LensRentals.com

Autofocus Reality Part 2: One vs. Two, Old vs. New

Posted by

The first part of this series confirmed what we already new: Live view focus is more accurate than phase detection AF.

This part is going to (sort of) look at something I've wondered about along with some related old wives' tales. And unlike the first article, at the end of this we'll have learned something today.

One Click or Two?

That something I'm wondering about is this: How much of a feedback loop is there between the camera and the lens? It might be that the system has totally one-sided communication:

Camera: OK, I see you're focused over here, I'm sending you a signal to focus over there. Then I'll take the picture.

Or there could be a feedback loop of rechecking going on like this:

Camera: I see you're focused over here. I'll send you a signal to focus over there. 

Lens: OK, I'm where you told me to go.

Camera: Well, not quite. You're still off a bit. I'll send you a signal to go a little bit this way. 

Lens: OK, I'm there. You should have gotten this right the first time. 

Camera: No, you didn't move where I told you to. We're really close now, just a bit this way. 

Lens: Who put you in charge anyway? OK, I'm there. 

Camera: Yes, that's perfect. I'll take the picture. And next time just do what I tell you the first time, OK?

Lens: Why don't you get things right the first time? I moved where you told me. You're using those off-center sensors again, aren't you? Like those things ever work! 

We're really not certain which way the phase detection AF systems in various cameras work. Most likely it's something in between those two examples. But it might be different with different bodies or different lenses. Some of the lenses in current use were designed 20 years ago and autofocus has been making a lot of advances since then.

If you believe scenario No. 1, AF may be more accurate if you "double focus", meaning you push the shutter button halfway down until the AF beeps, then release and push it halfway down again. The idea is that you're providing the camera a "recheck"of the AF point and a chance to fine tune focus.

I was taught to do this when I started photography but I have no idea if it really helps. So I thought we'd look at that using the same lenses we did in part one, the Canon 50mm f/1.4.

If the two-click AF method works better than one-click AF, that might give us some indication that the system is open and without feedback. Maybe. If it isn't I'm not sure it means there is a feedback loop. Maybe AF is as accurate as it gets no matter how many times you pre-focus.

The Setup

The last post showed that LiveView Contrast-Detection autofocus and my manual focus were about equal with the 50mm f/1.4 but that phase-detection AF was more random.

Using the same Imatest setup as before, we tried comparing sets of phase-detection AF: one group where the shutter button was halfway pushed once and another group where the shutter button was pushed down several times (with focus confirm beep after each) and then the shot taken.

Even before the numbers were processed I was pretty sure I knew what the answer was going to be. In watching the distance scale as I pushed down the AF button halfway and held it, I could see on some shots a quick, second movement of the scale before the focus confirm beep occurred.

I was pretty sure this indicated the lens and camera were talking to each other, meaning there was a feedback loop going on. The results shown below certainly demonstrate that pushing the button several times was no better than pushing it once. First is a typical 50mm f/1.4 on 5D Mk II graph (we did several and all were about the same).

We tried the same lenses on a 5D Mk III to see if the newer focus system made any difference. We didn't expect it would on single-point center AF with a still target in great light. However, the 5D III AF system is much better than the II. But this testing situation isn't meant to show off advanced AF features, just basic center point accuracy.

 

This one looks a little worse at a glance because this lens-camera combination could have used a bit of microfocus adjustment for backfocus, but certainly the single focus / multiple focus shots are not different.

Again we did this several times and the standard deviation of the results was the same for each camera with no difference between multiple focus and single focus.

What does it mean? Well, I can quit pushing the button multiple times to focus. I believe there is a feedback loop that the camera is repeatedly checking focus as things get close, but that's just what I believe. I don't think this proves it.

Old Versus New

We've been comparing an old and new lens: the 50mm f/1.4 was released in 1993, the 50mm f/1.2 in 2007. But some recent releases allow us the chance to compare older and newer.

The Canon 24mm f/2.8 lens was released in 1988 making it one of the oldest designed Canon lenses available, while the Canon 24mm f/2.8 IS USM was just released this year. So basically we set up to do the same type of tests comparing those two lenses.

First we put the old version on a 5D Mk II. Note that the scales are different. Since we're now shooting f/2.8 lenses the resolution is higher. Depth of field is also a bit greater but we assumed we'd have little trouble seeing changes in sharpness from focusing, and that was the case.

The only real difference is both manual and live view focus are a bit less accurate than with the 50mm lens. The lens has some spherical aberration and I certainly realized I was having a bit of difficulty focusing myself.

Apparently the camera did, too. Phase-detection AF is about the same as with the 50mm lenses, but it appears better largely because LiveView and manual focusing are worse.

Next, of course, we tried the new 24mm f/2.8 IS USM lens on the 5D Mk II.

 Both contrast (LiveView) autofocus and I did much better with this lens. Camera autofocus did not. Again, this lens is backfocusing just slightly, which is why the average resolution is a bit lower.

Repeating the test with different copies of the 24mm and 5D Mk II didn't change things at all. Neither did shooting the old version of the 24mm on the 5D Mk III with its new autofocus system.

Just to make everything complete, we mounted the new 24mm f/2.8 IS USM on the 5D Mk III and repeated the test. You've looked at the graphs, one after another, all the same.

I gathered the data not only from the graphs I've shown but from other copies we tested. So at this point, you can imagine how shocked I was when something different happened. So I repeated it with different copies of the 24mm IS USM and 5D Mk III and got the same results.

 It appears, and I've repeated the test on several copies, that the combination of new camera AF system in the 5D III and new lens makes a huge difference.

In this case, phase-detection AF is just as accurate as LiveView. I didn't put the numbers up to keep the post reasonably brief, but manual focusing and single-shot AF were just about the same for three different copies of the lens.

Something seems to be different with the 5D III and 24mm f/2.8 IS USM combination. The 5D III isn't better with the old lens. The 24mm f/2.8 IS USM isn't better with the 5D II camera. The two together, though, are better.

So What Does This Mean?

I'm not sure. It certainly seems that there are some camera-lens combinations that have more accurate phase-detection AF than others---when using center-point AF on a test target. It may actually be a big deal.

A given manufacturer is always going to say "new, more accurate autofocus" with a new model of camera. They don't have much motivation to add things like "on some lenses."

We reproduced the finding with a different 5D III and a couple of other copies of the 24mm lens, but perhaps there's some other explanation I haven't thought of.

I'll feel more certain if some of the other new lenses do the same thing on this camera, and we'll get that tested right away. I think the 28mm f/2.8 IS should certainly behave this way, and possibly the 40mm f/2.8.

Since we've only seen this on the one lens-camera combination I'd rate it as "probably, not certain" in accuracy. Since I had planned this as a series of posts, I'm posting them as I do them.

So you get to follow along, and we'll see if I've discovered something new or if there's just an odd red herring in my autofocus findings.

 Addendum

I was supposed to be doing constructive work today, but I couldn't stand not confirming what I saw above.

Although we'd done three copies of the 24mm lens on different 5D III copies, I wanted some more confirmation. It seemed logical that if what we were seeing was real, then the new 28mm f/2.8 IS would also AF more accurately on the 5D Mk III.

I was able to repeat the test this morning on three copies of the 28mm lens on a 5D III. I've printed two examples below with the axis expanded to emphasize differences. As you can see below the 28mm, just like the 24mm, is as accurate with phase-detection AF as it is with LiveView AF.

Obviously our next step will be running through the rest of the newer Canon lenses to determine which other ones have increased accuracy.

 

Roger Cicala

Lensrentals.com

July 2012

 

28 Responses to “Autofocus Reality Part 2: One vs. Two, Old vs. New”

MCO_970 said:

Really interesting results with the 24 & 5D3. I wonder if the same AF improvement pattern holds true for other newly released lenses, for ex. the new super-teles and the upcoming 24-70.

Stv said:

Old lenses just do what the camera says. There is no re-check. So either camera and lens get it right the first time or not.

Newer lenses (and like you found out newer autofocus systems) give distance information and also give feedback about actual movement. So there is a feedback loop.

This applies to Canon. Dunno about the other manufacturers.

lexvo said:

This could be a very interesting find. You're right that it should be backed up by testing with other lenses.

But it may very well be the case that Canon implemented an improved AF algorithm on new cameras and lenses.

Jesse said:

I would be curious how much of a difference continuous focus vs. single shot makes for static subjects...

Jason said:

The reason for two click AF wasn't for accuracy as much as it was for speed. If you want to take a picture and need to wait for AF, you can miss a moment. Also it can be more accurate if you are pressing down faster to take a full picture instead of a small press for half click. That would then give you more camera shake. I'm glad you tested, but to me a tripod mounted scenario misses the point of why two click is better. If I'm taking a landscape on a tripod, I do single click (with remote cable). But if I'm catching a child's face in a moment, I double click just so I can catch the moment instead of waiting for AF and missing that fraction of a second.

mshafik said:

Thanks for the research, but what about AI Servo, I use back button focusing, and I always used AI Servo with my 60D and 5D Mark II with great accuracy, and I usually release the shutter button while still pressing the focusing point.

Can you compare that as well?

Arun said:

How about the 5d3 with the new super-teles?

Ken Owen said:

The body and lens synergy conundrum is getting exciting!

Mark Marshall said:

This is great information! When I self-tested my 5DII and 50D a few years back I found with the 50mm f1.4 focus accuracy was awful. Focus success was only about 50% with the 5DII. With the 50D it was much improved but not great at around 75%. This accuracy made micro-lens adjustments very difficult to determine. The 50mm was my worst lens of all that I tested but in most cases the 50D did better than the 5DII. I got rid of the 50mm&50D, now have the 7D&5DII. I have found the 7D is immensely better at focus accuracy as compared to the 5DII. I find myself shooting the 85f1.2II wide open with 7D but am forced to stop it way down with the 5DII. As an example I find focus accuracy on the 7D is good with the 70-200f2.8II and worse with the 35mmF1.4. Also using an external flash seems to make an enormous improvement in accuracy, I assume since it lays out a grid of lines. Your testing is really helping me understand this very frustrating issue. I’m really looking forward to more test lenses and as compared to Nikon and others. The focus issues have been my single negative issue with my Canon equipment. Thanks so much for the info!

A said:

Take home: If you want to get the very best performance from your new camera, buy some new lenses to go with it.

Roger: Have you tried asking Chuck Westfall about this, and anything else Canon? He's always been helpful and informative in the past!

Wilba said:

Thanks for providing good data to support what we already know about this.

We know quite a lot about how PD AF works, see http://www.dpreview.com/articles/5402438893/busted-the-myth-of-open-loop-phase-detection-autofocus. The bottom line is, focus is confirmed when the AF sensor sees an in-focus subject.

Feedback from the lens is irrelevant to focus accuracy. It doesn't matter whether the lens was told to go to the right place, or whether it did, or whether it told the body that it did... it only matters what the AF sensor sees. The critical closed-loop is the feedback from the AF sensor.

One thing missing from your results is the difference between the first confirmation and the second. As far as I can tell (with my limited ability to accurately measure focus), even if the focus ring and/or distance scale might twitch on the second half-press, focus ends up at the same place as it does on the first. I'd be very interested to see what you can measure to confirm or deny that.

Stv - try the tests in the linked article.

Jason - yes, pre-focusing does set you up for a quick final focus, but rapid double pressing is claimed to improve accuracy and this data disproves _that_.

A - Chuck Westfall (and RDKirk) have so far refused to discuss evidence which disproves their claims.

Sakari said:

This is really good data! I wonder if it would be possible to get more varied tests? Seeing the difference between 1-series and 5-series would be interesting since I still prefer my 1Ds mk2 over 5dmk2 when i need to shoot bigger than f4. In my experience 1Ds mk2 rarely misses, even with 50/1.4. It's not perfect but far more trusty.

Jan said:

How about nikon? I heard nikon owners boasting about their superior autofocus.

fuc nem truong said:

too long dont reead like internet. death

RDKirk said:

Sheesh, why did my name get pulled into this? Wilba, dude, you're harping on a post that's a full decade old! And that was based on Canon documents even older. How long is that in digital camera lifetimes?

As I said in the very first line of that old, old post: "According to the major material in Canon's 'Lens Work III,' the description in their US patent application, and remarks by Chuck Westfall, to put it briefly:"

If Canon lied in their own documents...oh, well.

Wilba said:

RDKirk - Hello! Your name got pulled into this because it is frequently used, as an "appeal to authority", to support the open-loop myth.

I've asked you directly several times for comment on my findings and you've not done so, which makes it look like you're sticking by what you said over a decade ago.

I'd really like to hear what you think now that we can easily prove what's going on. I'd love to include a statement from you in the article about your results and conclusions from doing the tests. The world deserves to know! :-)

Happy to take a discussion offline if you like. Thanks, W

Dan said:

I wonder how accurate the phase detection AF systems are across all of the cameras newer than the 5Dii. For example, does the 7D's phase detection produce results sililar to the 5Dii, or to the 5Diii in your tests. There may be a magic date where all new camera bodies and new lenses released produce the high-quality results when used together.

Body: 1D-iv, 1Ds-iii, 7D, 60D, T3i, T4i, etc...?
Lens: The new super-telephoto ii lenses, 70-200 f/2.8L IS ii, 100 f/2.8L IS macro, etc...?

This kind of information will be extremely useful for considering all future lens purchase decisions, especially for sports photography. I'm certain that Canon marketing isn't very interested in clearly stating this list. It could result in a dramatic drop in sales of the older lenses if they are commonly known to be "not as good" for auto-focus. Of course for most situations, 9 out of 10 pictures will have no noticeable effect due to the reduced phase detection focus accuracy.

LensRentals Employee

Roger Cicala said:

Dan,

I hope to get exactly that information, as well as which lenses behave which ways. It's going to take several weeks to run all those tests, but I'll post as soon as I can.

Roger

Samuel Hurtado said:

Dear Roger:
You're expanding the realms of open human knowledge.
Just wanted to make sure you were aware of that.
Thanks a lot.

Lynn Allan said:

Informative articles ... thanks.

Something I've wondered about with my Canon 52m2: with LiveView, you can focus as "not zoomed 1x", 5x zoomed, and 10x zoomed. With a flat subject, is there a difference in focus accuracy between these options? If I've got a static subject and sufficient time, I tend to focus at 10x, but I'm unclear if there is an advantage to this.

With a non-flat subject, my impression is that the 10x zoom choice would allow me to be more selective on just where the focus is intended to be.

Spencer said:

Outstanding. Your business is the perfect testing grounds for this kind of work and I really appreciate the extra effort (read: information) that you provide. Thank you tons for sharing. I'd love to see a test along the lines that you are performing with the body manufacturer lenses with a popular third party lens (Sigma EX 50, perhaps) to see just how much precision might be lost when using a third party lens.

Aside: I know that the S50 is schizophrenic with some (most?) Canon bodies, but I have had mine remapped by Sigma with my 5D Mk III and now my mid and infinity focus is good, with only some front focusing at short distances. This goes to show how important that lookup table in the lenses firmware is to proper operation.

LensRentals Employee

Roger Cicala said:

Hi Lynn,

My impression is the same as yours: 10X, given enough time, allows the most critically accurate or selective focus. Personally I can't focus accurately at 1X, but 5X is a reasonable compromise for when I need orientation within the shot.

When we're doing critical testing, like with Imatest, only 10X gives us a fighting chance at perfect focus and even then we 'focus bracket' or take the best of several results depending upon the setup.

Best,

Roger

Elliotte Rusty Harold said:

Very interesting. I'm eager to see the results you get with the various supertelephotos and different bodies. I rarely shoot anything that needs such tight focus with short lenses, but if the new cameras and superteles have measurable autofocus improvements like these, then that could end up costing me a lot of money. :-)

I have heard anecdotal evidence from other bird photographers that the 5D Mark III and the the 300mm f/2.8L IS II outperform their previous rigs for autofocus, even when using a 2X TC III; but this is more convincing. I don't think anyone's had a new 500mm or 600mm lens for long enough to draw any conclusions on that.

focus said:

It would be interesting to see the EXIF focus distance versus Imatest.

Peter said:

Ill be really interested to see the results with the 70-200L IS II. Awesome test though, thanks.

ProfHankD said:

I have the A-mount Sigma 8-16mm, which I use on Sony A55, NEX-5, and NEX-7 (with LA-EA1 or LA-EA2 on NEX). Interestingly different focus behavior. It works on the A55, slowly on the NEX-5 + LA-EA1, but not at all on NEX-7. It looks like NEX-7 expects lenses to go where they're told, and the (old firmware) NEX-5 checks. No problem using screw-drive lenses on the NEX-7 + LA-EA2. I contacted both Sigma and Sony about this; Sigma basically said the lens works as intended, Sony said the Sigma isn't a Sony lens so it isn't their problem. ;)

Have you seen this with the 8-16mm or other non-screw-driven A-mounts on NEX?

(Incidentally, I still love the 8-16mm -- using manual focus. Harder to use well than my old 10-20mm, but worth the effort.)

Iorik said:

Thank you for such a refreshing and interesting articles. Maybe i didn't understood everything as a professional will, but... when i press half way the shutter button, to check focus,(i usually press once, check the aperture and speed, and press it again) the focus points always change. So, I have ALWAYS to press several times, till the focus points are on the correct place. Dont know if this makes any sense to you? Thank you again.

Iorik said:

HO, forgot to add that i use a 5DMKII and always felt sharpening/focusing issues with it. but always thought it was me... Never tried to shot in LiveView mode (despite the convenience of having the mirror lockup), because was warned (by the Canon Manual for instance) of overheating issues. Isn't that a problem?

Leave a Reply