There isn't much discussion the new VR E version of the 24-70/2.8 here so I figured I'd start a topic.
Perhaps my main practical issue with the G version is that there is considerable field curvature at 24mm, which I would often try to use for large group shots, with the result that many of the people close to the edges of the frame at the sides of the long axis were out of focus at apertures such as f/4 or f/5.6. This was a serious issue for me as it didn't require a big print to show the problem, and I eventually resorted to using either the Zeiss 28/2 or the Nikon 24/1.4 for these shots. Especially the 24/1.4 yielded stunningly sharp results at every point in the frame, at f/4. Of course if there is a lot of light (provided or existing) then f/9 or some such aperture could be used but I would often need to balance the flash light with indoor ambient light and so I needed usable results at wider apertures. With two rows f/4 can be acceptable on a shot of 20-30 people, if the field curvature is absent.
Ok, so one of the first things I tested with the 24-70/2.8E VR was how does it behave in this type of a context. I did a hand-held series of shots at 24mm focal length at apertures from f/2.8 to f/11 and found that to me quite good looking result was already obtained at f/4. Here it is (if you download the "original" it is 3000 pix wide. Cropping has only been applied to the top and bottom, not the right or left side of the image):
https://www.flickr.com/photos/ilkka_nissila/22252717550/in/dateposted-public/I will try to find time to make a comparison shot with the older 24-70/2.8G as well, to illustrate the nature and magnitude of the problem. Also I may bring a tripod to get a better chance of perfect alignment but in this case I was able to align acceptably using my eyes only.
I find the new lens to autofocus faster than the old. I didn't notice a need to AF fine tune the lens with my D810 but I will investigate this in detail later on and see if the optimal adjustment is focal length dependent as it is on my G version of the lens and the same camera body (in practice I've settled on a compromise value across the range of focal lengths).
The VR function seems to work as expected, but you have to calibrate to my cynical expectations of VR technology. ;-) I took a set of shots of a large restaurant menu board at 1/13s and 1/30s and some of the shots at 1/13s were sharp and some were blurred. At 1/30s the shots were sharper than at 1/13s. In another context there was a group of police officers in full riot gear (due to a demonstration that had just ended) and I took some shots of them at 1/50s with VR on. Only one of the three shots was in focus (the gear had autofocused on the building behind the officers in two shots and on the face of the officer that I was trying to focus on in one of the three shots). One of the shots was, however, sharp and the VR had worked on that. From another angle I did two shots but one of them had a double image, the other was sharp. I think taking 2-3 shots at a minimum is a good plan when working with slowish speeds and trusting VR to help reduce the shake. I find this mildly annoying and would normally prefer to set the shutter speed at 1/200s at least in such situations, but in this case I was just getting a feel for what the VR could do. I think it can be helpful in situations where hand holding is required and the subject doesn't move too much but I would recommend getting a few frames just in case, to get a good chance of a successful result. I apologize if I appear overly cynical about it, but to me reproducibility is important and as I pay great deal of attention to the emotion of the subject being photographed, I don't want to have to risk the facial expression changing while I'm doing a series to get a good chance of a sharp image. At 1/200s or better yet, 1/500s there is a much better consistency to the images, but of course then the ISO or depth of field must be sacrificed in some cases. Anyway: verdict is that VR seems to work as expected and can get shots that one wouldn't get otherwise but the caveat is that it doesn't guarantee every shot is sharp.
With regards to the overall feel to the images, I find so far that the 24mm end is excellent, and the images look vividly saturated and contrasty with minimal tendency to flare (for such a complex lens). Sharpness and CA are well controlled. I think the 70mm end also looks reasonably sharp but it didn't make quite such a positive impression of a high consistency of high level of sharpness as the wide end. It looks good but not quite as stunningly good as I might have hoped. I hope to have time to do a side by side comparison at 70mm as well. I do quite a lot of shooting with the 24-70 in the studio with the lens set at 70mm so at least stopped down the sharpness is important to me.
With regards to the handling, the new lens is quite a bit larger in length and somewhat wider in thickness than the old lens; the difference in length is quite noticeable. I find the zoom and focus rings, as well as the locking mechanism of the hood to be more affirmative and tighter than with the G version. I think the new lens is likely to be a bit more mechanically sturdy than the old one but of course without opening the lens up or gathering a lot of service statistics over several years I wouldn't be able to say something definite other than my personal impression and hope.
My very early overall impression is that the wide angle performance is excellent but I need much more material of the long end to give it due credit (or not). The VR seems to work more or less and the autofocus is very fast but can sometimes catch onto something in the background that I didn't intend (likely this is not a lens problem but can be a user problem ;-) After all it is normal that the autofocus sensor positions cover a larger area than indicated in the viewfinder and this can lead to problems with wide angle focal lengths when trying to focus on a face with building lines in the background).
It is just my subjective view that the lens is probably not quite worth the purchase cost to many people - it is very expensive and the old lens is overall still a very good performer (with its own quirks). If you need top 24mm performance then for sure the new lens is better but at 70mm the images didn't cause a "wow that's impressive" reaction at least not yet. I would like to see Nikon present a 24-70mm f/4 as an alternative with lower cost, smaller size and lower weight than the f/2.8E VR unless they plan to keep the 24-70/2.8G in the lineup, which is probably the easiest way they can offer a more affordable lens to those who need a high quality wide to telephoto zoom. (I am not that impressed by the 24-120/4).
This is just a conversation opening and initial impressions (do not take it as a review), I would like to hear other people's impressions of the new lens as well, if you get the opportunity to try it out or purchase it. I think it will take at least a year before the full performance characteristics are understood.
