D800 AF-ON and AF-C

Started Feb 11, 2014 | Discussions
fPrime
fPrime Senior Member • Posts: 1,929
Re: Ming Thein found the same thing about AF-C

ormdig wrote:

"Although Ming Thein to this day still believes the D800 lacks the PDAF focus consistency required for use outside of the studio,". Where does he say this and why?

See here: http://blog.mingthein.com/2012/10/27/nikon-d800e-midterm-report/

Something still doesn’t feel right with the autofocus system.
Although my camera no longer exhibits any asymmetry with its focus points following the recalibration and fix by Nikon Malaysia, it just doesn’t seem to be as positive or accurate as the D700 was (or D600 is now). There are situations in which the camera nails everything perfectly, and situations under which it just seems to miss by a hair; far more of the latter exist than the former. And no combination of AF settings seems to work; this means that the D800 is effectively an unviable proposition to me as a documentary/ reportage camera. Bottom line: I’m not 100% confident that it’s going to focus where I tell it to.

I can’t even get a consistently sharp image unless I’m over 1/2x focal length – and I’m certain I’ve got better technique than average. This, and the size of the files (a throughput issue) make it impractical for a documentary/ travel/ journalism camera.

 fPrime's gear list:fPrime's gear list
Canon EOS 5D Fujifilm FinePix S5 Pro Nikon D1X Nikon D200 Nikon D700
fPrime
fPrime Senior Member • Posts: 1,929
Re: And consider trying AF-C tracking with "Release + Focus" as well...

primeshooter wrote:

fPrime wrote:

I've found that I get slightly better focus accuracy with AF-ON and AF-C on the D700 on static subjects versus AF-S. The AF-C advantage may be true in general for cameras with the CAM 3500 FX module.

Interestingly on the D700, in addition to the above, I've also found even with static subjects that switching to the crosshair 9 point AF-C tracking with "release+focus" can often beat out both single point AF-C or 9 point AF-C tracking with "focus" enabled.

Has anyone else experimented along these lines... using AF-C tracking for static subjects? I'm obviously testing focus performance under extreme conditions to evaluate optimal configuration... f/1.4 lens in low, bad light with medium to high contrast targets.

fPrime

Interesting indeed, I have always preferred the AF-C or AF-S single point never liked the area modes but I can see why folk use them. I guess the subjects I shoot don't move that erratically that I cannot just simply follow them with the focus box and save the bother of switching modes?

I know, it seems that 9 point AF-C tracking would pose a higher risk that an off-axis area of brighter reflection or better contrast might distract the focus in that direction. But for some reason the 9 point AF algorithm seems to work slightly better for me in low light than pure single point AF on static subjects with the D700.

Again "focus+release" with 9 Point AF tracking paradoxically seems to be slightly more accurate than forcing "focus"... exactly opposite what I'd expect. This is the reason I'm interested to hear if anyone else can confirm my test results with their own D700.

As to what might explain this behavior, the better performance could boil down to less processing delay from when the last AF-C PDAF measurement is made to when the shutter actually trips. Since "focus" presumably requires a longer confirmatory processing loop than "focus+release", using the latter setting may allow AF-C to more surely time the optimal lens focus with the shutter release. I believe the same rationale is the underlying basis for setting AF Lock On to OFF... less AF processing delay allows AF-C to catch optimal focus under tracking slightly better.

fPrime

 fPrime's gear list:fPrime's gear list
Canon EOS 5D Fujifilm FinePix S5 Pro Nikon D1X Nikon D200 Nikon D700
primeshooter
OP primeshooter Veteran Member • Posts: 4,315
Re: And consider trying AF-C tracking with "Release + Focus" as well...

fPrime wrote:

primeshooter wrote:

fPrime wrote:

I've found that I get slightly better focus accuracy with AF-ON and AF-C on the D700 on static subjects versus AF-S. The AF-C advantage may be true in general for cameras with the CAM 3500 FX module.

Interestingly on the D700, in addition to the above, I've also found even with static subjects that switching to the crosshair 9 point AF-C tracking with "release+focus" can often beat out both single point AF-C or 9 point AF-C tracking with "focus" enabled.

Has anyone else experimented along these lines... using AF-C tracking for static subjects? I'm obviously testing focus performance under extreme conditions to evaluate optimal configuration... f/1.4 lens in low, bad light with medium to high contrast targets.

fPrime

Interesting indeed, I have always preferred the AF-C or AF-S single point never liked the area modes but I can see why folk use them. I guess the subjects I shoot don't move that erratically that I cannot just simply follow them with the focus box and save the bother of switching modes?

I know, it seems that 9 point AF-C tracking would pose a higher risk that an off-axis area of brighter reflection or better contrast might distract the focus in that direction. But for some reason the 9 point AF algorithm seems to work slightly better for me in low light than pure single point AF on static subjects with the D700.

Again "focus+release" with 9 Point AF tracking paradoxically seems to be slightly more accurate than forcing "focus"... exactly opposite what I'd expect. This is the reason I'm interested to hear if anyone else can confirm my test results with their own D700.

I may have inadvertently noticed the same actually. It's like Nikon have made a few mistakes with this body or something? Focus is always more important to me than speed etc.

As to what might explain this behavior, the better performance could boil down to less processing delay from when the last AF-C PDAF measurement is made to when the shutter actually trips. Since "focus" presumably requires a longer confirmatory processing loop than "focus+release", using the latter setting may allow AF-C to more surely time the optimal lens focus with the shutter release. I believe the same rationale is the underlying basis for setting AF Lock On to OFF... less AF processing delay allows AF-C to catch optimal focus under tracking slightly better.

fPrime

fprime, which setting do you have, short medium long or off? Which do you find gives more accurate results?

fPrime
fPrime Senior Member • Posts: 1,929
Re: And consider trying AF-C tracking with "Release + Focus" as well...

primeshooter wrote:

fPrime wrote:

primeshooter wrote:

fPrime wrote:

I've found that I get slightly better focus accuracy with AF-ON and AF-C on the D700 on static subjects versus AF-S. The AF-C advantage may be true in general for cameras with the CAM 3500 FX module.

Interestingly on the D700, in addition to the above, I've also found even with static subjects that switching to the crosshair 9 point AF-C tracking with "release+focus" can often beat out both single point AF-C or 9 point AF-C tracking with "focus" enabled.

Has anyone else experimented along these lines... using AF-C tracking for static subjects? I'm obviously testing focus performance under extreme conditions to evaluate optimal configuration... f/1.4 lens in low, bad light with medium to high contrast targets.

fPrime

Interesting indeed, I have always preferred the AF-C or AF-S single point never liked the area modes but I can see why folk use them. I guess the subjects I shoot don't move that erratically that I cannot just simply follow them with the focus box and save the bother of switching modes?

I know, it seems that 9 point AF-C tracking would pose a higher risk that an off-axis area of brighter reflection or better contrast might distract the focus in that direction. But for some reason the 9 point AF algorithm seems to work slightly better for me in low light than pure single point AF on static subjects with the D700.

Again "focus+release" with 9 Point AF tracking paradoxically seems to be slightly more accurate than forcing "focus"... exactly opposite what I'd expect. This is the reason I'm interested to hear if anyone else can confirm my test results with their own D700.

I may have inadvertently noticed the same actually. It's like Nikon have made a few mistakes with this body or something? Focus is always more important to me than speed etc.

Agreed on the bias towards focus, primeshooter.

As to what might explain this behavior, the better performance could boil down to less processing delay from when the last AF-C PDAF measurement is made to when the shutter actually trips. Since "focus" presumably requires a longer confirmatory processing loop than "focus+release", using the latter setting may allow AF-C to more surely time the optimal lens focus with the shutter release. I believe the same rationale is the underlying basis for setting AF Lock On to OFF... less AF processing delay allows AF-C to catch optimal focus under tracking slightly better.

fPrime

fprime, which setting do you have, short medium long or off? Which do you find gives more accurate results?

I personally keep AF Lock On to OFF which, according to consensus, provides the highest AF accuracy with subjects that are rapidly change focus distance (like when the target is moving towards or away from the camera). It's my opinion that keeping it OFF doesn't hurt accuracy in static scenarios and may actually help AF-C by eliminating yet another processing load in focus calculation.

In shooting static subjects under medium to low incandescent & CFL lighting at f/1.4 I expect, and get, rock solid AF accuracy even with outer, non cross type, AF points on my D700.

fPrime

 fPrime's gear list:fPrime's gear list
Canon EOS 5D Fujifilm FinePix S5 Pro Nikon D1X Nikon D200 Nikon D700
primeshooter
OP primeshooter Veteran Member • Posts: 4,315
Re: And consider trying AF-C tracking with "Release + Focus" as well...

fPrime wrote:

primeshooter wrote:

fPrime wrote:

primeshooter wrote:

fPrime wrote:

I've found that I get slightly better focus accuracy with AF-ON and AF-C on the D700 on static subjects versus AF-S. The AF-C advantage may be true in general for cameras with the CAM 3500 FX module.

Interestingly on the D700, in addition to the above, I've also found even with static subjects that switching to the crosshair 9 point AF-C tracking with "release+focus" can often beat out both single point AF-C or 9 point AF-C tracking with "focus" enabled.

Has anyone else experimented along these lines... using AF-C tracking for static subjects? I'm obviously testing focus performance under extreme conditions to evaluate optimal configuration... f/1.4 lens in low, bad light with medium to high contrast targets.

fPrime

Interesting indeed, I have always preferred the AF-C or AF-S single point never liked the area modes but I can see why folk use them. I guess the subjects I shoot don't move that erratically that I cannot just simply follow them with the focus box and save the bother of switching modes?

I know, it seems that 9 point AF-C tracking would pose a higher risk that an off-axis area of brighter reflection or better contrast might distract the focus in that direction. But for some reason the 9 point AF algorithm seems to work slightly better for me in low light than pure single point AF on static subjects with the D700.

Again "focus+release" with 9 Point AF tracking paradoxically seems to be slightly more accurate than forcing "focus"... exactly opposite what I'd expect. This is the reason I'm interested to hear if anyone else can confirm my test results with their own D700.

I may have inadvertently noticed the same actually. It's like Nikon have made a few mistakes with this body or something? Focus is always more important to me than speed etc.

Agreed on the bias towards focus, primeshooter.

As to what might explain this behavior, the better performance could boil down to less processing delay from when the last AF-C PDAF measurement is made to when the shutter actually trips. Since "focus" presumably requires a longer confirmatory processing loop than "focus+release", using the latter setting may allow AF-C to more surely time the optimal lens focus with the shutter release. I believe the same rationale is the underlying basis for setting AF Lock On to OFF... less AF processing delay allows AF-C to catch optimal focus under tracking slightly better.

fPrime

fprime, which setting do you have, short medium long or off? Which do you find gives more accurate results?

I personally keep AF Lock On to OFF which, according to consensus, provides the highest AF accuracy with subjects that are rapidly change focus distance (like when the target is moving towards or away from the camera). It's my opinion that keeping it OFF doesn't hurt accuracy in static scenarios and may actually help AF-C by eliminating yet another processing load in focus calculation.

In shooting static subjects under medium to low incandescent & CFL lighting at f/1.4 I expect, and get, rock solid AF accuracy even with outer, non cross type, AF points on my D700.

fPrime

Sounds like a plan I will change it back to OFF. To give you a little background, pre firmware update (when focusing was AWFUL - despite what these fanboys on here tell me), I had moved it to OFF but after a while I got worried I was making it worse so switched it back to the default setting. Now since the firmware update I will try what you say, as it does make most sense that it'll work best at OFF.

primeshooter
OP primeshooter Veteran Member • Posts: 4,315
Re: And consider trying AF-C tracking with "Release + Focus" as well...

fPrime wrote:

primeshooter wrote:

fPrime wrote:

primeshooter wrote:

fPrime wrote:

I've found that I get slightly better focus accuracy with AF-ON and AF-C on the D700 on static subjects versus AF-S. The AF-C advantage may be true in general for cameras with the CAM 3500 FX module.

Interestingly on the D700, in addition to the above, I've also found even with static subjects that switching to the crosshair 9 point AF-C tracking with "release+focus" can often beat out both single point AF-C or 9 point AF-C tracking with "focus" enabled.

Has anyone else experimented along these lines... using AF-C tracking for static subjects? I'm obviously testing focus performance under extreme conditions to evaluate optimal configuration... f/1.4 lens in low, bad light with medium to high contrast targets.

fPrime

Interesting indeed, I have always preferred the AF-C or AF-S single point never liked the area modes but I can see why folk use them. I guess the subjects I shoot don't move that erratically that I cannot just simply follow them with the focus box and save the bother of switching modes?

I know, it seems that 9 point AF-C tracking would pose a higher risk that an off-axis area of brighter reflection or better contrast might distract the focus in that direction. But for some reason the 9 point AF algorithm seems to work slightly better for me in low light than pure single point AF on static subjects with the D700.

Again "focus+release" with 9 Point AF tracking paradoxically seems to be slightly more accurate than forcing "focus"... exactly opposite what I'd expect. This is the reason I'm interested to hear if anyone else can confirm my test results with their own D700.

I may have inadvertently noticed the same actually. It's like Nikon have made a few mistakes with this body or something? Focus is always more important to me than speed etc.

Agreed on the bias towards focus, primeshooter.

As to what might explain this behavior, the better performance could boil down to less processing delay from when the last AF-C PDAF measurement is made to when the shutter actually trips. Since "focus" presumably requires a longer confirmatory processing loop than "focus+release", using the latter setting may allow AF-C to more surely time the optimal lens focus with the shutter release. I believe the same rationale is the underlying basis for setting AF Lock On to OFF... less AF processing delay allows AF-C to catch optimal focus under tracking slightly better.

fPrime

fprime, which setting do you have, short medium long or off? Which do you find gives more accurate results?

I personally keep AF Lock On to OFF which, according to consensus, provides the highest AF accuracy with subjects that are rapidly change focus distance (like when the target is moving towards or away from the camera). It's my opinion that keeping it OFF doesn't hurt accuracy in static scenarios and may actually help AF-C by eliminating yet another processing load in focus calculation.

In shooting static subjects under medium to low incandescent & CFL lighting at f/1.4 I expect, and get, rock solid AF accuracy even with outer, non cross type, AF points on my D700.

fPrime

fprime, I take it you have abandoned AF-S? I just noticed you have a 24mm 1.4, would love one of those babies! It's my most used focal length for landscapes.

Martin Grecner Contributing Member • Posts: 520
Re: D800 AF-ON and AF-C
1

I never had problems with AF-S on D800 unless my subject was moving.

I do not know if D800 tries to implement predictive autofocus in AF-S mode like the D70 claimed to do, probably not, so be careful to avoid moving subjects with AF-S.

Martin

primeshooter
OP primeshooter Veteran Member • Posts: 4,315
Re: D800 AF-ON and AF-C

Martin Grecner wrote:

I never had problems with AF-S on D800 unless my subject was moving.

I do not know if D800 tries to implement predictive autofocus in AF-S mode like the D70 claimed to do, probably not, so be careful to avoid moving subjects with AF-S.

Martin

Fair enough but depends on your lens etc. The problem here is that several bodies I have tested I've seen this and other seasoned pros have noticed what I describe, as do many users here (many that know what they are talking about aswell). That even in AF-S mode with non movement targets - it's mince, especially in anything but broad daylight. *If you are using fast primes (f/1.4) in any sort of slight lowered contrast situation. The thing is you shouldn't need a damned zebra to focus in lower light with pdaf these days.

flbrit Veteran Member • Posts: 4,025
Re: D800 AF-ON and AF-C

primeshooter wrote:

The thing is you shouldn't need a damned zebra to focus in lower light with pdaf these days.

Very funny.  

Brian

 flbrit's gear list:flbrit's gear list
Olympus E-M1 II Carl Zeiss Planar T* 2/50 ZM Olympus 40-150mm F2.8 Pro Carl Zeiss Distagon T* 1,4/35 ZM Olympus 7-14mm F2.8 Pro +7 more
(unknown member) Veteran Member • Posts: 3,985
Re: D800 AF-ON and AF-C

flbrit wrote:

primeshooter wrote:

The thing is you shouldn't need a damned zebra to focus in lower light with pdaf these days.

Very funny.

Brian

Great. Now you've done it. Now we're going to see another rash of proper AF fine Tuning targets followed by commentary from Lennie

-- hide signature --

http://www.nightstreets.com
-
"Sick cultures show a complex of symptoms such as you have named...but a dying culture invariable exhibits personal rudeness. Bad manners. Lack of consideration for others in minor matters. A loss of politeness, of gentle manners is more significant than a riot."
This symptom is especially serious in that an individual displaying it never thinks of it as a sign of ill health but as proof of his/her strength. ...Friday, it is too late to save this culture--this worldwide culture... Therefore we must now prepare the monasteries for the coming Dark Age. Electronic records are too fragile..."
--Robert A. Heinlein in "Friday"

primeshooter
OP primeshooter Veteran Member • Posts: 4,315
Re: D800 AF-ON and AF-C

flbrit wrote:

primeshooter wrote:

The thing is you shouldn't need a damned zebra to focus in lower light with pdaf these days.

Very funny.

Brian

Great. Now you've done it. Now we're going to see another rash of proper AF fine Tuning targets followed by commentary from Lennie

-- hide signature --

http://www.nightstreets.com
-
"Sick cultures show a complex of symptoms such as you have named...but a dying culture invariable exhibits personal rudeness. Bad manners. Lack of consideration for others in minor matters. A loss of politeness, of gentle manners is more significant than a riot."
This symptom is especially serious in that an individual displaying it never thinks of it as a sign of ill health but as proof of his/her strength. ...Friday, it is too late to save this culture--this worldwide culture... Therefore we must now prepare the monasteries for the coming Dark Age. Electronic records are too fragile..."
--Robert A. Heinlein in "Friday"

Only matter of time before said poster shows up and quotes the nikon manual....then he will disappear and lie in wait of another af thread. A shameless fanboy.
--
facebook https://www.facebook.com/stevenrphotographs
flickr http://www.flickr.com/photos/steverphotographer/
google + https://plus.google.com/u/1/110463150518351139559

fPrime
fPrime Senior Member • Posts: 1,929
Re: And consider trying AF-C tracking with "Release + Focus" as well...

primeshooter wrote:

fprime, I take it you have abandoned AF-S? I just noticed you have a 24mm 1.4, would love one of those babies! It's my most used focal length for landscapes.

Correct, the C/S/M switch on my D700 may as well be glued on C. I wonder if that's why Nikon felt it was no big deal to downgrade the C/S/M switch to AF/M on the D800... because so many photographers rarely use anything other than AF-C?

I bought the 24mm 1.4G specifically for landscape and it is sweet in that role.  Focus isn't as snappy as my 50mm and 85mm Siggy's but for sharpness across the FX frame it rules.  I can see why Nikon have it on their recommended list for the D800.

fPrime

 fPrime's gear list:fPrime's gear list
Canon EOS 5D Fujifilm FinePix S5 Pro Nikon D1X Nikon D200 Nikon D700
BJN
BJN Veteran Member • Posts: 5,073
On a tripod, use contrast detect (live view).

If you really need critical focus, don't mess with the phase-detect system at all. Live view is the only really precise focusing tool on the camera. PDAF points aren't indicated precisely in the viewfinder, and they're not perfect no matter what mode you're using. The reflex finder isn't very useful for manual focusing.

-- hide signature --

BJ Nicholls
SLC, UT

 BJN's gear list:BJN's gear list
Panasonic Lumix DMC-GM5
sd40 Senior Member • Posts: 1,553
Re: Ming Thein found the same thing about AF-C

I want to express this lurker's appreciation for those who contribute to these AF-ON threads.  I have been following them since Ray Soares first made waves. I have tried the suggested techniques and had some success with them.  For the time being, at least, I have settled into using AF-C all the time (landscape and action).  I buy the argument that AF-C compensates for the photographer's motion as well as subject motion.  And there's something intuitive about letting the little black box fly around the viewfinder until it settles down.  My hands are still steady enough that it generally does settle down.

I have continued to use the half-press, however.  I'm comfortable using the Fn button on the front of the D600 for focusing, but unless I'm going to focus and recompose I don't see the advantage vs. using the half-press.

I'm still reading and learning.

 sd40's gear list:sd40's gear list
Nikon D750 Nikon AF-S Nikkor 24-70mm f/2.8G ED Nikon AF Fisheye-Nikkor 16mm f/2.8D Nikon AF Nikkor 105mm f/2D DC Nikon AF-S Nikkor 85mm f/1.4G +8 more
Martin Grecner Contributing Member • Posts: 520
Re: D800 AF-ON and AF-C

primeshooter wrote:

Martin Grecner wrote:

I never had problems with AF-S on D800 unless my subject was moving.

I do not know if D800 tries to implement predictive autofocus in AF-S mode like the D70 claimed to do, probably not, so be careful to avoid moving subjects with AF-S.

Martin

Fair enough but depends on your lens etc. The problem here is that several bodies I have tested I've seen this and other seasoned pros have noticed what I describe, as do many users here (many that know what they are talking about aswell). That even in AF-S mode with non movement targets - it's mince, especially in anything but broad daylight. *If you are using fast primes (f/1.4) in any sort of slight lowered contrast situation. The thing is you shouldn't need a damned zebra to focus in lower light with pdaf these days.

I have never tried f/1.4 lens yet. My fastest yet are 35mm f/1.8 G and 85mm f/1.8 D.

These work fine with AF-S on my D800 (well, at least not any worse than AF-C as far as I can tell).

But it might be different story on your camera, of course.

Martin

Rob Tomlin
Rob Tomlin Regular Member • Posts: 431
Re: Ming Thein found the same thing about AF-C

sd40 wrote:

I want to express this lurker's appreciation for those who contribute to these AF-ON threads. I have been following them since Ray Soares first made waves. I have tried the suggested techniques and had some success with them. For the time being, at least, I have settled into using AF-C all the time (landscape and action). I buy the argument that AF-C compensates for the photographer's motion as well as subject motion. And there's something intuitive about letting the little black box fly around the viewfinder until it settles down. My hands are still steady enough that it generally does settle down.

I have continued to use the half-press, however. I'm comfortable using the Fn button on the front of the D600 for focusing, but unless I'm going to focus and recompose I don't see the advantage vs. using the half-press.

I'm still reading and learning.

I just got my D800E earlier this week.  In the past I've always used AF-S with the half-press for focus.  I have rarely used the AF-ON button on my D700.

My confusion regarding using AF-C (and the AF-On button) is how do you focus on something and lock that focus if you want to recompose?

Obviously if you let off the AF-On button, the camera will stop focusing and theoretically focus will be locked, but when you go to press the shutter release the camera will focus again...other than where you wanted it locked.

 Rob Tomlin's gear list:Rob Tomlin's gear list
Sony RX1 Nikon D700 Nikon D800E Nikon AF-S Nikkor 70-200mm f/2.8G ED VR Nikon AF-S Nikkor 14-24mm f/2.8G ED +4 more
ormdig
ormdig Senior Member • Posts: 1,774
Re: Ming Thein found the same thing about AF-C
1

Hi Rob, go to Custom Setting Menu, Autofocus, AF Activation, AF-On Only. This separates the focus from the shutter button. The shutter button will now meter and shoot, the AF-On button will do the focusing. If you use AF-c the camera will constantly focus while you hold down the button. If you let it up, focus is locked. The shutter button will not refocus the image.

This is why everyone uses this method. You have AF-c and AF-s with one button and you don't have to switch anything. Either hold the AF-On button down (AF-c) or let it up (AF-s).

-- hide signature --

Pete

 ormdig's gear list:ormdig's gear list
Nikon D7000 Nikon D800 Nikon AF-S Nikkor 24-70mm f/2.8G ED Nikon AF-S Nikkor 70-200mm f/2.8G ED VR II Nikon AF-S 105mm F1.4E ED +7 more
Rob Tomlin
Rob Tomlin Regular Member • Posts: 431
Re: Ming Thein found the same thing about AF-C

ormdig wrote:

Hi Rob, go to Custom Setting Menu, Autofocus, AF Activation, AF-On Only. This separates the focus from the shutter button. The shutter button will now meter and shoot, the AF-On button will do the focusing. If you use AF-c the camera will constantly focus while you hold down the button. If you let it up, focus is locked. The shutter button will not refocus the image.

This is why everyone uses this method. You have AF-c and AF-s with one button and you don't have to switch anything. Either hold the AF-On button down (AF-c) or let it up (AF-s).

-- hide signature --

Pete

Great explanation and makes perfect sense.  Thanks a lot Pete!

 Rob Tomlin's gear list:Rob Tomlin's gear list
Sony RX1 Nikon D700 Nikon D800E Nikon AF-S Nikkor 70-200mm f/2.8G ED VR Nikon AF-S Nikkor 14-24mm f/2.8G ED +4 more
Keyboard shortcuts:
FForum MMy threads