hyken wrote:
Well, maybe some of us bought into the marketing and reviewers! If the AF was competent, who wouldn't want more MP. Most if not all stated that the AF was very close to the a1ii but with that caveat of it only processing 60 cal/sec. Maybe that was there out. I really think they mostly showed the tack images that I got yesterday because there were many and they were superb, it's just the 40% or so that are not usable because of misses.
I have been shooting Sony professional bodies and glass for many years now and always been disappointed with the R series when it came to AF. I blame myself for trying the latest iteration. The a1ii is a beast, a9iii is crazy fast in every way and my go to in studio. ...Show more →
Don't blame the marketing and reviewers - I never saw any that suggested this was a replacement for an A1ii/A9iii for serious action shooters. My takeaway was that it was likely good enough for those with "occasional use" needs only and where its primary purpose would be for high res/high DR situations.
Sony's own marketing showed a bee-eater with warped wings - BIG RED FLAG right there
And Sony claimed 5.6x faster readout speed than the A7R5 and only 60 af calculations per second - neither of which are in the action camera territory of the A1ii/A9iii.
For wildlife I don't expect the same hit rate as the A1ii or A9iii - if I need critical focus I will use the A1ii or A9iii every time.
Having said that my experience to date is that the A7RM6 seems to be at least as good as the A1ii with nailing focus - but we haven't really had good enough conditions to properly test that. I would be very surprised if that turns out to be the case - perhaps with large slow subjects it might be close.
For reference here is a link to my test of the a6400 on a whippet using mechanical shutter at 7fps (for live view evf for following the action). He's a lot older now so I am not sure he has any sprints left in him for testing the A7RM6 but I'll ask...
Hey we know it can shoot swallows IF. So what else do we need? That is the benchmark all cameras should be judged by. I remember when I first got the Z9 and tried swallows and just wanted to toss that camera in the lake
Who is blaming anybody, I'm just asking a question of other shooters who might actually own the body. Maybe there are some who are using it that might be seeing some of the same issues. If not, maybe I have a lemon or bad copy. This post was not at all about sensor readout speed, which isn't that bad to be honest. It was about AF potential issues. The only comparison with the a1ii is that they both show the AF red/white dot on the eyes but only one only actually maintains tack focus for the most part, not at ~40% or so...at least with my early testing. In my brief hands on, this body is no where close to either a1.
jtra wrote:
Do you have a raw format set to "compressed" (implies 12 bit processing in continuous drive)? In other modes and 30 fps the AF system will not be able to do 60 calculations per second on a7rvi because 14 bit processing is too slow for it.
hyken wrote:
Don't start blasting please, I did some simple tests on day one to compare to my a1ii and a9iii with same glass, 300mm f/2.8GM and 100-400 f/4.5. Same settings across the board. I'm not gonna post images because after shooting about 1000 images of my golden retriever running full speed directly at me chasing a ball, the same thing was happening consistently. Shooting wide open with both lenses.
The 300 was slightly better regarding hit rate but even with this superb AF giant, only about 60%, the 100-400 was about 40%. When I say this, I was hoping for tack sharp eyelashes/eyes. 1/2500 on all tests.
When the a7r6 was missing, it was always front focusing. Her tongue would be tack sharp consistently even though the AF spot was locked on her eyes.
Is anyone seeing this with super fast glass and action? Maybe I'm so spoiled by the a9iii and a9ii, I was expecting more?
I tried with pre-capture on/off because if you have pre-capture on, you can't enable 'priority set to release' to "AF".
Also, paired with my a9iii, the 100-400 f/4.5 rocks! Not perfect but pretty darn good, probably about 90% with fast action...maybe 95%.
Sure I'm forgetting something but those are my initial thoughts after receiving the camera today, 6/23/26.
Here is my experience with A7R6 with bird subject ( have not tried photographing my great pyrenees sprinting yet )
discliamer first : Of course any camera can produce jaw dropping tack sharp images, in the right hands, especially after post processing and posting small size on an Internet forum. But that tells us nothing about the keeper ratio or comparative performance of a given camera rather it is a testament to the photographer's technique and experience.
with that out of the way, I think the A7R6 AF is pretty capable but it is NOT at the same level of A1 II or A9 III in certain situation. That situation includes when the bird is coming at you at high speed or taking off from a perch. the A7R6 only does 60 AF cycles per second (calculate AF position and drive the lens 60 times per second) , the A1 series and A9 III do 120 cycles a second, and there us no arguing with hard math.
When the bird is flying parallel to the back of the camera or at a shallow angle 120 cycle / second is overkill as the focus distance does not change rapidly, the A7R6 is similar to the A1 in these conditions in my experience. But when bird is coming right at you the focus distance is changing rapidly and when you use a shallow DOF lens like the 300 f/2.8 you will see a difference in the percentage of tack sharp images. I would say in my experience 70% tack sharp vs. 90% sounds about right...of course I have made and can post many sharp incoming images with both cameras but the keeper ratio is not the same. I think this is expected from the 2x disparity in the spec's as well. But for flying parallel I would say there should not be much of a difference in keeper ratio oif your settings are correct and the right technique is used (stable handhold , correct AF pattern etc.)
this experience resembles yours with the golden retriever to some degree at least
Maybe 1/2000 / 1/2500 isn't enough with these large files. Gonna try really pushing the shutter tomorrow. Of course, that kind of defeats my use case in that I shoot a lot of action sports under the lights and I'm not talking about pro arena's. If I'm having to push the shutter, my ISO will really suffer. Anyway, on to more testing this weekend with soccer, daytime.
We'll see how the camera handles humans running towards the shooting position. Thank you!
hyken wrote:
Maybe 1/2000 / 1/2500 isn't enough with these large files. Gonna try really pushing the shutter tomorrow. Of course, that kind of defeats my use case in that I shoot a lot of action sports under the lights and I'm not talking about pro arena's. If I'm having to push the shutter, my ISO will really suffer. Anyway, on to more testing this weekend with soccer, daytime.
We'll see how the camera handles humans running towards the shooting position. Thank you!
I recommend using at least 1/3200 sec, I would also use non tracking zone pattern with subject recognition set to animal .... this settings seemed to do the best for me at the dog park
Here are some running dog shots taken just now - 1/2500 is too slow, a lot of motion blur is visible.
Not much to complain about the AF accuracy though - EyeAF boxes I have on a separate video if anyone wants to see them I can post that video. Basically it is 99%+ and the misses are ones at the beginning and right at the end of the sequence which is to be expected.
Dog Eye AF - WARNING HEAVY WITH IMAGES SO YOUR BROWSER MAY TAKE SOME TIME TO LOAD
There is a link at the bottom of each page to the next page.
Not sure what anyone else thinks but this looks about as good as it gets. I don't think my A9iii would do any better. To be fair I wasn't shooting at f/2.8 which might pose more of a challenge.
Maybe I will do that tomorrow - if Harley the Whippet will oblige. He doesn't have quite the enthusiasm he used to have I can see him thinking "Aw shit, he's got a camera, has he forgotten I am 77(in dog years) !"
I downsampled to 6K - 10K is just f'n ridiculous! Gonna break AWS Cloud Flare!
Yeah, that's not going to answer the question of focus accuracy.
duncangr wrote:
Here are some running dog shots taken just now - 1/2500 is too slow, a lot of motion blur is visible.
Not much to complain about the AF accuracy though - EyeAF boxes I have on a separate video if anyone wants to see them I can post that video. Basically it is 99%+ and the misses are ones at the beginning and right at the end of the sequence which is to be expected.
Dog Eye AF - WARNING HEAVY WITH IMAGES SO YOUR BROWSER MAY TAKE SOME TIME TO LOAD
There is a link at the bottom of each page to the next page.
Not sure what anyone else thinks but this looks about as good as it gets. I don't think my A9iii would do any better. To be fair I wasn't shooting at f/2.8 which might pose more of a challenge.
Maybe I will do that tomorrow - if Harley the Whippet will oblige. He doesn't have quite the enthusiasm he used to have I can see him thinking "Aw shit, he's got a camera, has he forgotten I am 77(in dog years) !"
I downsampled to 6K - 10K is just f'n ridiculous! Gonna break AWS Cloud Flare!
Harley and friend with the 300 +1.4 (420 f/4). Cropped to the head, not downsampled or anything.
A total of around 230 frames with just a few included from two of the three sequences. Manually modify the URL to get other frames in the sequences (2976-3030, 3047-3129, 3271-3362)
Eye detect is basically 100% - there are a few misses when things get toothy with the black greyhound (second link) about 8 out of the 90 frames.
If you're not happy with this performance from the highest resolution, highest DR FF camera you seriously need to go see a doctor.
I take it back, maybe it can't do swallows IF. Those all seem to be lacking detail even though it did technically focus on the swallow for most of them.
Maybe it can do perpendicular swallows only
Jun 25, 2026 at 06:30 AM
Steve Spencer Offline • • • • • • • Upload & Sell: On
arbitrage wrote:
I take it back, maybe it can't do swallows IF. Those all seem to be lacking detail even though it did technically focus on the swallow for most of them.
Maybe it can do perpendicular swallows only
From the shots from folks so far including you and Duncan, it looks to me like the camera, "can" do swallows in flight. The problem is that at times it will mess them up in a couple of ways. First, as in these shots sometimes the detail will be lacking. Second, at times the the wings will be distorted, as they clearly were in a few of Duncan's shots he linked. Third, at times objects in the background will be tilted as you quickly pan. So, "can" it do swallows in flight? Yes, and you might even get some good shots with no problems. Is it reliable for swallows in flight? Not really, there are some real potential problems that very well may show up.
So can you use this camera and get some good shots even on really tough subjects with lots of movement? Sure, and especailly so if you are willing to slow down the frames per second and use the mechanical shutter if all else fails. If you shoot these things regularly, however, it seems to me it is worthwhile for most people to get a camera that allows them to use the electronic shutter with a faster sensor scan speed and more reliable results.
I'm not sure if you were being serious or not, but I'll explain. After looking at your dog images, I see you were at f9 and it looks like about 257mm. I'm just guessing here, but say you were 30 feet from the dog, my depth of field calculator puts your area of focus at about 1.7 feet. This would probably put the dogs entire head in focus. At 2.8, the are of focus is about 1/2 foot or 6". If you were closer to 50 feet, the f9 gets you almost 5 feet of focus, whereas 2.8 gets you 1.5 feet or so. So getting the dogs eyes in focus at f9 should be much easier than at f2.8. The margin of error at f9 means you can't really ascertain how accurate the focus on the moving target is compared to 2.8. For focus accuracy, I like to minimize my DOF as much as possible.
