Sony a 9

[arthurking83]

1. My experience with the Sony lenses is that IQ is not an issue. As I don't have the lens I can't be sure, but I expect that the intent is that it will have very good IQ.
1b. Shutter speed. Also depends on ISO, so if the Sony has better noise reduction, then ...

(all other settings to be static) It's a 99.9% sure thing that a lens set to wide open isn't going to produce as good IQ as a lens stopped down one or two clicks!

2. I don't think your pet hates are relevant. Most reviewers (and myself) think that being able to see a good image is far more important than having a great image that you can't see. I have got used to the EVF and I find it to more more than adequate, particularly in low light where it is demonstratively better than optical.

Fair enough .. it was just a side note and an opinion, but the argument still holds that due to the way viewfinders work, with an EVF image brightness depends more so on maximum aperture set, than do SLRs(in std configurations).
I'm not a fan of std configs, and I prefer a faster(f/2.8 brightness) screen matte. So my fast lenses show bright images. But when I mount slower lenses(eg. f/5.6 or slower) the hit to image brightness is both magnified, and subsequently more obvious.
My pet hate(ie. opinion) was more so for the purpose of highlighting the fact that at f/5.6, the EVF needs to boost it's ISO level to compensate for the two stops of light loss through the aperture(compared to f/2.8). That means grain(my opinion centred around the graininess in these situations).

... I don't know how many pro sports shooters also take video, but if they do it becomes a no brainer. If you have ever shot video on a DSLR you will know that trying to follow a sports person with a screen is almost impossible. But with an EVF it is realistic. Add the other video features and you really do have a winner.

I do suspect that sports photographers are like the rest of us in that they don't want to change unless there is a compelling reason, but given what Sony are doing, I think they will start to change. Maybe it will wait until Sony release a long f2.8, but maybe it will be sooner. We should pick a big sporting event next year and survey the photographers.

This is what may cause a gradual evolution of change in the industry(of CaNikon don't step up to the mark). Japan already has 8K capable broadcasting ability(not used much, but a glimpse into the future).
CaNikon are both stuck in the dark ages of HD video recording. Looking into the immediate future, 4K is becoming old hat already(just as it's starting to take off! ) and CaNikon(more so Nikon than Canon) are dragging their feet.

I reckon what will happen in pro terms in many situatons.
They'll stick with tried and tested CaNikon gear for stills, and have access to an A9 plus (say) a 100-400 for video if needed .. or something to similar effect.

ps. clarification of 4K becoming 'old hat' ... while Canon and Nikon both do 4K video, they still stuck in the lower quality modes, and low frame rates. 4K and 60fps should be the minimum feature/ability nowadays. I don't know the Canon's specs, but Nikon's D5 is stuck at 4K and 30fps(and probably a lowly bit rate at that too!).
D5's biggest issue is the recording limit tho.

If the A9 gains a foothold, it'll almost certainly be due to their video capability more than anything else.

Sony 400/2.8 is unlikely in the next 6 months, as they have nothing to base it on .. so it'll be a start from scratch endeavour.
300/2.8 would be a different mater tho. Like Swifty said, focusing mechanism would be their main engineering challenge, and big lens elements need high torque, this also implies high power usage. The A9 is already going to be battery limited due to the EVF, so to multiply this effect with a lens that strains the power system even more is their main challenge.
The question is, if the their 300/2.8 A-mount has the optics base to allow easy transference of an acceptable focusing mechanism for the A9.

400/2.8 if we see one within 2 years will be a miracle (if they haven't started the process yet). 300/2.8 if easily adaptable, in 6 months shouldn't be hard.
Next Olympics(2020) is where we will definitely see how serious Sony is about this A9 market segment .. and how long CaNikon have to up their respective antes to minimise any impact the A9 has on their turf.

Other things to note about the A9 too:
Traditionally, Nikon have a preference to use Sony sensors, so they're more than likely to be in a position soon to use a similar/same sensor .. for example in a D5s. It's the Sensor that's allowing those elevated performance abilities.
So there's no reason to eliminate the possibility of a 20fps capable D5s, with the same video features too in a year or so.
In a sense, Nikon almost has a bit of an advantage over Canon here using this historical perspective as a guide.
Canon need to start from scratch to achieve the same impact. New sensor, and all the other engineering distractions that the top end product entails.

[Steve Axford]

One of the best things about an EVF is that you can use it as night vision goggles. I can see in a dark forest at night with an EVF. I can't see anything with an optical viewfinder and precious little with my naked eyes. You really should try that, it's quite an eye opener.

I don't really disagree with most of what you say. None of us have a crystal ball, so we all watch with interest. As for the Nikon (sony) sensors, I think that Sony have taken the line that Nikon will pay a premium for the new sensors if they let Nikon have them at all. That could prove to be very difficult for Nikon if Sony want to take their market share. Time will tell and I suspect that Nikon will find a way to survive irrespective of whether Sony let them have their sensors or not. Anyway, the new sensor is designed to be an EVF sensor, so Nikon would have to change to mirrorless anyway, or keep the mirror just for the opticl viewfinder. I really do think that the A9 will be another nail in the DSLR coffin. They have been great cameras, but what's the point if the focusing is better done on sensor?

[swifty]

A few observations from looking up a lot of 400/2.8 images. It seems night sports are where the 400's are being used wide open mainly. DOF appears to be adequate but I have no idea how much they're cropping in. I guess sitting on the sidelines, you're position limited so you crop as required and the distance to where the action is may vary considerably.
But in day events, there appears to be more stopping down which of course makes sense as adequate shutter speeds are maintained even at lowish ISO settings.
For wildlife, it appears most are stopping down a stop or so from max for the smaller creatures, not sure whether for DOF or the increase in acuity but probably both. But I'm seeing many paired with TC's too so maybe 400's a bit short for birding and other small creatures?
Didn't seem to see many big game types of wildlife photography but I'd imagine shooting around dusk or dawn when the big predators hunt might see shutter speed limitations so perhaps f2.8 becomes more handy. But maybe 400 isn't quite right for big game either?

But I did read elsewhere the 400/f2.8, along with 70-200/f2.8 and 200-400/f4 are crucial pro sport lenses. So if Sony are serious about pro sports, maybe the 400/2.8 or 200-400/f4 are the tele's being worked on.

But to add a comment regarding the sensor scenario. If it is in relation to the interview with the Sony Imaging manager, then you have to understand that's exactly what it means. Sony imaging will keep sensors they develop (and pay for) for themselves. Nikon and others aren't buying from Sony imaging but from Sony semiconductor.
It is also not so clear cut which sensor patents are Sony semiconductor, which are Sony imaging, which are Nikon's, which are licenced from other third parties etc.

Either way, the A9 has a heck of a fast sensor. It appears to have some similarities to the E-M1 Mk II sensor which is also a Sony and also very fast too.
The sensor readout speed gains bodes very well for mirrorless bodies of the future.

[Steve Axford]

Interesting! This is what I would expect and leaves things open as to the uptake of the A9 for sports. I would guess that in the future the A9 will play a big part, but until then we wait and see, though perhaps some will jump sooner than others.

[MissionMan]

A few observations from looking up a lot of 400/2.8 images. It seems night sports are where the 400's are being used wide open mainly. DOF appears to be adequate but I have no idea how much they're cropping in. I guess sitting on the sidelines, you're position limited so you crop as required and the distance to where the action is may vary considerably.
But in day events, there appears to be more stopping down which of course makes sense as adequate shutter speeds are maintained even at lowish ISO settings.
For wildlife, it appears most are stopping down a stop or so from max for the smaller creatures, not sure whether for DOF or the increase in acuity but probably both. But I'm seeing many paired with TC's too so maybe 400's a bit short for birding and other small creatures?
Didn't seem to see many big game types of wildlife photography but I'd imagine shooting around dusk or dawn when the big predators hunt might see shutter speed limitations so perhaps f2.8 becomes more handy. But maybe 400 isn't quite right for big game either?

But I did read elsewhere the 400/f2.8, along with 70-200/f2.8 and 200-400/f4 are crucial pro sport lenses. So if Sony are serious about pro sports, maybe the 400/2.8 or 200-400/f4 are the tele's being worked on.

But to add a comment regarding the sensor scenario. If it is in relation to the interview with the Sony Imaging manager, then you have to understand that's exactly what it means. Sony imaging will keep sensors they develop (and pay for) for themselves. Nikon and others aren't buying from Sony imaging but from Sony semiconductor.
It is also not so clear cut which sensor patents are Sony semiconductor, which are Sony imaging, which are Nikon's, which are licenced from other third parties etc.

Either way, the A9 has a heck of a fast sensor. It appears to have some similarities to the E-M1 Mk II sensor which is also a Sony and also very fast too.
The sensor readout speed gains bodes very well for mirrorless bodies of the future.

The stopping down is probably also to do with playing within the lenses peak sharpness. At f5.6 the f2.8 would be tack sharp.

[Steve Axford]

I really doubt that, MM. A good 2.8 will be easily sharp enough at 2.8.

[MissionMan]

I really doubt that, MM. A good 2.8 will be easily sharp enough at 2.8.

Sharp enough but still not at it's sharpest. Sharpest is normally 2-4 stop lower than maximum. I.e. if you don't have to use it at f/2.8, don't because f/5.6 will be sharper

[Steve Axford]

If I buy a 400/f2.9 for $15,000 I would expect that I would never have to stop it down for sharpness. For dof yes, but not for sharpness. I have top quality lenses that are supposedly sharper at f4 or f5.6, but I can't tell the difference, even with pixel peeping.

[swifty]

The stopping down is probably also to do with playing within the lenses peak sharpness. At f5.6 the f2.8 would be tack sharp.

Yes, I would expect that but my experience with the best lenses I've got to try are that they peak much closer to wide open.
Sorry folks for continuing to deviate a bit OT but this is quite interesting. I haven't had the privilege of experiencing using one of these exotics but a quick google search reveals some pretty interesting behaviour.
Eg. According to Nassim's figures: https://photographylife.com/reviews/...00mm-f2-8e-vr/
Nikon's latest version equal peaks at f4 and f5.6. So unless you needed the extra DOF, there would be no reason to stop down past one stop.
No wonder you see so many shots at f4.
With TC 1.4 and TC 2 it's peaking around 1 stop down too whilst with the TC1.7 it peaks 2 stops down.

Whats very interesting is the Nikon 800/5,6 peaks wide open!!! What the... probably something to do with TC's and f8 focusing I'm guessing.

- - - Updated - - -

I have top quality lenses that are supposedly sharper at f4 or f5.6, but I can't tell the difference, even with pixel peeping.

You need an A9r with 100MP

[Steve Axford]

That would be nice, but would make dof even more challenging. I did read something recently where a reviewer tested the peak sharpness of a variety of good lenses and found that the peak was usually f4 to f5.6. I did read that the sharpness degrades after f8 for my 21mm loxia. I can't see it, so I will sometimes use f16 with that lens. I think that measured sharpness is interesting but you need to test if it makes a noticeable difference. Often, it doesn't.

[MissionMan]

That would be nice, but would make dof even more challenging. I did read something recently where a reviewer tested the peak sharpness of a variety of good lenses and found that the peak was usually f4 to f5.6. I did read that the sharpness degrades after f8 for my 21mm loxia. I can't see it, so I will sometimes use f16 with that lens. I think that measured sharpness is interesting but you need to test if it makes a noticeable difference. Often, it doesn't.

I think the degradation levels are normally dependent on the lens quality. Sharpness wide open and degradation on a $200 lens vs a $2000 vs a $15000 lens will be vastly different, largely because the sharpness wide open on each of those will be different. The cheap lens will obviously increase substantially whilst the expensive lens is pretty sharp to start with. The lens sharpness doesn't normally degrade much beyond wide open at f/11 so the difference would be barely noriceable on a good lens.

[Steve Axford]

Most labs report that sharpness declines as you reach high fstops. Most say that f16 is too high with an A7R2 and you will get noticable degredation due to diffraction effects. They tell you f11 is the highest you should go (if you want perfect sharpness). I have tested this and I cannot tell the difference up to f16. F18, however is too far, though I use up to f22 for 4k video. You really don't want to believe all the test results without verifying it yourself.

[arthurking83]

Most labs report that sharpness declines as you reach high fstops. Most say that f16 is too high with an A7R2 and you will get noticable degradation due to diffraction effects. They tell you f11 is the highest you should go (if you want perfect sharpness). I have tested this and I cannot tell the difference up to f16. .....

On the whole, most of those types of testing results ... ie. that I've seen and verified with my own equipment .. seem to be accurate.

as an example: TDP - Canon 400 f/2.8 image quality comparison at f/4 vs f/11

The above link should(hopefully) take you to the The Digital Picture website to compare the differences in IQ at f/4(about max IQ) and f/11(where IQ is visibly degraded) for the Canon 400/2.8 IS II lens.
Hove the mouse over the image to show the same resolution chart at f/11, then mouse off to show the same image at f/4.
At f/11 there is very visible IQ degradation, and it really starts to manifest at f/8!!(where you'd expect maximum detail rendering)
Camera used is the 5Dsr, so the issue of diffraction limitation is at the top end of the spectrum.

For the vast majority of the lenses I have, I've read the test results, and pretty much experience the same results.

What's more important tho is what is the desired result you're after. More DOF in a single shot(I'm not overly keen on focus stacking) or maximum detail rendering for an isolated area of the subject.

But on this topic, the single most important element in the entire equation is reproduction ratio(or viewing size).
The reproduction ratio you intend to view the image with determines how much IQ loss you can bear to see.
If you view the image zoomed in at 100% pixel level, then diffraction effects are more noticeable, then say if you view the image as a whole, where the pixel zoom level is going to be more like 20%.

To see the difference of how this works, use Cntrl - in your browser to zoom that webpage out(to say 25% .. Cntrl - three times) and the effect of diffraction, whilst still there, doesn't look as bad as it did at 100% view.

video, even at 4K is pretty low res stuff(at these sensor sizes), and the issue of diffraction limitation is all about resolution.
So while 4K is renown for being 'hi resolution' .. this is a relative term, and only with respect to past video specification.

[Steve Axford]

What a cool site. Pity it doesn't do Sony gear.
I would suggest that you have used the wrong camera. If you change the input to Canon 1DS MkIII (instead of the Canon 5Ds R), then I can't see any change. Canon do not sell the 5Ds R as a wildlife or sports camera and the 400/f2.8 II isn't designed for that camera.
Unfortunately that site doesn't have any Sony lenses or cameras, so we can't see what they would get for those. All I can say is that I have tested my Sony 90mm G OSS macro lens and the Zeiss 21mm loxia, and I can see only the slightest of differences. But, both lenses are the best I have ever used in their class.

Of course 4k video is low res compared to an 8k still frame. That was my point when I said I could use a higher fstop..

[Lance B]

last time I checked mirrorless was still growing and DSLR was still declining and Sony knocked Nikon for the number two slot, even if it was only for a couple of months. That's the big impact and I think Canikon have every reason to be worried. This, if it actually works has the potential to take a lot of business. Even traditionally pro DSLR people have been impressed with the specs thus far.

As for the pros switching, Fuji has gained a fair amount of market share already in wedding photography where long lens like the 400 f/2.8 aren't required. How many pros need long lens? Wedding ohotographers? No. Portrait? No Studio? No. Landscape? no. It's only Wildlife and sports that require it.

Actually, mirrorless has flatlined for the last 3 years:

http://promuser.com/markets/dslr-vs-...t-share-report

Nikon and Canon need not be concerned.

- - - Updated - - -

Sharp enough but still not at it's sharpest. Sharpest is normally 2-4 stop lower than maximum. I.e. if you don't have to use it at f/2.8, don't because f/5.6 will be sharper

Actually, no. My 400 f2.8E FL VR is super sharp wide open and peak sharpness is at f4, 1 stop down. I regularly shoot at f4 and will shoot wider if necessary. In fact, it is just as sharp at f2.8 as it is at f5.6:

https://www.ephotozine.com/article/n...s-review-26266

Over at Lenscore, the Nikon 400 f2.8E FL VR is the sharpest of all the super teles and the 6th sharpest of any of the lenses tested and 2nd of any lens for overall score.

http://www.lenscore.org/

f3.2:

[arthurking83]

.....
I would suggest that you have used the wrong camera. If you change the input to Canon 1DS MkIII (instead of the Canon 5Ds R), then I can't see any change. Canon do not sell the 5Ds R as a wildlife or sports camera .....

Yep! totally got that, but the reason I used the higher res camera was to highlight the issue.
That's why I made the comment with regards to display purposes, ie. output size and acceptable level of IQ loss and then the point to 'zoom out'(which is akin to using a lower res sensor by way of comparison)

The effect is there; it's simply a matter of how acceptable is one aspect of performance balanced against other sets of performance targets.
ie. as you said earlier for your situation .. do you forgo ultimate sharpness over DOF ... or as in some situations I've had to deal with, chromatic aberration manifests badly on the Nikon 105VR Micro lens at approx f/5.6 - f/8 .. close to the ideal sharpness aperture too

I doubt very much that diffraction limitation is going to be an issue on the A9 tho. It is a 'low res' sensor by way of comparison to a few higher res sensors for the same format nowadays.

[Steve Axford]

Diffraction can always be an issue, it just depends on what fstop it starts to be noticeable. My macros lenses go to f24 and that is almost always a problem, though on some occasions it may be worth it. Of course, some lenses only go to f16 and then it is rarely a problem (cept for the MPE 60mm where f16 is really f64 at 5x magnification)
.

[MissionMan]

Actually, mirrorless has flatlined for the last 3 years:

http://promuser.com/markets/dslr-vs-...t-share-report

Nikon and Canon need not be concerned.

- - - Updated - - -



Actually, no. My 400 f2.8E FL VR is super sharp wide open and peak sharpness is at f4, 1 stop down. I regularly shoot at f4 and will shoot wider if necessary. In fact, it is just as sharp at f2.8 as it is at f5.6:

https://www.ephotozine.com/article/n...s-review-26266

Over at Lenscore, the Nikon 400 f2.8E FL VR is the sharpest of all the super teles and the 6th sharpest of any of the lenses tested and 2nd of any lens for overall score.

http://www.lenscore.org/

f3.2:

http://www.pbase.com/lance_b/image/1...1/original.jpg

mirrorless has flatlined whilst their share of the market is increasing steadily. I'd say they have every reason to be concerned. They saying it will surpass 30% by the end of this year. I don't think the switchers market is that big, id say it's more entrants to the market and that's not just cameras then, it's glass as well. Even Thom Hogan who is a massive Nikon fanboy has expressed concern in this space. The challenge as you know is when people pick a system, it's incredibly difficult to get them off unless there is something missing and not that many people need or can afford a 400 f2.8. The more likely scenario is a 100-400 which most mirrorless systems have.

[Steve Axford]

I've had to deal with, chromatic aberration manifests badly on the Nikon 105VR Micro lens at approx f/5.6 - f/8 .. close to the ideal sharpness aperture too

That surprises me. Mind you, I would very rarely use my macro lens at f5.6 to f8 so I would never notice if it was less than perfect at those apertures. I would assume that the design is focused on high fstops, since it is a macro lens. Also, does this vary with the distance from the subject, ie magnification? I find it curious to see the Lenscore site rating the Nikon macro as being better than the Sony on colour and similar on chromatic aberration though less on resolution. Perhaps this highlights that you cannot just compare lenses with a standard formula. You need to take the use into account and macros generally have a very specific uses.

[arthurking83]

On a side note to the topic of the thread:

That surprises me. Mind you, I would very rarely use my macro lens at f5.6 to f8 so I would never notice if it was less than perfect at those apertures. I would assume that the design is focused on high fstops, since it is a macro lens. Also, does this vary with the distance from the subject, ie magnification? ...

As far as I've noticed, no difference with respect to magnification nor focused distance ... it's just plain annoying across the range(when it rears it ugly head).

here's a quote from Photozone on CA issues with the 105VR.

"Chromatic aberrations (color shadows at harsh contrast transitions) range from 1 to almost 1.5 pixels on the image borders, slowly increasing by stopping down."

CA on their graphed chart increases as aperture is stopped down. So, like you said, being a macro lens, you'd think that it's most commonly used aperture would be macro-ish values .. f/11, or f/16(despite any diffraction effects), but CA gets worse as you stop down!

Most lens designs(if not all that I've ever seen in tests!) have a reducing CA level as aperture is closed up .. but this one works the other way around.
Strange thing about the lens tho, is that it's still a great lens. works well and all that(focus a little iffy-slow) but it renders beautifully, and while I have plans to replace it as my macro lens of choice, I still intend to keep it too.

On a more topical level:
I can't ever recall a 400/2.8 lens test where CA could be an issue in any real world sense, but CA can be a common issue with 100-400 type lenses from all manufacturers(if the photographer is pedantic about image quality).
Sure CA can be processed out more easily with software nowadays, but that's not really the 'same thing' as getting it right in camera is it?
In fact as I recall, Sony have a no so raw raw file, that is always preprocessed in some way by the camera, even takign into account the use of thirdparty software to process those images. Some thirdparty raw converters(eg. RawTherapee and other DCRaw based software) do ignore the Sony in camera raw processing routines tho, so you get to see the real quality ability of the lens when those types of raw converters are used!
(as I remember too tho, Olympus and Panasonic also have a similarly processed raw file going into most 'popular' raw converters .. so the issue is not unique to Sony.