j4nu
Offline
Upload & Sell: Off
|
 Re: Sony 85mm f/1.4 GM II Announced—Pre-Orders Now Open! | |
Steve Spencer wrote:
j4nu wrote:
Steve Spencer wrote:
j4nu wrote:
Steve Spencer wrote:
j4nu wrote:
Steve Spencer wrote:
j4nu wrote:
Steve Spencer wrote:
JD07 wrote:
I'm happy to be corrected, but I thought the eye AF issues with the A7IV were supposed to be corrected by a firmware update (possibly v1.1 I think) some time ago?
Steve Spencer wrote:
j4nu wrote:
Orsonneke wrote:
I have the 85 GM ii in use since last week in Belgium.I am astonish how exact the AF is compared to the sigma 85.The 85 GM ii is exact on the eye ball , ALSO on fast moving people.Corner sharpness is better too (perhaps not relevant for portrait) and much less barrel distortion compared to the sigma.I only tested the 85 GM ii on the a9iii.I would definitely buy the lens again.I sold the sigma
I hope this doesn't mean that you need A9III to get accurate eyeAF with the GM II  ...
Do you hear any AF noise like @Donbioh@@@@@@@@@@@@ does?
Keep in mind Don was using the A7 IV which is behind other current offerings from Sony in just the area he was having problems. Here is a quote from the DPreview discussion of this camera's AF:
"However, we did find that the a7 IV can tend to very slightly front-focus when using Face/Eye detection, leaving the eye-lashes perfectly in focus but the pupil very slightly soft. This isn't going to be an issue for everyday social photography, but if you're trying to shoot portraits with shallow depth-of-field, the a7 IV doesn't seem to be quite as dependably accurate as the previous generation of Sony cameras."
It seems quite possible that he isn't seeing a better hit rate for 85 GM II over the Sigma 85 DG DN because the camera he is using can't deliver that hit rate.
It may not take an A9 III to see that hit rate either. Here is a quote from the DPreview discussion of the A7r V's AF:
"The good news is that, for the most part, you don't need to think about or edit much of this at all. Choose AF-C and a small tracking AF point and the camera will track whatever you point it at pretty well. Specify that you want it to focus on Humans (just faces and eyes, perhaps) and it'll do extremely well. In our practice focus didn't always perfectly hit the subject's pupil but it's only the camera's high pixel count that allowed us to see any imprecision: we can't think of another camera or AF method that would have delivered better results."
Of course we will have to see, but there is good reason to hope, IMO, that the 85 GM II will have great AF with the latest cameras that are capable of providing the same precision that the lens is capable of providing.
I am sure it has been improved with firmware, but the A7 IV does not have the sensor scan speed of the A1 or A9 cameras and the extra calculations that provides, nor the advance AI module of the A7r V, so the 85 GM II might well work with those more advanced cameras even though it struggles with the A7 IV. That struggle could be the camera more than the lens.
Yes, you are right, but the Sigma focuses on the eye just fine in this case so it shouldn't be the matter of the camera here.
I'm sure that A7RV will recognize the face faster due to AI and A9III will track better due to faster sensor read out, but come on - A7IV should be perfectly capable of capturing kids doing their stuff...
I also expected GM II to be excellent when it comes to AF and I am surprised reading this report...
Also, these are 2 different things:
* eye(lash)AF in static scenes shooting from an angle, etc. - this is mostly the luck of the draw in my experience
* eyeAF in dynamic scenes (esp. close rapid movement towards the camera) - this is where both the camera and the AF motors in the lens play a big role (and where 85DN struggles)...
I also read Don's comments differently. I saw him saying the new 85 GM II wasn't better at capturing kid's eyes than the Sigma DG DN and both struggled. If anything the Sigma "might" have been better, but it "might" have been luck of the draw. That sounds to me exactly what I would expect if the AF performance was limited by the camera rather than the lens, and it fits with other reports that the new 85 GM II is clearly better than the Sigma DG DN with a top level camera like the A9 III.
I think we need to await more systematic tests and that is hard with AF, but over time I still expect that at least when you really push the AF system the new 85 GM II with its 4 super fast motors moving two different groups (as I understand it) is really going to provide outstanding AF. I say that as someone who isn't going to buy the lens and who recently picked up an 85 GM I. I will be very surprised if exceptional AF isn't one of the clear advantages of the GM II over both the first version and the Sigma DG DN. One things we know for sure, both the first version of the GM and the Sigma are going to be a lot more limited in frames per second than the new lens and that will matter for some, but obviously only those who have cameras that can take advantage of that ability for higher frames per second (i.e., it won't matter on an A7 IV or A7r V that can only do 10 fps). I think we will find a similar things for general AF performance as well. The better AF will be a much better benefit on newer cameras with more advanced AF performance.
Yes, I also had the same expectations regarding AF.
... but if you need A9III to make use of it in a non-extreme scenarios like kids running around the playground, I can't say it's what I expected...
But that is just the point: there is zero evidence you need an A9 III, just evidence that an A7 IV might not be enough. An A9 III seems to be enough, but I think it is likely that all the cameras that Guy mentioned here (https://www.fredmiranda.com/forum/topic/1871097/11#16643573) will be more than enough to get the benefits of the better AF from this new 85 GM II and I won't change my expectations until I see evidence that those cameras don't gain much by using the lens.
Well, my point is that A7IV should be enough for that scenario...
It's also possible it was a defective copy, due to that audible AF sound which I think no reviewer has mentioned...
And that's where you and I differ. I am not bothered if it turns out that some of the AF advantage of the new version requires a more advanced camera to fully utilize those advantages. Obviously the number of frames per second this lens is capable of shooting is only going to be relevant for some cameras and not others. The 85 GM II is certainly capable of shooting at many more frames per second than the Sigma 85 DG DN, but you are not going to see that on an A7 IV, and I wouldn't say that isn't the way it is supposed to be. You will see that difference in capability on any A9 series camera, and the A1, but in this case you won't see it on the A7r V either.
In a similar way I would not be surprised if some of the other speed related aspects of AF are only going to be relevant for some cameras and not others. I expect that some of the AF advantages of this new lens will require a camera with more AF capabilities and even the A9 III might not be able to take full advantage of everything this lens can do speed wise. That might await future cameras and that would not be a deficiency of the lens.
I think the point I wanted to get across is that the biggest difference between the DN and GM II seemed to be AF accuracy in action. That is why I'd get GM II over DN, even if I had to pay for that with 1cm longer lens (and a lot more $$$ of course).
Now, if you need (for a lens that doubles as a portrait lens) A9III or A1 (I intentionally don't include A7RV/ACR here as these have slower readout speeds than A7IV and AI helps in subjects recognition, which wasn't the problem in this scenario) to see any improvement in that area in a not-that-demanding scenario (no BIFs, etc.), then a lot of people are going to think twice about such "upgrade"...
I don't know. I would expect that almost anyone who wants the lens for increased AF accuracy in action is already going to have a camera that can handle fast action (an A9 series or A1 camera) and would be unlikely to have an A7 IV, and of course almost nobody would plan to use an 85mm lens for BIF, but more demanding situations like sports photographers, or perhaps in very low light like wedding photographers shooting dancers at a wedding reception, we might expect the AF advantages of this lens to emerge. I wouldn't expect them shootings kids outdoors at a park. Sure the extra fps of something like an A9 III or even an A1 might allow some differences to emerge, but even AI isn't going to help all that much as I doubt we will be getting a seesaw AF mode any time soon 
I would also agree with Don that the Sigma 85 DN does really well for AF accuracy for the types of things he was shooting, so expecting better probably doesn't make much sense, IMO. If you are going to spend the substantial money this lens costs, then it makes sense that you plan to use the lens in a way that requires its capabilities. I think that for what Don described as his shooting, the lens isn't going to get him much advantage over the Sigma. I don't think I would use the lens in a way that takes advantage of its capabilities either. That's why I am not getting the lens. I would use the lens primarily for portraits and for that I don't need what the GM II provides and I actually prefer the GM I. If I shot much sports, however, or if I shot landscapes (where the edge to edge sharpness of this lens might matter), or even if I shot weddings or events I would be really interested in this lens. I think you see those varied interests reflect the different enthusiasm for this lens and that is how it should be.
Well, try shooting wide open a kid running (or swinging) towards you not too far away and you'll see there's plenty to be improved in the Sigma . This is really not a complicated motion but somehow Sigma's AF can't keep up with properly updating the distance...
Ok, I see others don't like this discussion so I'm stopping on my part at least .
|
