Nikon vs. Canon: Auto-focus speed in Low Light: 50D vs. D300s; 1D Mk III vs.

[Chris7]

A guy at the local photo store said that the AF processors on the D700 and D3 were the same, so it seemed appropriate to compare the more expensive 1D Mk. III with the D700. (A comparison of the other cameras below is more relevant to what I buy, but I'm still interested at what's happening nearer to the top).

I was told that the low light auto-focus speed of the 50D and D300s were "close," but I'd like to hear some other people's experiences.

The 50D has a lot less AF points than the D300s, so the comparison might be helpful if talking about single-point AF.

In a review of the D300 (not "s"), The Imaging Resource said (here), "The Nikon D300 wasn't able to focus in this light without the AF assist light (which is why his eyes are slightly open here), something the Canon 40D did achieve..."

I think the D300s is said to have slightly improved low light autofocus over the D300, but I don't know by how much...

BTW, is the does the 7D do faster low light AF than the 50D, in single-point mode? How do low light AF speed of these two compare?

 

[OreoCookie]

If I remember correctly, you've already invested into Canon gear, right (the 7D if I remember correctly)?
If so, why do you care? Just be happy with the gear you got

FYI, the D300(s) basically uses the same AF sensor as the D3(s), D3x and D700. It's the crop version to be precise. The cpu which controls the AF, though, is not as beefy so the AF is a tad slower. I wouldn't be surprised if the D300 has better overall AF performance than the 50D. For most people the difference (be it either way) is just an academic exercise.

On the Canon side, things are much, much harder to compare. All the three cameras you've mentioned use very different AF modules.

 

[Westside guy]

Autofocus speed varies significantly on a lens-by-lens basis - and it's probably a much larger effect than any difference between the bodies' capabilities.

 

[El Cabong]

You're concerned with this why, exactly?

Are you going to be shooting in near-dark conditions without an AF illuminator on a regular basis?

Even then, the reviewer doesn't specify whether the 51-point area autofocus was active or if the camera was set to a less complicated autofocus mode. On top of that, the reviewer says, "It was taken with the 18-70 f/3.5, so that's probably why the D300 had trouble focusing, frankly" without specifying what lens was used on the 40D, and, if the lens on the D300 was faulted by the reviewer for affecting the autofocus, it's probably safe to say that the one on the 40D had a larger aperture (and that the difference was in the lenses, not the cameras).

The autofocus speeds of these cameras, given similar lenses and autofocus modes (i.e. 51-point autofocus points takes more time than single-point autofocus), should differ by only a few milliseconds.

Do you still care?

 

[pdxflint]

Certainly the higher Nikon models (D300/D300s/D700/D3, etc) have so many variation of AF set-up that it really is a wide-open question. Also, as mentioned, a wide-aperture lens will make a huge difference in low light for AF speed, among other things. On the Nikons you can have 51 pt. 3-D color tracking, 51, 21 or 11-point dynamic, or of course single (user selected) point. How you set the camera up for different custom shooting styles (banks) can make a big difference in AF performance, depending on subject, action and lighting. So, there's a lot more to it than simply saying one model is faster in low light than another...

 

[Chris7]

If I remember correctly, you've already invested into Canon gear, right (the 7D if I remember correctly)?
If so, why do you care? Just be happy with the gear you got

OCD? Anyway, my wife wasn’t real into the idea of the 7D (though I may still get one), and decided to go with the t2i, until I realized it did not have the ergonomics or AF performance I wanted. I like the feel of the 50D but it’s a bit noisy. Then realized I should really see what the 60D had to offer before I invested in one line or the other. (Etc., etc, etc.) I own only one Canon product, the 28mm 1.8, although I do borrow a friend’s cheap Canon zooms and XT.

I wouldn't be surprised if the D300 has better overall AF performance than the 50D. For most people the difference (be it either way) is just an academic exercise...

Re: the 50D vs. the D300, are you speaking of center-point AF? Re: "academic exercise," in the review the D300 could not even lock focus with out an AF assist lamp when the 50D could (although we don't know if they were using the same speed lenses in the comparison). So your experience with these cameras is that they're not that different with AF in low light?

Autofocus speed varies significantly on a lens-by-lens basis - and it's probably a much larger effect than any difference between the bodies' capabilities.

Low light shooting would be done primarily with an EF 50mm f/1.4 USM or AF Nikor 50mm f/1.4D (the D is actually said to be faster than the AF-S "G" series), if that helps.

You're concerned with this why, exactly?

Are you going to be shooting in near-dark conditions without an AF illuminator on a regular basis?...

I would like to shoot bursts when photographing moving people indoors with an f1.4, ISO 1600-3200, and shutter of 1/250 (likely single point AF). Is this light bright enough that none of these cameras would slow down from the AF speeds they get in bright light? If this is the case, it would seem Canon would be much faster simply due to the lens.

The autofocus speeds of these cameras, given similar lenses and autofocus modes (i.e. 51-point autofocus points takes more time than single-point autofocus), should differ by only a few milliseconds.

Do you still care?

About a few milliseconds? Depends on how many.

 

[flosseR]

Ok I really don't like these comparisons because they are ALL just academic as it has been mentioned. I don't know the reviewer of the D300, but comparing the 5d mark 2 and the D700 I can safely say the 51 point AF on the D700 achieves autofocus in near dark conditions and its FAST.

now i would like to see where it "says" that the 50mm f1.4D is faster to AF than the AF-s part as ANY AF'-s lens will focus faster.. that's the point of the built in SWM instead of using the screw drive to focus. however that being said, in real life unless you do extremely fast action sports, you won't notice and it won't matter that much. I have both AF and AF-s lenses and you can tell the difference but not by much.. achieving low light focus on the D90 and the D700 that I know and own(ed) is awesome and i have yet to encounter any low light situation where the nikons did not achieve focus...

but, thats just personal experience.. YMMV.

 

[El Cabong]

now i would like to see where it "says" that the 50mm f1.4D is faster to AF than the AF-s part as ANY AF'-s lens will focus faster..

Ask and ye shall receive

http://www.dpreview.com/lensreviews/nikon_50_1p4g_n15/page2.asp

I wouldn't have thought so either. Again, as the review states, the difference in real situations should be negligible.

[I agree with the rest of your post. I was actually shooting this weekend with a D300 in a dark nightclub and autofocus was just dandy.]

 

[pdxflint]

I can easily see where some AF-D lenses will AF faster than some AF-S ones, depending on the body being used. Not all built-in AF motors are the same. On my D300 the same AF-D lenses get "quicker" than they did on my D50. Faster built-in screw drive motor makes the difference. And I image the available electrical current from the battery(s) might also make a difference from one body to another...

 

[OreoCookie]

OCD? Anyway, my wife wasn’t real into the idea of the 7D (though I may still get one), and decided to go with the t2i, until I realized it did not have the ergonomics or AF performance I wanted. I like the feel of the 50D but it’s a bit noisy.

Get over it. No matter what you buy, you'll be unhappy and worrying that you're missing out on feature abc.

Re: the 50D vs. the D300, are you speaking of center-point AF? Re: "academic exercise," in the review the D300 could not even lock focus with out an AF assist lamp when the 50D could (although we don't know if they were using the same speed lenses in the comparison). So your experience with these cameras is that they're not that different with AF in low light?

Modern AF systems are so difficult to handle properly, because you need to use the right settings for the right circumstance. My camera (a lowly D80) has 3x3x2 focus modes (not sensors, different modes!). Bigger Nikons (and Canons, of course) have even more.

If you're getting substandard AF performance, you're most likely in the wrong mode anyway.

Regarding AF assist lights, I recommend you switch them off. Get an external flash, the decent ones include IR AF assist lights that are way, way, way better than the camera's own AF assist light. My SB-600 projects a cross pattern which is ideal to get a quick focus lock. Couple that to a fast lens (I have a 30 mm f/1.4 Sigma), and you're alright in low light. Again, that's with a D80 which surely has a worse AF system than the aforementioned cameras. Also, speed isn't everything, accuracy is still king. Handling a camera in a low light situation is very difficult. Sometimes it's easier to go manual.

I would like to shoot bursts when photographing moving people indoors with an f1.4, ISO 1600-3200, and shutter of 1/250 (likely single point AF). Is this light bright enough that none of these cameras would slow down from the AF speeds they get in bright light? If this is the case, it would seem Canon would be much faster simply due to the lens.

That doesn't make sense to me: you're thinking of a specific situation. And you're thinking of a specific AF mode (which is still unambiguous as there are two single-shot AF modes on modern Nikons).

 

[peskaa]

Guy at your local camera store is half right, half wrong. The D700 and D3 share the same AF sensor, but they have a different CPU doing the processing, with the D700's being slower. In real life terms this results in the D700 having a definite lag to the AF, particularly compared to the D3.

I used a D700 and a D3s to shoot a press gig a while back (cameras were loaners from Nikon), coupled with top end lenses (14-24, 70-200 etc.) with basically identical AF performance. The D3s was fast, and was responsive enough to grab the shots I wanted - there was hardly any delay between squeezing the shutter and the shutter releasing.

The D700 on the other hand was noticeably slower. Sure, it would AF and lock on, but the time taken to actually complete the whole shoot sequence was long enough to actually miss shots. I was not impressed in the slightest to be honest, and wasn't expecting that kind of result. The D700 just isn't as responsive, and this tells in high speed shooting environments.

D700s work great in a studio, and on shoots where you're not relying on tight timing. If you're shooting reportage, action, sports etc. then it just doesn't cut the mustard and you want a D3.



The 1D Mark III, which I own as my own personal kit, is on a par with the D3 in terms of AF speed and responsiveness.

 

[flosseR]

The D700 on the other hand was noticeably slower. Sure, it would AF and lock on, but the time taken to actually complete the whole shoot sequence was long enough to actually miss shots. I was not impressed in the slightest to be honest, and wasn't expecting that kind of result. The D700 just isn't as responsive, and this tells in high speed shooting environments.

I am (almost) calling bull on this one. The module (AF) is exactly the same. I know several press photographers (including big events like the G summits) that swear by the D700 and the D3s for that matter and cannot notice really any difference between the AF systems (other aspects sure but the shooting itself, no, i just called 2 of them to check).Also, you are comparing a camera from 2008 to a camera from 2010 At 5-8fps I wouldn't consider the D700 a Studio cam and it is not marketed as one.

dpreview (http://www.dpreview.com/reviews/nikond3s/page2.asp Specifications for D3 and http://www.dpreview.com/reviews/nikond700/page2.asp specifications for D700) and Luminous Landscape (http://www.luminous-landscape.com/reviews/cameras/d700.shtml) both go by "identical AF systems" .

Which CPU are you talking about since the module contains a CPU and they both have Expeed Imaging processors?

 

[peskaa]

I am (almost) calling bull on this one. The module (AF) is exactly the same. I know several press photographers (including big events like the G summits) that swear by the D700 and the D3s for that matter and cannot notice really any difference between the AF systems (other aspects sure but the shooting itself, no, i just called 2 of them to check).Also, you are comparing a camera from 2008 to a camera from 2010 At 5-8fps I wouldn't consider the D700 a Studio cam and it is not marketed as one.

Which CPU are you talking about since the module contains a CPU and they both have Expeed Imaging processors?

The system is identical, in terms of the number of sites, spread over the viewfinder, etc. But I can't just shake the fact it's slower somehow. The only fact I've heard is that the CPU is slower in the D700, and that's why it is a smidge slower in real world performance. Whilst they may both have Expeed processors, perhaps there is a small clock speed difference?

http://www.fredmiranda.com/forum/topic/822617
Other people with exactly the same performance as me. The D3 feels snappier, but it is hard to quantify. I'd agree I get more keepers out of a D3 than a D700.

Now, I'm not rubbishing the D700. However, I'm stating that in my work environment where we are shooting studio and PR, and use a mix of D700 and D3 (D3, D3s, D3x) the D700s have been relegated to studio work and the D3 series go wherever. Obviously it's still a very competent camera, but give me a choice and I'll take the D3 if I have to rely on AF.

 

[El Cabong]

Rhymes with 'coal'...

Quick glance at threads started by the OP:

January:
1) 2 stops better low light performance from Canon Rebel to 7D and from 7d to 5D Mk. II?
2) Used Canon 5D Mk II? Used Nikon Nikon D700?
3) Canon T1i vs. Nikon D5000 at high ISO

March:
4) My Nikon/Canon dilemma
5) Used price for a Canon 7D in 2 years?

Now:
6) Nikon vs. Canon: Auto-focus speed in Low Light: 50D vs. D300s; 1D Mk III vs. [D700/D3?]

Are you just asking all this out of curiosity? Do you have a DSLR yet/still? Or are you just trying to stir the Nikon vs. Canon pot?

 

[compuwar]

The system is identical, in terms of the number of sites, spread over the viewfinder, etc. But I can't just shake the fact it's slower somehow. The only fact I've heard is that the CPU is slower in the D700, and that's why it is a smidge slower in real world performance. Whilst they may both have Expeed processors, perhaps there is a small clock speed difference?

Have you tried turning off AF assist on the D700? It could be that using it slows the system down a bit. Alternately, it could be the battery, especially if you're using AF-D lenses- are the D700's you're using gripped and using teh same batteries?

Paul

 

[ManhattanPrjct]

Quick glance at threads started by the OP:

January:
1) 2 stops better low light performance from Canon Rebel to 7D and from 7d to 5D Mk. II?
2) Used Canon 5D Mk II? Used Nikon Nikon D700?
3) Canon T1i vs. Nikon D5000 at high ISO

March:
4) My Nikon/Canon dilemma
5) Used price for a Canon 7D in 2 years?

Now:
6) Nikon vs. Canon: Auto-focus speed in Low Light: 50D vs. D300s; 1D Mk III vs. [D700/D3?]

Are you just asking all this out of curiosity? Do you have a DSLR yet/still? Or are you just trying to stir the Nikon vs. Canon pot?

Zing! I guess people tend to fall into three categories:

1) Impulse buyers
2) Diligent research, then an informed decision
3) Too much research, agonize over a decision

 

[Chris7]

Thanks for your time.

Thanks pescar and flosseR.

...If you're getting substandard AF performance, you're most likely in the wrong mode anyway...

Helpful, thanks.

Zing! I guess people tend to fall into three categories:

1) Impulse buyers
2) Diligent research, then an informed decision
3) Too much research, agonize over a decision

...Are you just asking all this out of curiosity? Do you have a DSLR yet/still? Or are you just trying to stir the Nikon vs. Canon pot?

Your criticism is at least partially warranted. If my indecisiveness have rubbed you the wrong way, it may be because you and others here have graciously taken he time to give some very thoughtful, accurate, and meaningful responses.

Some of what I have asked is indeed out of curiosity of what for me is unaffordable mind-blowing technology (e.g. D3 & 1D Mk. III, technology that will be trickling down) or fanciful longer term “planning.” However, I have tried to be forthright about my lack of experience and what equipment I own.

The reason I am still borrowing my friend’s XT, and obsessing over a decision is largely to do with the current inability of Canon consumer/prosumer cameras to allow the user to set the aperture and minimum shutter, and have the camera auto expose with auto ISO in 1/3 stops or less, (EDIT) a "feature" standard with Nikon (The 7D will do Auto ISO with fixed shutter and aperture, but will not do minimum shutter, so will overexpose without adjustments, say, when the camera is pointed out the window. It also does not allow for exposure compensation or max ISO in this mode). While I would really like this feature, my wife would not buy a camera without it. This is weighted on the other hand with a personal distaste for the sound of Nikon’s quality but squeaky primes, and I do not want to buy third party lenses (I realize third party lenses an area of heated debate here, and not one I want to start on this thread).

So I’m trying to find something to tip the scale for me between the two brands. Hence this thread about low light autofocus speed. Maybe only this Fall’s 60D will tell if Canon decides to move forward with a more sophisticated Auto ISO. The decision should be relatively straightforward from there.

Anyway, I sincerely appreciate the thoughtful advice of the people here. I have gotten a lot of information that I need here, for example about high ISO performance at pixel vs. image level, the importance of camera ergonomics (ruling out the Rebel series for me), info about autofocus unreliability with certain third party lenses, what it takes to get adequate OOF blur, and recommendations on photography books.

I may be back after I’ve made a decision and a purchase.

-Chris

 

[peskaa]

Have you tried turning off AF assist on the D700? It could be that using it slows the system down a bit. Alternately, it could be the battery, especially if you're using AF-D lenses- are the D700's you're using gripped and using teh same batteries?

Paul

I'll have a look and see tomorrow. We're using the D700s gripless for size reasons, with the standard ship batteries. I think we may have one grip kicking round that I can try.

 

[flosseR]

Hey Chris, My D700 does auto ISO with fixed minimum Shutter speed and Aperture Priority..

works great thought eh Auto ISO is a bit aggressive every now and then...

 

[Chris7]

Hey Chris, My D700 does auto ISO with fixed minimum Shutter speed and Aperture Priority..

works great thought eh Auto ISO is a bit aggressive every now and then...

Thanks. Looks like I forgot to mention that I know this feature to be standard on Nikon cameras, at least from the D90 and up. I meant to say this in my last post, but I just re-read it and realize I never explicitly stated this.

 

[flosseR]

Thanks. Looks like I forgot to mention that I know this feature to be standard on Nikon cameras, at least from the D90 and up. I meant to say this in my last post, but I just re-read it and realize I never explicitly stated this.

Ok never mind then

 

[Chris7]

works great thought eh Auto ISO is a bit aggressive every now and then...

I didn't think to ask -- what do you mean by Auto ISO being "aggressive"?

ISO on the D90 does minimum increments of 1/3 stop. If I set the minimum shutter to 1/250, it seems that, as light gets progressively darker, the shutter would fall from just under 1/320, down to 1/250; and then the shutter would jump back up almost 1/3 stop (say, to 1/310), as the ISO jumps up 1/3 stop (hope I said that right).

Is this how it works? Or by "aggressive," do you mean that sometimes the shutter never gets down to the set minimum shutter (e.g. 1/250) before the ISO jumps up a 1/3 stop (just guessing here)?

Thanks for your time.

 

[Ruahrc]

As I recall (I don't use Auto ISO but my D80 has it too) you can define the minimum shutter speed for Auto ISO. So if you set the Auto ISO limit at 1/250, it will stay at the lowest ISO possible as long as it can make 1/250. If it needs to go to say 1/200 at the current ISO, then it will bump up the ISO until it can meet the minimum 1/250 setting. It's only as aggressive as you set it.

Ruahrc

 

[ManhattanPrjct]

As I recall (I don't use Auto ISO but my D80 has it too) you can define the minimum shutter speed for Auto ISO. So if you set the Auto ISO limit at 1/250, it will stay at the lowest ISO possible as long as it can make 1/250. If it needs to go to say 1/200 at the current ISO, then it will bump up the ISO until it can meet the minimum 1/250 setting. It's only as aggressive as you set it.

You can also set a maximum ISO (like on a D90, it will perform without much noticeable noise until 800 or so) this way.

 

[Chris7]

...it will stay at the lowest ISO possible as long as it can make 1/250. If it needs to go to say 1/200 at the current ISO, then it will bump up the ISO until it can meet the minimum 1/250 setting. It's only as aggressive as you set it.

That works. Thanks.

You can also set a maximum ISO (like on a D90, it will perform without much noticeable noise until 800 or so) this way.

Humm. With, for example, aperture set wide open, ISO set at max 800, and shutter set at minimum 1/250... when the light starts to get too dark for these settings, will the camera just underexpose, or will the max ISO setting "override" the minimum shutter setting, and lower lower the shutter?