Evaluation of Auto focus performance among popular cameras

[Interceptor121]

I have done some research using data from imaging resource quite interesting

Autofocus Systems for Underwater Photography

 

[Compressor]

Thank you @Interceptor121.

 

[stuartv]

I don't think that article really gives the best analysis of the situation. But, maybe my experience is too limited.

In my experience, the speed of autofocus underwater is not nearly as critical as it is for shooting BIF (birds in flight). The subjects I have shot do not move nearly as fast as BIF, generally, and they don't usually appear with so little warning as you get with BIF.

The vast majority of the time, I have my focus locked onto my subject well before I actually release the shutter. And, yes, I am talking WA, as that is almost exclusively what I have been shooting.

I do agree that accuracy of focus is critical. To me, that is THE MOST important characteristic of the AF system for shooting underwater. Personally, I am just not able to manage manual focus when shooting moving subjects underwater, so I need it to be pinpoint accurate.

Also, the chart is no longer current. One of the reported improvements from the Sony a7rIII to the a7rIV is speed of AutoFocus. I have shot some BIF with my a7rIV. Some shots have been where a bird suddenly dived into the water after a fish. For those shots, I whip my camera up and shoot as it dives. The whole scene is done in less than a second. I have gotten 2 - 4 frames in that time, all in focus. So, I don't think a metric of >0.25 seconds is close to accurate for an a7rIV. But, I could be wrong, as I don't know what the test conditions were for the numbers in that chart. And, of course, what I've said is purely anecdotal.

My other anecdote on the subject is letting a friend play with my camera who shoots semi-professionally using some top shelf Canon (I can't remember which model). After a minute or two of shooting his dog running around the yard, his first comment was an emphatic, "man! This thing focuses FAST!"

Another thing that does not seem to match my experience is that you've reported that the Sony does well at focusing in low light. With mine, that is one of my biggest complaints. Without a focus light, it really seems to struggle on focusing on things underwater that look to me like they are easily lit well enough to get a focus on. But, I am talking about an rIV and you are talking about an rIII, of which I have no direct experience. So maybe that is a difference between them.

 

[Gene]

This thread title caught my eye as I often photograph people moving quickly in low-lit scenes, (on film and television productions). And to make it harder, there's usually smoke added along with some back light.
I'm a long-time Nikon guy, but have also used Fuji mirror-less cameras before Nikon finally came out with the Z7. (Which I plan on getting an underwater housing for.)
Underwater, I currently use a Nikon D800 and I usually shoot in caves, so of course that involves a lot of lights and patience. While shooting underwater, I usually focus on my subject(s) right before shooting. And we try to limit their movement. On set, that's a different story entirely! I shoot what I can, as I can and thankfully we usually do various takes but not always. There's always those moments that are the special ones though, and in my decades of shooting, I've learned that you can actually motor-drive right around that and miss it. So for years, I've been pretty successful with "feeling" the moment and capturing it by timing, shooting a single frame. (AKA, Henri Cartier Bresson's decisive moment.)
I've been shooting since I was a kid in the early 1970s, so I've seen technology come a long way. Shooting film with no auto-focus to today's seemingly unlimited number of photos you can shoot and see immediately. It's amazing really. Do I miss film? Yes, but I've been shooting digitally since 2003 and I'm pretty happy with it.
By the way, I remember Canon really pushing into the photojournalist market in the early 1990s while I was working for the Syracuse Newspapers in upstate New York. Canon made some major progress with auto-focus and metering. That was also about the time I first used Photoshop. I've been a photographer for most of my adult life and I grew up with my father being a photojournalist. I can't imagine what work I'd do or even what my life would be like if I couldn't be a photographer.

Gene Page
www.pagephotography.smugmug.com

 

[Interceptor121]

The data

I don't think that article really gives the best analysis of the situation. But, maybe my experience is too limited.

In my experience, the speed of autofocus underwater is not nearly as critical as it is for shooting BIF (birds in flight). The subjects I have shot do not move nearly as fast as BIF, generally, and they don't usually appear with so little warning as you get with BIF.

The vast majority of the time, I have my focus locked onto my subject well before I actually release the shutter. And, yes, I am talking WA, as that is almost exclusively what I have been shooting.

I do agree that accuracy of focus is critical. To me, that is THE MOST important characteristic of the AF system for shooting underwater. Personally, I am just not able to manage manual focus when shooting moving subjects underwater, so I need it to be pinpoint accurate.

Also, the chart is no longer current. One of the reported improvements from the Sony a7rIII to the a7rIV is speed of AutoFocus. I have shot some BIF with my a7rIV. Some shots have been where a bird suddenly dived into the water after a fish. For those shots, I whip my camera up and shoot as it dives. The whole scene is done in less than a second. I have gotten 2 - 4 frames in that time, all in focus. So, I don't think a metric of >0.25 seconds is close to accurate for an a7rIV. But, I could be wrong, as I don't know what the test conditions were for the numbers in that chart. And, of course, what I've said is purely anecdotal.

My other anecdote on the subject is letting a friend play with my camera who shoots semi-professionally using some top shelf Canon (I can't remember which model). After a minute or two of shooting his dog running around the yard, his first comment was an emphatic, "man! This thing focuses FAST!"

Another thing that does not seem to match my experience is that you've reported that the Sony does will at focusing in low light. With mine, that is one of my biggest complaints. Without a focus light, it really seems to struggle on focusing on things underwater that look to me like they are easily lit well enough to get a focus on. But, I am talking about an rIV and you are talking about an rIII, of which I have no direct experience. So maybe that is a difference between them.

The data is from imaging review as I wrote the low light situation low contrast is maybe the only one that is interesting the only parameter that is of some interest is the first one that is an indication of total shutter time. If you are prefocussed all camera are fast if you switch focus distance then you run into issues.
Sony cameras are very good at adjusting focus quickly but here we are talking at first time focus not predictions or burst so is a different story
As I wrote at the end there has been a lot of progress and this topic is massively overrated and the key issue is the operator not knowing how to use the tools, I wrote this article simply because I wanted to demystify it

 

[Nemrod]

Thanks to Covid 19 I have not been able to go diving. Our trip was cancelled and since Florida is closed my back up is not available either. So, I have been in a private pool playing around, now that I have both of my Inon strobes back and working. The pool was hand dug! It is 15 feet deep. Is it is not well lighted. It has no stripes, no contrasting tiles, it is pale blue. I ran into the side of the pool twice before I realized I had run out of pool, so my human eye AF was not working well in the lighting and lack of contrast! The A6400 would hunt and not give a good focus pointed at the blue pool side or bottom. I toss a penny or a dive weight on the bottom or use my lens board, the A6400 snaps into focus. Shooting up into the Snells Window it snaps into focus. But, with nothing to focus on it does hunt. If I wave my hand across in front it snaps to focus. I am set in AF-S and Prefocus Off. I am using Center focus and Center exposure. I am set for Back Button focus which I love. I do notice that my Sigma lens focuses more slowly than the two Sony lenses I have but still quickly enough. Regardless of the IR data, the newest Sony cameras have much improved AF speed. During the last two years as I sampled cameras trying to decide which one I could live with, none of them focused as fast as my new A6400 in any noticeable amount. So, I guess I do not agree with the IR data and will throw down with Sony. You know, all the data had Johnny Ringo faster than Doc Holiday. Apparently the data was wrong--- of course it is all legend, lost in the smoke and fog of time. But in this time, Olympus OMD-XYZ-@%$# my Sony A6400 will be your daisy.

 

[Interceptor121]

As I wrote the article has the purpose of removing the hype from autofocus

For the 6XXX series the concern is the popup flash once you have remote trigger is a camera like any other and it won't focus on the sky or a pool with no tiles etc etc

 

[stuartv]

There is another reason that I don't think your article gives the best analysis of the situation - a reason it does not really demystify in a way that has practical application.

AF speed is very dependent on the lens used. It is well known (in some circles, anyway) that the Sony 90mm macro lens is the best macro lens available for a Sony camera. But, it is also very slow to AF. In contrast, other lenses on a Sony achieve focus very quickly.

The article does not acknowledge this fact at all, much less caution the reader to take this into account when choosing a system. The chart that gives numbers does not say what lenses were used. Overall, it is completely misleading because the numbers could be completely different if the cameras were tested with a different set of lenses, leading to a totally different conclusion.

By not even mentioning that, it would mislead the less-knowledgeable into potentially buying into a platform that "makes sense" based on the information presented, but in reality could be the absolute worst choice for what that person wants to do.

 

[Interceptor121]

There is another reason that I don't think your article gives the best analysis of the situation - a reason it does not really demystify in a way that has practical application.

AF speed is very dependent on the lens used. It is well known (in some circles, anyway) that the Sony 90mm macro lens is the best macro lens available for a Sony camera. But, it is also very slow to AF. In contrast, other lenses on a Sony achieve focus very quickly.

The article does not acknowledge this fact at all, much less caution the reader to take this into account when choosing a system. The chart that gives numbers does not say what lenses were used. Overall, it is completely misleading because the numbers could be completely different if the cameras were tested with a different set of lenses, leading to a totally different conclusion.

By not even mentioning that, it would mislead the less-knowledgeable into potentially buying into a platform that "makes sense" based on the information presented, but in reality could be the absolute worst choice for what that person wants to do.

Sorry but now you are being picky at least there is some foundation and explanation how to read information compared to the crap that is already been written everywhere including review posted by people that sell the equipment

I always get responses from those guys whose system came worse than what they expected and then there is a start of if and buts to build an argument that what I wrote is flawed, you hear nothing from those whose camera appears to do well, wonder why

For what concerns Sony cameras they are quite popular among new comers but extremely rare within pro and people that win competitions so for what matters the marketing did not convince those guys yet. Has this got a link with what I wrote am not sure but those are the facts

If you are happy with your camera just go and get some images to proof the point that is a great camera

 

[Nemrod]

Sorry but now you are being picky at least there is some foundation and explanation how to read information compared to the crap that is already been written everywhere including review posted by people that sell the equipment

I always get responses from those guys whose system came worse than what they expected and then there is a start of if and buts to build an argument that what I wrote is flawed, you hear nothing from those whose camera appears to do well, wonder why

For what concerns Sony cameras they are quite popular among new comers but extremely rare within pro and people that win competitions so for what matters the marketing did not convince those guys yet. Has this got a link with what I wrote am not sure but those are the facts

If you are happy with your camera just go and get some images to proof the point that is a great camera

Professionals by and large are still in the dSLR camp. Since Sony does not produce a dSLR it is no surprise there and they have no legacy brand loyalty from the olden times. Olympus, not having a dSLR is a different matter because Olympus does build some cameras that can live in the real world, rugged, and the brand has strong loyalty from the film days, I owned an OM-1 and OM-2. I favor the brand.

As to Sony AF, it is discussion that will always have two side at this time of development. I like the Hybrid AF and subjectively from my limited experience, the Sony cameras, the newest iterations in particular, are subjectively as fast or faster with AF as any. Now, where it gets muddy is shot to shot and where the "shutter" lag is actually generated. For folks using the camera built in strobe, yes, that will slow the system shot to shot vs a camera using a hot shoe and external strobe (surface) or a LED trigger. The "shutter" lag is not from the AF but from the strobe recycling and from the preflash delay.

IR has some good data, it is a good resource, it is just not always complete or regimented. They are comparing apples and oranges and pears and the results are not as useful as could be hoped for. Sony claims worlds fastest AF, this seems to perturb those who do not have a Sony or have a camera with a slower AF. Nonetheless, the Sony AF is fast and getting faster and more intelligent and it is fast enough.

And frankly, for me, I would ditch the auto-everything and especially AF in a heart beat. Give me a true MF system with a split prism or course ground glass and a tactile, mechanical focus ring on the lens. The manual focus on digi-cams is a shadow of what it was on a OM-1 or Nikon F and is difficult to use with no tactile feel and no direct focus screen. So to be truthful, I do not like any of the AF systems and in particular I do not like the Olympus AF at all, again, subjective comparisons.

James