How good is the 7D autofocus system?

[xopher]

I was recently talking to a professional photographer friend about D800 vs 5D3 (both of which he uses extensively) and he strongly believes the 5D3 new autofocusing system is MUCH more accurate at low light in comparison its Nikon counterpart. He also told me about how Canon skin tones tend to better.. etc.. list goes on.. By the end I think im going to join the Canon side pretty soon.

Coming from a Nikon background, I'm not too familiar with anything Canon. Shortly after, I picked up my friend's 7D and i noticed the autofocusing was just bang on fast and amazing! Keep in mind I know the focusing is highly dependant on the lens too, i truly believe the 7D autofocusing system is solid.

Based on the rumours, it seems like the new 6D will inherit the 7D autofocusing system and the D600 will inherit the D7000 autofocusing system. I have already used the D7000 extensively in low light and it does its job decently, however, after picking up the 7D, I'm really unsure which one has the edge over the other.

A lot of the photography I do is done low light areas which is why this accuracy and speed in lightly lit areas will be the deciding factor. Can one of you more experience canon guys give me an unbiased opinion on the AF system in comparison to the D7000? Does the 7D low light accuracy resemble the godly 5D3 system I've heard about?

Thanks in advance!

 

[bdunbar79]

The 7D's AF system is really good. I don't know if it's any worse than the 5D Mark III's in low light, but it is really good. I don't think it is desirous though, using any camera, to need rapid AF in low light. The lowest light I go in is in a gym or night football and usually I wouldn't consider that as "low light" regarding AF, considering the 7D and 5D3 will do well in both situations.

 

[RLPhoto]

The 7D's AF is brilliant, best in class. The 5D3 is even better than the 7D's AF, which isn't a slouch. 8)

 

[justsomedude]

A lot of the photography I do is done low light areas which is why this accuracy and speed in lightly lit areas will be the deciding factor. Can one of you more experience canon guys give me an unbiased opinion on the AF system in comparison to the D7000? Does the 7D low light accuracy resemble the godly 5D3 system I've heard about?

Thanks in advance!

Not even close.

I own both cameras, and in my experience the 7D doesn't even come close to the 5D3's AF accuracy. Don't get me wrong, the 7D is a strong performer, but for my low light event photography, the 5D3 is my go-to.

Before I got my 5D3 I used my 7D in plenty of low-light event shoots, however, the keeper rate was less than ideal. When I couldn't use a speedlite for an AF beam, and had to rely solely on the sensor/lens, the 7D left a lot to be desired. At least in my opinion. Missed focus was a constant issue (check some of my old 2011 posts for rantings on the 7D's AF), and I ended up throwing away more shots than I ever had even on my 40D. With that said, the 7D is still my backup, but 10 times out of 10 I'll grab my 5D3 first for low-light shoots where no flash or AF beam is allowed. It's just that good.

Also - I hate to stir the pot, but in order to be fair I feel I must disclose all sides... hop on Google and search for "7D AF issue." Yes, there are some people who get a little crazy when discussing the 7D's AF performance, but there is definitely a consistent group of people who feel the 7D's AF is shoddy. Info here, and here.

If I were you I'd rent both, and judge for yourself. But if low-light AF is your thing, there's really no discussion as far as I'm concerned. The 5D3 is your only answer.

Also, check the post just a couple topics down from this one if you want to review some of the science behind both AF systems. According to Roger Cicala, the standard deviation performance of the 7D was the worst of all Canon bodies he tested (tied with the T3i). As for the 5D3... well... it was the best: http://www.lensrentals.com/blog/2012/08/autofocus-reality-part-3b-canon-cameras#more-8458

 

[Menace]

I went from 450d to a 7d - the AF performance difference between the two was amazing. Now I have a 5d III and the AF is even better

I don't think you will be disappointed. Happy shooting

 

[LetTheRightLensIn]

I was recently talking to a professional photographer friend about D800 vs 5D3 (both of which he uses extensively) and he strongly believes the 5D3 new autofocusing system is MUCH more accurate at low light in comparison its Nikon counterpart. He also told me about how Canon skin tones tend to better.. etc.. list goes on.. By the end I think im going to join the Canon side pretty soon.

Coming from a Nikon background, I'm not too familiar with anything Canon. Shortly after, I picked up my friend's 7D and i noticed the autofocusing was just bang on fast and amazing! Keep in mind I know the focusing is highly dependant on the lens too, i truly believe the 7D autofocusing system is solid.

Based on the rumours, it seems like the new 6D will inherit the 7D autofocusing system and the D600 will inherit the D7000 autofocusing system. I have already used the D7000 extensively in low light and it does its job decently, however, after picking up the 7D, I'm really unsure which one has the edge over the other.

A lot of the photography I do is done low light areas which is why this accuracy and speed in lightly lit areas will be the deciding factor. Can one of you more experience canon guys give me an unbiased opinion on the AF system in comparison to the D7000? Does the 7D low light accuracy resemble the godly 5D3 system I've heard about?

Thanks in advance!

From my experience so far, the 7D AF isn't nearly as good as the 5D3 AF.

 

[LetTheRightLensIn]

A lot of the photography I do is done low light areas which is why this accuracy and speed in lightly lit areas will be the deciding factor. Can one of you more experience canon guys give me an unbiased opinion on the AF system in comparison to the D7000? Does the 7D low light accuracy resemble the godly 5D3 system I've heard about?

Thanks in advance!

Not even close.

I own both cameras, and in my experience the 7D doesn't even come close to the 5D3's AF accuracy. Don't get me wrong, the 7D is a strong performer, but for my low light event photography, the 5D3 is my go-to.

Before I got my 5D3 I used my 7D in plenty of low-light event shoots, however, the keeper rate was less than ideal. When I couldn't use a speedlite for an AF beam, and had to rely solely on the sensor/lens, the 7D left a lot to be desired. At least in my opinion. Missed focus was a constant issue (check some of my old 2011 posts for rantings on the 7D's AF), and I ended up throwing away more shots than I ever had even on my 40D. With that said, the 7D is still my backup, but 10 times out of 10 I'll grab my 5D3 first for low-light shoots where no flash or AF beam is allowed. It's just that good.

Also - I hate to stir the pot, but in order to be fair I feel I must disclose all sides... hop on Google and search for "7D AF issue." Yes, there are some people who get a little crazy when discussing the 7D's AF performance, but there is definitely a consistent group of people who feel the 7D's AF is shoddy. Info here, and here.

If I were you I'd rent both, and judge for yourself. But if low-light AF is your thing, there's really no discussion as far as I'm concerned. The 5D3 is your only answer.

Also, check the post just a couple topics down from this one if you want to review some of the science behind both AF systems. According to Roger Cicala, the standard deviation performance of the 7D was the worst of all Canon bodies he tested (tied with the T3i). As for the 5D3... well... it was the best: http://www.lensrentals.com/blog/2012/08/autofocus-reality-part-3b-canon-cameras#more-8458

+1

 

[dr croubie]

Just a pointer:

The 1D and 1Ds series both used the same AF system. On the FF 1Ds, the AF spread was a lot less of the frame, on the 1D it covered more of the frame (because the 1Ds has the larger frame than the 1D. But it wasn't so bad on either camera (i've got the same AF on my eos 3, it's spread out more than enough).

Now, the AF points on the 7D cover about the same (on its crop sensor) as the 1Ds on its FF sensor.
Put that 7D AF on the 6D (or whatever the new FF will be called, i'd guess 6D), and the AF point coverage is going to be a smaller part of the frame than any FF camera (smaller than the 5D2's AF points?).
Neuro's got a a nice graphic of all that, maybe he'll see this thread and chime in with it.

But anyway. The AF points themselves may work beautifully. But if they don't cover much of the frame you'll notice that more than the number of points.

(that said, my 7D focuses great, on its aps-c sized frame at least)

 

[M.ST]

The AF system in the 7D can´t reach the AF system build in the 5D Mark III or 1D X.

 

[Kernuak]

For me, while the 7D performs well in good light, when the contrast drops (such as low light or a background similar to the subject), it isn't as accurate. So far, I've only had limited chance to the the 5D MkIII AF, but there's no comparison, it has easily focused in situations where the 7D would struggle. One area where the 7D may have an edge (although manual focus is more relevant anyway) is macro. The spot focus seems to be better in that situation, but it isn't something I've tried a great deal on either camera. As for the D7000, I don't have any experience.

 

[JigokuTrigger]

First I should point out that I don't own either the 7D or a 5D3. I recently sold me 550D in order to buy a 7D though, in the process of waiting for my shipment.

A few weeks back I borrowed my Friends D7000 and played with it for a day, in particular I did a bit of night time extended exposure. What really stuck out at me during the actual test was, how fast the AF on the D7000 but somewhat inaccurate(4/10 were on target). In compared to my 550D which took twice as long to focus using a L lens (70-200 f4) but at least got it right 8/10 times, I'd rather go with my 550D. You shouldn't take this as absolute though, this has only been my experience.

 

[sproggit]

Fickle, very Fickle

It's been a real education to read some of the replies posted in response to this question, such that I feel compelled to add my own 2 cents' worth.

I upgraded from an EOS40D to the 7D in 2010, partly because of the reportedly much better quality sensor in the 7, partly because of the increase from 6.5 to 8.4fps [I shoot a lot of hawks, owls, buzzards]. I carried a set of L Series glass with me to the new camera: 16-35mm f2.8, 24-70mm (Mk I) f2.8, 70-200mm (Mk I) f2.8, 100-400mm f4.5-5.6.

In almost all respects I find the 7D auto-focus to be a step back from the 40D. I think there are several reasons for this:-

1. You have to partner it with the right lens. Shooting raptors with my 70-200 I get 2-4 times as many pin-sharp "keepers" than with the 100-400mm. Yesterday I shot 650 exposures of the Olympic Sailing in Weymouth with the 100-400mm and less than 1% are keepers. In *all* cases it's [poor] focus that has spoiled the shot. Note that the 70-200 has USM, the 100-400 does not: that makes a *big* difference.

2. I think [I've read but now cannot find] that the 7D actually uses a different, "newer" AF when compared with the 40D. Perhaps this is the equivalent of a "Microsoft Version 1.0 Release"? [ i.e. they shipped it in order to get poor unsuspecting users to help identify the bugs].

3. This might be complete nonsense, but in my highly subjective view the worst possible thing you can give this AF system to work with is a tree-line in sunlight. The broken "surface" of the trees coupled with the hot-and-cold spots caused by sunlight and shade are enough for the "You gotta be kidding me!" warning light to appear in the HUD (if there was one, which there isn't). In other words, perhaps my experience is pushing the camera/lens combination to the limit.

4. I wonder - and I don't know - if the 7D is particularly fickle to "micro-adjustment" requirements for Canon lenses. I have not tried this, and I have no way of knowing if it's good, bad or indifferent, but last night, after a poor result from the Olympic shoot, I went in search of enlightenment and found this:

http://www.northlight-images.co.uk/article_pages/cameras/1ds3_af_micoadjustment.html

As the article explains that there is a "reset" option to keep everything safe, I am tempted to give it a try...

Bottom line, I find that when the 7D gets it right, it is capable of breathtaking shots, head and shoulders above my 40D. Unfortunately, my 40D returned 5-6 times as many "keepers" as the 7, and in *all* cases the reason for ditching 7D exposures is focus.

By way of comparison, I recently decided to treat myself to a 'go-anywhere' camera upgrade. I'd been using a Canon Powershot G12 since they came out [G9 before that, Ixus before that], but on the recommendation of a semi-professional photographer friend of mine, was lured to the Panasonic, and bought a Micro-Four-Thirds DMC-GX1 mirror-less. Wow. It doesn't have the frame rate of the 7D, but *every* shot is pin sharp. Even the kit lens turns in very respectable images... and put a decent bit of glass on it [I recently purchased the 7-14mm ultra-wide] and it just blows my 7D away...

I'm so hacked off with my 7D that I want to sell it. Would love a 1DX but need to wait for the price to come down and the savings to build up to match it... I asked my Father [who inherited my 40D, if he'd swap it for my 7. He knows the trouble I've had with this thing and the rotter won't bite...]

 

[Ryan708]

Re: Fickle, very Fickle

It's been a real education to read some of the replies posted in response to this question, such that I feel compelled to add my own 2 cents' worth.

I upgraded from an EOS40D to the 7D in 2010, partly because of the reportedly much better quality sensor in the 7, partly because of the increase from 6.5 to 8.4fps [I shoot a lot of hawks, owls, buzzards]. I carried a set of L Series glass with me to the new camera: 16-35mm f2.8, 24-70mm (Mk I) f2.8, 70-200mm (Mk I) f2.8, 100-400mm f4.5-5.6.

In almost all respects I find the 7D auto-focus to be a step back from the 40D. I think there are several reasons for this:-

1. You have to partner it with the right lens. Shooting raptors with my 70-200 I get 2-4 times as many pin-sharp "keepers" than with the 100-400mm. Yesterday I shot 650 exposures of the Olympic Sailing in Weymouth with the 100-400mm and less than 1% are keepers. In *all* cases it's [poor] focus that has spoiled the shot. Note that the 70-200 has USM, the 100-400 does not: that makes a *big* difference.

2. I think [I've read but now cannot find] that the 7D actually uses a different, "newer" AF when compared with the 40D. Perhaps this is the equivalent of a "Microsoft Version 1.0 Release"? [ i.e. they shipped it in order to get poor unsuspecting users to help identify the bugs].

3. This might be complete nonsense, but in my highly subjective view the worst possible thing you can give this AF system to work with is a tree-line in sunlight. The broken "surface" of the trees coupled with the hot-and-cold spots caused by sunlight and shade are enough for the "You gotta be kidding me!" warning light to appear in the HUD (if there was one, which there isn't). In other words, perhaps my experience is pushing the camera/lens combination to the limit.

4. I wonder - and I don't know - if the 7D is particularly fickle to "micro-adjustment" requirements for Canon lenses. I have not tried this, and I have no way of knowing if it's good, bad or indifferent, but last night, after a poor result from the Olympic shoot, I went in search of enlightenment and found this:

http://www.northlight-images.co.uk/article_pages/cameras/1ds3_af_micoadjustment.html

As the article explains that there is a "reset" option to keep everything safe, I am tempted to give it a try...

Bottom line, I find that when the 7D gets it right, it is capable of breathtaking shots, head and shoulders above my 40D. Unfortunately, my 40D returned 5-6 times as many "keepers" as the 7, and in *all* cases the reason for ditching 7D exposures is focus.

By way of comparison, I recently decided to treat myself to a 'go-anywhere' camera upgrade. I'd been using a Canon Powershot G12 since they came out [G9 before that, Ixus before that], but on the recommendation of a semi-professional photographer friend of mine, was lured to the Panasonic, and bought a Micro-Four-Thirds DMC-GX1 mirror-less. Wow. It doesn't have the frame rate of the 7D, but *every* shot is pin sharp. Even the kit lens turns in very respectable images... and put a decent bit of glass on it [I recently purchased the 7-14mm ultra-wide] and it just blows my 7D away...

I'm so hacked off with my 7D that I want to sell it. Would love a 1DX but need to wait for the price to come down and the savings to build up to match it... I asked my Father [who inherited my 40D, if he'd swap it for my 7. He knows the trouble I've had with this thing and the rotter won't bite...]

How about a 60D with under 4k clicks? Ill bite on the 7d

 

[Kernuak]

Re: Fickle, very Fickle

It's been a real education to read some of the replies posted in response to this question, such that I feel compelled to add my own 2 cents' worth.

I upgraded from an EOS40D to the 7D in 2010, partly because of the reportedly much better quality sensor in the 7, partly because of the increase from 6.5 to 8.4fps [I shoot a lot of hawks, owls, buzzards]. I carried a set of L Series glass with me to the new camera: 16-35mm f2.8, 24-70mm (Mk I) f2.8, 70-200mm (Mk I) f2.8, 100-400mm f4.5-5.6.

In almost all respects I find the 7D auto-focus to be a step back from the 40D. I think there are several reasons for this:-

1. You have to partner it with the right lens. Shooting raptors with my 70-200 I get 2-4 times as many pin-sharp "keepers" than with the 100-400mm. Yesterday I shot 650 exposures of the Olympic Sailing in Weymouth with the 100-400mm and less than 1% are keepers. In *all* cases it's [poor] focus that has spoiled the shot. Note that the 70-200 has USM, the 100-400 does not: that makes a *big* difference.

2. I think [I've read but now cannot find] that the 7D actually uses a different, "newer" AF when compared with the 40D. Perhaps this is the equivalent of a "Microsoft Version 1.0 Release"? [ i.e. they shipped it in order to get poor unsuspecting users to help identify the bugs].

3. This might be complete nonsense, but in my highly subjective view the worst possible thing you can give this AF system to work with is a tree-line in sunlight. The broken "surface" of the trees coupled with the hot-and-cold spots caused by sunlight and shade are enough for the "You gotta be kidding me!" warning light to appear in the HUD (if there was one, which there isn't). In other words, perhaps my experience is pushing the camera/lens combination to the limit.

4. I wonder - and I don't know - if the 7D is particularly fickle to "micro-adjustment" requirements for Canon lenses. I have not tried this, and I have no way of knowing if it's good, bad or indifferent, but last night, after a poor result from the Olympic shoot, I went in search of enlightenment and found this:

http://www.northlight-images.co.uk/article_pages/cameras/1ds3_af_micoadjustment.html

As the article explains that there is a "reset" option to keep everything safe, I am tempted to give it a try...

Bottom line, I find that when the 7D gets it right, it is capable of breathtaking shots, head and shoulders above my 40D. Unfortunately, my 40D returned 5-6 times as many "keepers" as the 7, and in *all* cases the reason for ditching 7D exposures is focus.

By way of comparison, I recently decided to treat myself to a 'go-anywhere' camera upgrade. I'd been using a Canon Powershot G12 since they came out [G9 before that, Ixus before that], but on the recommendation of a semi-professional photographer friend of mine, was lured to the Panasonic, and bought a Micro-Four-Thirds DMC-GX1 mirror-less. Wow. It doesn't have the frame rate of the 7D, but *every* shot is pin sharp. Even the kit lens turns in very respectable images... and put a decent bit of glass on it [I recently purchased the 7-14mm ultra-wide] and it just blows my 7D away...

I'm so hacked off with my 7D that I want to sell it. Would love a 1DX but need to wait for the price to come down and the savings to build up to match it... I asked my Father [who inherited my 40D, if he'd swap it for my 7. He knows the trouble I've had with this thing and the rotter won't bite...]

I also went from the 40D to the 7D and simply put, there is no comparison. When it comes to tracking objects and flexibility, the 7D is better. However, get it wrong and the higher pixel desnsity of the 7D is far less forgiving, showing up every error. When first released, there was alot of furore over the soft appearance of the 7D, but mostly it was down to slightly off focus or motion blur, which was much more obvious than earlier, lower resolution APS-C sensors. Choosing the correct AF mode is also critical, choose the wrong one and the focus is likely to be off. Take the time to learn the best focus mode for different circumstances and the 7D will perform well. In addition, AFMA is important, as again, slight focus inaccuracies will be shown in greater detail than on the 40D. In general, I have found that the 7D requires faster shutterspeeds to freeze motion.

 

[thfifthcrouch]

I just read sproggit's entry. My experience with the Canon 40d and the Canon 100-400mm was completely different. Sproggit mentioned that the said lens was not a USM, I just checked mine out and is says Ultrasonic and the literature says it is a USM. Maybe not. Nothing but soft images with this combination, began to think it was the lens. But once in a great while I would get a very sharp image and wondered why that was such if in fact I got a soft copy of the 100-400.
Last December I left my home in New England for a couple of weeks in California. I just purchased a Canon 60d and took that with me. My wife and I like bird watching. We noticed that a local state park was doing a bird watch. On the watch we spotted a California Scrub Jay. Took out the 60d and the 100-400mm and took two pictures expecting the worst. To my shock both pictures were perfect. Took a lot of pictures of Canyon Towhees, Acorn Woodpeckers and such. When I went inside to get out of the California Sun I was surprised at how good the captures were. It was then I decided that the Canon 40d and said lens has some front or back focus issue. I am more certain of that now having read sproggit's entry and his experience with the 40d.
I decided to acquire a Canon 7d. It focused just as well as the Canon 60d except in AI Servo in which to me was a little faster then the 60d going near to far. Pictures are just as sharp but the 7d seems to have a slight banding issue. To me there wasn't a lot of difference between the two. The points do cover more area and I kinda of gotten some decent picture of birds in flight. Personally for me it's more about good mono pod and technique but I am no master of that. NOW should I get my 40d fixed?????????????

 

[dr croubie]

That depends on what is your definition of "acceptable".
I've gotten 'acceptable' (to my eyes/needs) photos out of my 7D at iso3200. Sure, I got 'better' shots at the same gig with my mate's 5D2 at 3200, but for what I wanted, both the 7D and 5D2 were 'good enough', and the AF on the 7D was also working better than the 5D2.

 

[xopher]

Haha no im not getting the 7D, im considering the 6D WITH the rumoured 7D autofocusing module.

Well from what you guys are saying... i might as well save pennies for the 5d3...

 

[paul13walnut5]

I have the 7D and when you hit the right settings the AF is excellent, the trouble is too much choice.

I've settled mainly on spot AF or single point AF with manual selection (tracking the subject in the VF with the cursor/joystick) in AiServo mode, which also works best with a little set up and practice.

I also tend to use bright lenses. When AF performance is an issue you really need to give your camera (any canon EOS camera) as much light as possible, this means f2.8 or faster. Yes some bodies work fine with f4 or f5.6 or even f8 lenses, but they all work better with f2.8 or faster lenses.

I have two lenses that I used for performance subjects, the f2.8 USM 70-200 and the f2.0 100mm.

The body is only half the story.

Re: Noise.. I tend to shoot RAW and apply a little luma NR and a little sharpening in post with my 7D and T3i.

I wouldn't use any higher than 3200 for stills except in a real emergency, and no higher than 800 for video.

I know that the FF and higher / more recent bodies do better in low light, but then they are a lot more money, except for the 5D2 which has other concessions.

I don't need camera that goes beyond these very high ISO settings, so I'm happy.
Also, being more interested in video, if it's too dark, I add light.

 

[Lichtgestalt]

Also, check the post just a couple topics down from this one if you want to review some of the science behind both AF systems. According to Roger Cicala, the standard deviation performance of the 7D was the worst of all Canon bodies he tested (tied with the T3i). As for the 5D3... well... it was the best: http://www.lensrentals.com/blog/2012/08/autofocus-reality-part-3b-canon-cameras#more-8458

well i like roger.. but he has tested with one lens (28mm IS) and no tracking.
so you can´t say much about the overall AF perfomance from that test.

sure the 5D MK3 AF is better.. well it´s newer and the camera cost twice as much, so it better is.

and at the end he says something important:

Otherwise I might have put my trusty old 85mm f/1.8 on a 5D III and thought “not better at all, they’re lying to me again.”