products don't have imaginary '3D pop'.

Sony colours are different to Canikon colours, if you don't know how to post a Sony output, it's your fault, not Sony's, Sony JPG's are sweet, they look different to Canikon, Leica and Blad, but they are, in my subjective opinion, just as great in a different way.

There needs to be a new rule. Unless you know what 'hipster' means, you shouldn't be allowed to use it in snarky internet comments.

Today's definition of hipster is those who try to look like old school sailers... but take baths on a regular basis...

Photos shot from the hip, right?

Why yes, I am old...

There needs to be a new rule. DPR staff needs to know what the definition of 'Troll' is.

Oh I assure you, we extend that entry every day ;)

I like this definition from a 2009 article quoted on Wikipedia: in PopMatters magazine, Rob Horning asserted that the hipster might be the "embodiment of postmodernism as a spent force, revealing what happens when pastiche and irony exhaust themselves as aesthetics."

I have downloaded jpgs and raws, and inspected the files at 100% and 200% view on a calibrated monitor, but I am unable to spot this claimed softness. Contrast is high, dynamic range is marvellous, and the amount of fine detail is great, even for shots taken at 200 mm focal length.

But then, the pictures might look less sharp for people used with much lower Mp sensors, since those are great for surprssing aberrations = they don't resolve that much fine detail.

@noflashplease

Clean up the spit from your monitor, it should help.

Indeed, they look really soft 1:1, to me as well.

However, resample to 24MP, they look pretty sharp. Probably the lens can not resolve 42MP.

Regards

These images are just plain soft for a 70-200mm F/2.8 lens. I'm not talking about high ISO noise and I'm well away of the resolution. Nothing in these images is in the sort of sharp focus that I'd expect from a lens in this category. If we were talking about a DSLR, I'd suspect an issue with front/back focus, but that's not possible with a mirrorless body? Regardless of price, assuming that this copy is representative, the FE 70-200mm F2.8 GM OSS would appear to be unworthy of my camera bag. Sony needs to do better.

noflashplease: You are repeating yourself here! Again.

'Sony needs to do better'. People say such stupid things. Have you seen the MTFs of this lens?

'Unworthy of my camera bag' Which is full of Nikons, anyway. Go and look at the Nikon mtfs and then come back and tell us how soft this is.

@photomedium When you have some years of photography you quickly spot what is a good lens and a bad lens, in terms of sharpness this lens is obviously lacking on the finer, high frequency details, there are sufficient photos at diferent lights that should be enough to spot.

@Marcelobtp
nonsense.

1) Are you used with evaluating files from a high density sensor like the Sony A7rII at 100% pixel?

2) Looks like there is going on some kind of compressing/softing when large files are posted on Dpreview galleries - is the same true for their jpg samples?

I'm not the only one seeing the fundamental softness in these samples.

Professional sports shooters, as overwhelmingly shown by their choices.