Industry News: Sony officially announces the α9 II

Del Paso

M3 Singlestroke
Aug 9, 2018
684
708
I think you use lag to describe the time it takes between something happing in reality and it being displayed on the display.

Assuming there is no initial delay for starting the imaging pipeline and display, this depends only on how long it takes to process and display a single frame.

In that case, higher frame rates are the way to reduce lag, not increase it.

The faster you can process the image, the faster you can display it, given an appropriate display. And that reduces lag. At 60 Hz for example, each image is displayed for 16.7 ms and the processing time can not be higher than that. At 120 Hz, each image is displayed only 8.3 ms, but processing must be twice as fast to allow such a frame rate.

Reducing lag and perceived blur this way does make tracking objects considerably easier though. I use a gaming display with a 144 Hz refresh rate on my computer and it is a night and day difference between that and 60 Hz.

The EOS R viewfinder has a resolution of only 1.280 x 960 pixels. It only does 120Hz 720p video, (1280 x 720 pixels) but I'm wondering: do we know the refresh rate of the R? I had assumed 60 Hz because that's what the LV display is, but the viewfinder spec isn't mentioned on Canon's site I believe.

In any case the M6 II does 120 Hz 1080p (1920 x 1080 pixels) video so it should have the processing power to drive a high res, high refresh viewfinder. A future high res R may give us just that.
Isn't the the resolution you named the resolution of the back-screen?
The R's viewfinder has a resolution of approximately 3,9 MP.
 

amorse

EOS 7D MK II
Jan 26, 2017
600
652
www.instagram.com
Thom Hogan, who is a pro sports photographer and writes sensible, informative articles has written a positive response http://www.sansmirror.com/newsviews/2019-mirrorless-camera/october-december-2019-newsv/sony-a9-mark-ii-announced.html in contrast to 15 pages from angry Sony addicts on Fredmiranda. But, there is a real reason for the difference in opinion. He has written in the past how it is crucial for him to send off small files quickly, and this is different from most Sony A9 users. They tend to be bird and nature photographers who love the AF of the A9 but are frustrated by its low resolution and were led by the rumour sites to expect a 36 Mpx sensor.

It is amusing to read their comments castigating Sony and thinking of moving to Canon - it's the mirrorless mirroring the mirrored!
That makes a lot of sense. I have no horse in this race, but it seemed to me that the a9ii was what Sony needed (though it is not a product I'd be in the market for so take all this with a grain of salt). Sony has been constantly lambasted for poor ergonomics and their weather sealing, so targeting both of those things seems like the right choice. Maybe Sony has come to the conclusion that if they want to see a mass transition to the a9 from other flagship bodies, they need to cater to the needs of larger businesses who buy a lot of cameras (i.e. connectivity/reliability improvements) and that means focusing on utility-type upgrades and less flashy or newsworthy upgrades.

I mean, at the end of the day sports cars are incredible machines that create a ton of desire, but mini vans get a lot of tasks done without hassle. Maybe Sony thinks they need a bit more utility and a little less flash, and this is the result of that decision.
 

criscokkat

EOS RP
Sep 26, 2017
313
286
Madison, WI
It'll be funny if after all this gnashing of teeth if the next top of the line sports camera war is a 24mp A9II and Nikon d6 vs a 28-30 mp Canon. Canon just showed a major breakthrough in sensor readout speed with their new aps-c cameras. If they apply that tech to a larger but less dense sensor there's no reason they shouldn't be able to hit the metrics those cameras have and still be an extra 20-30% more MP for the occasional crop. I think that's important because Sony has had a history of releasing new cameras on a rather random schedule. I suspect that the A9 II is an attempt to try to capture as many pro's as possible while the iron is hot, before professional Nikon and Canon sports mirrorless cameras come out. To do this, they had to change things outside the sensor more than in. Rather than wait until next year when they have new sensors, they did it this year. The olympics are next year and now that their lens selection is fully fleshed out they want to win as many pro converts as possible.

I suspect we'll see an a9III before fall of 2021 with a newer, larger sensor. Canon's mirrorless sports camera will probably use the same sensor as the 1dxiii, so they will most likely need that extra mp to be in the same range.
 
  • Like
Reactions: SteveC

mpmark

EOS 80D
Aug 9, 2016
113
125
With video. This appears to be real.


Same sensor, same 20 fps electronic shutter; mechanical shutter now 10 fps

- A
What Sony fail to mention is that its actually only 12 FPS in uncompressed RAW, 20 Compressed RAW is misleading, The 1DXII is actually faster. I dont like this misleading advertising as everywhere you read it says 20FPS RAW, not entirely true.

1DXii 14FPS RAW and 16FPS RAW Mirror lockout
A9 and A9II - 12FPS RAW (20 Compressed RAW) or 10FPS RAW with shutter.

Who the heck spends that much money to shoot at 20FPS compressed RAW file and lose dynamic range and other factors, I want the full RAW file all the time! I'd never shoot with anything less. The older Canon 1Dxii is still faster!
 
Last edited:

Joules

EOS 7D MK II
Jul 16, 2017
588
543
Hamburg, Germany
Isn't the the resolution you named the resolution of the back-screen?
The R's viewfinder has a resolution of approximately 3,9 MP.
Canon likes to state their resolution in dots, instead of pixels. Each Pixel is made out of 3 dots. Quote from Canon on the R viewfinder resolution:

"3.69 Million dots (1280x960)"

The first figure is meaningless to me. When I think about resolution, I think about pixels, which can display each color. Not individually colored dots.
 

Graphic.Artifacts

EOS 7D MK II
Aug 1, 2017
522
356
I think you use lag to describe the time it takes between something happing in reality and it being displayed on the display.

Assuming there is no initial delay for starting the imaging pipeline and display, this depends only on how long it takes to process and display a single frame.

In that case, higher frame rates are the way to reduce lag, not increase it.

The faster you can process the image, the faster you can display it, given an appropriate display. And that reduces lag. At 60 Hz for example, each image is displayed for 16.7 ms and the processing time can not be higher than that. At 120 Hz, each image is displayed only 8.3 ms, but processing must be twice as fast to allow such a frame rate.

Reducing lag and perceived blur this way does make tracking objects considerably easier though. I use a gaming display with a 144 Hz refresh rate on my computer and it is a night and day difference between that and 60 Hz.

The EOS R viewfinder has a resolution of only 1.280 x 960 pixels. It only does 120Hz 720p video, (1280 x 720 pixels) but I'm wondering: do we know the refresh rate of the R? I had assumed 60 Hz because that's what the LV display is, but the viewfinder spec isn't mentioned on Canon's site I believe.

In any case the M6 II does 120 Hz 1080p (1920 x 1080 pixels) video so it should have the processing power to drive a high res, high refresh viewfinder. A future high res R may give us just that.
I don't think the refresh rate of the EVF and the sample rate of the sensor are the same thing or even directly related. My TV redraws the screen at 120 HZ but most of the time it's just redrawing the same image several times in a row as there aren't 120 images per second of content. 24 FPS movies are shown at the cinema at 72 or 96 HZ by drawing the same image 3 or 4 times. Gaming is different because you are instructing the computer to generate content at a higher rate. A camera can't really do that. Or it won't just because you use a display with a higher refresh rate. The higher refresh rate of 120 or 240 HZ EVF is simply to avoid flicker and more closely approximate the experience of an optical viewfinder.

The real time lag is caused by the camera having to read the entire sensor, demosaic the image, process it , down-sample and then present it to the viewfinder in a form it can display. A measurable amount of time will have passed while it's accomplishing this. Anything that adds to that process increases the time lag between when an event is actually happening and when it gets displayed in the viewfinder. Greater lag makes it more difficult to track moving targets. In my experience that is the biggest problem with EVFs.

As a demonstration of this. Put your camera on a tripod with the rear panel display on. Step back a little bit and then do a couple of quick pans. You will notice that the image on the rear display noticeably lags behind where the lens is currently pointed. With your eye up to the EVF this same things happens but since you are looking through the EFV and can't see what is happening in real time it is harder to perceive. Admittedly the lag is vastly improved over a few years ago but if you are trying to capture the right moment in an action sequence it can be a big deal if you are used to an OVF that functions at the speed of light.
 
Last edited:

maxfactor9933

EOS M50
Apr 18, 2018
43
28
Malaysia
I was expecting to see a new and more ergonomic form factor. it could improve the handling once a 600 mm is attached and better heat dissipation allow them to crank up the FPS
 

mpmark

EOS 80D
Aug 9, 2016
113
125
Sony is slowing down or it's just a trick? Anyways, Canon will need to be faster than a ray of light to catch up. C'mon, Canon, do something besides ridiculous amount of EOS Rebel 90000s and G5 Mark 349s.
what exactly do they need to catch up with? The 1dxii actaully shoots faster then the new a9, its marketing, look up the "compressed raw crap". It only shoots up to 12fps uncompressed raw, 3 year old 1dxii shoots up to 16 uncompressed raw.
 

privatebydesign

Would you take advice from a cartoons stuffed toy?
Jan 29, 2011
8,166
1,514
119
what exactly do they need to catch up with? The 1dxii actaully shoots faster then the new a9, its marketing, look up the "compressed raw crap". It only shoots up to 12fps uncompressed raw, 3 year old 1dxii shoots up to 16 uncompressed raw.
It also has a mechanical shutter and full AF tracking at 14fps, and 4K 60p, the Sony can’t manage either.
 
The A9II is certainly a disappointment to wildlife/bird photographers who crave more MPs and were hoping for 30+ while still maintaining 20FPS (or more). But Sony is dead set on the sports/photojournalist market (just read the press release from the A9 and A9II). The changes in the A9II seems like an obvious list of features Sony would have gotten from pros already shooting their system and polling pros still shooting CaNikon.

Sony usually throws giant press events for even a single lens release or their dinky APS-C bodies. They didn't do a single thing other than their press release and some phone calls to the YouTube guys/gals for the A9II. Sony knows this isn't a typical type of Mark II release. They probably shouldn't have used the II moniker but in the end it really doesn't matter one way or the other.

I was one waiting for a higher MP A9II....I had a standing preorder in with my dealer before even seeing the real specs. That preorder is cancelled for now but I haven't totally written off the A9II.

There are a couple things buried in the press that the A9II improved upon other than the ergo, weather sealing, better card slots and other connectivity items.....those relate to Sony updating the processor in the camera. The new processor is claimed to improve AF tracking (yet to be seen if it is significant) but more importantly it is claimed to decrease the EVF lag and most importantly (to me having used the A9 for over a year now) they can now use the full EVF resolution in the 120Hz refresh mode. The EVF picture in the A9 at 120Hz is pretty lame but 60Hz mode is too difficult to follow fast moving objects. Being able to use the full 3M dot EVF in the fast mode will be a big difference. Many people are upset the A9II didn't get the 5M dot EVF from the A7RIV but it seems that there would be no way to make use of that resolution when at 120Hz and feeding 20FPS shooting through the EVF.

It is because of these EVF changes in addition to the ergonomic changes (which I've tried now with the A7RIV and have put the cameras at an equal ergo level to the Z7 I used) that keep me slightly interested in an A9II.
 
  • Like
Reactions: pj1974
what exactly do they need to catch up with? The 1dxii actaully shoots faster then the new a9, its marketing, look up the "compressed raw crap". It only shoots up to 12fps uncompressed raw, 3 year old 1dxii shoots up to 16 uncompressed raw.
True, but who shoots these types of bodies (whether for sports or wildlife) at ISO 100-200 anyways. Extreme level testing by Jim Kasson shows no compromise to the IQ of the Sony compressed RAW above ISO 200.

What they need to catch up on is an AF system. I've owned and used both cameras (A9 and 1DXII) for years and anyone claiming the 1DXII system is competitive with the A9 is too far entrenched in the Canon fanboy cult that they can't extricate themselves and likely never will.

The A9II is a hard sell to me at $6K CDN....spending $7.5K CDN on a 1 series DSLR in 2020 is a virtually impossible sale. I'm much more interested in a Canon 1DX MILC but I think we will all have to be very patient....maybe Winter Olympics in China??

If Canon release a 1DX MILC competitive with what the A9 is and the 600 f/4 DO RF, I will be back in Canon land faster than you can blink an eye....until then I will soldier on with my kaleidoscope of gear brands....
 
It also has a mechanical shutter and full AF tracking at 14fps, and 4K 60p, the Sony can’t manage either.
After shooting 20FPS e-shutter, blackout free for the past 1.5 years alongside my 1DX2, D500, D850, I can't imagine why anyone would care what a mechanical shutter can do unless you do use flash for your work. Granted every other MILC out there is garbage for e-shutter work so certainly when considering an EOS R or Z7 or A7RIV it should be an important factor but the A9 is different.
 

Graphic.Artifacts

EOS 7D MK II
Aug 1, 2017
522
356
I don't have much experience with electronic shutters for stills. Information on electronic shutter speeds for stills isn't easy to come by but this topic interested me so I've done a bit of research. Electroniic shutter speeds are the time it takes the electronic (rolling) shutter to read the entire sensor. These numbers aren't perfect but so far they are the best I could do.

Canon R --- 80 ms
Canon M6 Mark II --- 46 ms
FujiFilm X-T3 --- 20 ms
Sony A9 --- 6 ms

Smaller numbers would seem to be better.
80 ms is approximately 1/12th of a second.
6 ms is approximately 1/150th of a second.

Canon appears to be making progress on this but they probably have some work to do before they can make a mirrorless that can compete with the IDX's using an electronic shutter. I'm not sure what the upper limit for mechanical shutter would be without a mirror. It's possible that Canon could do a 20 FPS mechanical shutter sans/mirror but presumably you would have to blackout the image while the shutter is closed.

edit: Hard to put a number on the 1DX Mark II because all of the tests are for 4K video. 4K video tests put the 1DX Mark II at about 15 ms which sounds about right. Rolling shutter for video in the 1DX Mark II is much better than any other Canon DSLR/Hybrid that I've worked with any very similar to the X-T3.
 
Last edited:

YuengLinger

EOS 5D MK IV
Dec 20, 2012
2,792
992
Southeastern USA
I, for one, wish Sony would once and for all make that perfect camera. Why? Because Sony fans don't come here to gloat, but to discharge their envy and disappointment. If Sony did make a camera that large numbers of working professionals and other actual photographers (not smartphone dreamers who just think about photography) decided to buy, then Sony fans would have no more reason to troll around.

Keep trying, Sony!
 
  • Like
Reactions: SteveC

Graphic.Artifacts

EOS 7D MK II
Aug 1, 2017
522
356
What Sony fail to mention is that its actually only 12 FPS in uncompressed RAW, 20 Compressed RAW is misleading, The 1DXII is actually faster. I dont like this misleading advertising as everywhere you read it says 20FPS RAW, not entirely true.

1DXii 14FPS RAW and 16FPS RAW Mirror lockout
A9 and A9II - 12FPS RAW (20 Compressed RAW) or 10FPS RAW with shutter.

Who the heck spends that much money to shoot at 20FPS compressed RAW file and lose dynamic range and other factors, I want the full RAW file all the time! I'd never shoot with anything less. The older Canon 1Dxii is still faster!
at 20 fps my IDX Mark II does 8 bit jpg's. Actually I guess that's 24 to 60 FPS and it's cropped to 4K. Sounds like somebody is getting their talking points from DPR's comment section. My understanding is that the electronic shutter in these cameras works pretty well and Canon is going to have to work pretty hard to catch up. Assuming they don't go with mechanical shutters or somehow manage to leap ahead with a global shutter.
 
Last edited:

AlanF

Canon 5DSR II
Aug 16, 2012
6,230
4,114
True, but who shoots these types of bodies (whether for sports or wildlife) at ISO 100-200 anyways. Extreme level testing by Jim Kasson shows no compromise to the IQ of the Sony compressed RAW above ISO 200.

What they need to catch up on is an AF system. I've owned and used both cameras (A9 and 1DXII) for years and anyone claiming the 1DXII system is competitive with the A9 is too far entrenched in the Canon fanboy cult that they can't extricate themselves and likely never will.

The A9II is a hard sell to me at $6K CDN....spending $7.5K CDN on a 1 series DSLR in 2020 is a virtually impossible sale. I'm much more interested in a Canon 1DX MILC but I think we will all have to be very patient....maybe Winter Olympics in China??

If Canon release a 1DX MILC competitive with what the A9 is and the 600 f/4 DO RF, I will be back in Canon land faster than you can blink an eye....until then I will soldier on with my kaleidoscope of gear brands....
How are you finding the A7RIV for AF?
 

3kramd5

EOS 5D MK IV
Mar 2, 2012
3,083
404
Interesting that Sony can still only achieve 10fps with a mechanical shutter even with no mirror, Nikon and Canon have bested that with a mirror and full AF tracking functionality for years.
I doubt that’s all they *can* achieve, but they seem to want to feature electronic shutters. I’d wager they’re using the same purchased Nidec shutter in this as they used in A7Riii, not a new Sony design.
 

navastronia

EOS RP + 5D Classic
Aug 31, 2018
379
418
I don't have much experience with electronic shutters for stills. Information on electronic shutter speeds for stills isn't easy to come by but this topic interested me so I've done a bit of research. Electroniic shutter speeds are the time it takes the electronic (rolling) shutter to read the entire sensor. These numbers aren't perfect but so far they are the best I could do.

Canon R --- 80 ms
Canon M6 Mark II --- 46 ms
FujiFilm X-T3 --- 20 ms
Sony A9 --- 6 ms

Smaller numbers would seem to be better.
80 ms is approximately 1/12th of a second.
6 ms is approximately 1/150th of a second.

Canon appears to be making progress on this but they probably have some work to do before they can make a mirrorless that can compete with the IDX's using an electronic shutter. I'm not sure what the upper limit for mechanical shutter would be without a mirror. It's possible that Canon could do a 20 FPS mechanical shutter sans/mirror but presumably you would have to blackout the image while the shutter is closed.

edit: Hard to put a number on the 1DX Mark II because all of the tests are for 4K video. 4K video tests put the 1DX Mark II at about 15 ms which sounds about right. Rolling shutter for video in the 1DX Mark II is much better than any other Canon DSLR/Hybrid that I've worked with any very similar to the X-T3.
This subject also interests me. I assume you've also come across Jim Kasson's blog where he tests cameras' e-shutter readout speeds? Because of its e-shutter performance (and because it's less of an investment than an a9), I have contemplated getting an X-T3, though ideally, I would love to purchase a Canon system with an e-shutter speed fast enough for normal, non-action use without artifacts. Shooting stills on movie sets (where being silent is mandatory) without needing a blimp, that's the dream for me.
 

Kit.

EOR R
Apr 25, 2011
1,662
1,005
I don't have much experience with electronic shutters for stills. Information on electronic shutter speeds for stills isn't easy to come by but this topic interested me so I've done a bit of research. Electroniic shutter speeds are the time it takes the electronic (rolling) shutter to read the entire sensor. These numbers aren't perfect but so far they are the best I could do.

Canon R --- 80 ms
...
80 ms is approximately 1/12th of a second.
It's not what is seen in its EVF, whose panning lag I could not distinguish from the panning lag of A9's EVF.

By the way, A9's IBIS goes haywire with slow or moderate panning, both horizontally and vertically. Hopefully they fixed that in A9II, otherwise for most sports its IBIS needs to be turned off.
 

Aussie shooter

@brett.guy.photography
Dec 6, 2016
599
692
As a bonafide sony hater I have to say well done on this camera. They seem to be learning that there is more to a camera than a self pleasuring spec sheet. Still a long way to go but maybe they are finally heading in te right direction.