Home · Register · Search · View Winners · Software · Hosting · Software · Join Upload & Sell

Moderated by: Fred Miranda
Username   Password

  New fredmiranda.com Mobile Site
  New Feature: SMS Notification alert
  New Feature: Buy & Sell Watchlist
  

FM Forums | Canon Forum | Join Upload & Sell

  

5D3 AF vs Spot AF inconsistencies
  
 
skid00skid00
Offline
• •
Upload & Sell: Off
p.1 #1 · p.1 #1 · 5D3 AF vs Spot AF inconsistencies


I noticed last night that I was getting severe front-focus with my 5D3 & 24-70ii. This after spending hours getting to know how to AFMA and fine-tuning AF.

It seems that the reduced area of the Spot AF point can see things differently than can Live View AF in 10x mode.

After trying to blow out the back of the main mirror, the small secondary AF mirror, and the AF array (can't really see this, but blew a lot of air thru the interior with the main mirror slightly raised, I am seeing better results. I should point out that I stumbled across this 'fix' when, after 3 attempts, Canon couldn't get my 1Ds and 3 lenses to AF consistently. I haven't had any noticeable sensor dust...

I'm wondering if anyone else has noticed a difference between these AF modes?

In doing these tests (with a Manfrotto 3021 Pro and a 410 geared head), I also saw that there is *always* a slight amount of image blur from the mirror motion, vs Live View mode. Something to think about if you need pixel-level sharpness. I also used the quiet-mode shutter. I need to test if I can use mirror-lockup when shooting astro...

Unlike everyone else, I AFMA (by shooting a 13" x 19" white sheet with a 1" thick line thru the center) by watching for distance scale movement while changing from LV to AF point. I find this method to be very quick (minutes) and as accurate as is possible.




Aug 23, 2014 at 09:19 PM
kevindar
Offline
• • • •
Upload & Sell: Off
p.1 #2 · p.1 #2 · 5D3 AF vs Spot AF inconsistencies


blow back of main mirror, and small secondary af mirror? curious if you have any reference or pictures or video of what that entails. I guess you can blow the back of the mirror buy activating sensor cleaning


Aug 26, 2014 at 01:23 PM
TeamSpeed
Offline
• • •
Upload & Sell: On
p.1 #3 · p.1 #3 · 5D3 AF vs Spot AF inconsistencies


There are indeed differences between AF and live view, they use different AF systems.


Aug 27, 2014 at 09:59 AM
Ian.Dobinson
Offline
• • • • • •
Upload & Sell: Off
p.1 #4 · p.1 #4 · 5D3 AF vs Spot AF inconsistencies


skid00skid00 wrote:
It seems that the reduced area of the Spot AF point can see things differently than can Live View AF in 10x mode.

I'm wondering if anyone else has noticed a difference between these AF modes?

Unlike everyone else, I AFMA (by shooting a 13" x 19" white sheet with a 1" thick line thru the center) by watching for distance scale movement while changing from LV to AF point. I find this method to be very quick (minutes) and as accurate as is possible.



I think you may be using the 'spot' AF wrongly .

lets say you are using just the single centre point .
there are 2 modes for this , there is the normal AF point and the reduce area Spot AF .

spot AF is not to try and AF a small target but rather to be used to avoid something that may be picked up by the af point (generally in the foreground) . ie shooting thru a fence or long grass etc .
if you have no obstructions then the standard point will be more accurate

this was shot spot AF (on a 7D) to avoid the grass IMG_0499.jpg by dobbie99, on Flickr
if your trying to MA for spot AF then you probably would be better of targeting something that had a good texture like a brick rather than a high contrast target like a black white checker board .



Aug 27, 2014 at 10:55 AM
scalesusa
Offline
• • • •
Upload & Sell: Off
p.1 #5 · p.1 #5 · 5D3 AF vs Spot AF inconsistencies


Live View is going to be much more accurate, since it uses the actual image from the sensor. Phase Detect autofocus, be it spot or other is what you are adjusting with AFMA. As long as you are using a standardized target, Spot AF should be pretty much the same as normal AF, but it is not always as accurate, so turn it off when doing AFMA.



Aug 27, 2014 at 05:56 PM
 

Search in Used Dept. 



Inga
Offline
• • •
Upload & Sell: Off
p.1 #6 · p.1 #6 · 5D3 AF vs Spot AF inconsistencies


Ian.Dobinson wrote:
if you have no obstructions then the standard point will be more accurate


This is interesting to me.

[Sorry to hijack the thread]

When I'm doing portraits and headshots of relatively still and motionless subjects I often switch to the more precise AF spot method, thinking that it would be more accurate. I do often find that my images are slightly OOF when using fairly shallow DOF (shooting at f/2.8 at 200mm with the subject a few meters away). More often than not the camera seems to front focus slightly making the nose sharper than the eyes, and that continues even when I push the AFMA back further and further.

I have been using the FoCal software to AFMA my lenses, but maybe I should go with something more manual. Their method recommends using 50x the focal length as the distance to the target, which is much further than my usual shooting distance for portraits and headshots.




Aug 28, 2014 at 01:43 AM
Ziffl3
Offline
• • • •
Upload & Sell: Off
p.1 #7 · p.1 #7 · 5D3 AF vs Spot AF inconsistencies


Inga ... When you go to spot or small spot AF mode ... were is the AF sensor placed on the subject?
Even in spot mode (not the small spot) the actual active sensor area will extend slightly out side the AF box.
The system will want to find closest object to camera. so depending were you locate the AF spot, it might be picking up the nose.



Aug 28, 2014 at 01:55 AM
Inga
Offline
• • •
Upload & Sell: Off
p.1 #8 · p.1 #8 · 5D3 AF vs Spot AF inconsistencies


Ziffl3 wrote:
Inga ... When you go to spot or small spot AF mode ... were is the AF sensor placed on the subject?
Even in spot mode (not the small spot) the actual active sensor area will extend slightly out side the AF box.
The system will want to find closest object to camera. so depending were you locate the AF spot, it might be picking up the nose.


I always focus on the nearest eye. Usually this does require a slight re-composition. I am using the outer focus points on my 5D3 (the cross-type sensors only). I use BBF to help with the focus/recompose process.

The chance that the AF point is on the nose seems highly unlikely. The AF point grabbing the eyelashes, which are obviously forward of the eye itself, seems more likely given the 'expanded' nature of the focus point per your suggestion.



Aug 28, 2014 at 02:40 AM
schlotz
Offline
• • • •
Upload & Sell: On
p.1 #9 · p.1 #9 · 5D3 AF vs Spot AF inconsistencies


Inga wrote:
This is interesting to me.

[Sorry to hijack the thread]

When I'm doing portraits and headshots of relatively still and motionless subjects I often switch to the more precise AF spot method, thinking that it would be more accurate. I do often find that my images are slightly OOF when using fairly shallow DOF (shooting at f/2.8 at 200mm with the subject a few meters away). More often than not the camera seems to front focus slightly making the nose sharper than the eyes, and that continues even when I push the AFMA back further and further.

I have been using
...Show more

Inga,

I believe FoCal uses the phase detect AF process with multiple sampling to determine a proper AFMA setting. Someone correct me if I'm wrong. Having successfully used FoCal, I will say that after discovering the simplicity in using the Dot Tune method, I've stopped using FoCal's tethered computer based process and solely use the Dot Tune manual method. BTW: there is a caveat to the 50x recommendation: if you usually shoot at a specific distance then by all means do your AFMA at that distance.

Dot Tune > http://www.fredmiranda.com/forum/topic/1187247/0

Matt



Aug 28, 2014 at 11:17 AM
Inga
Offline
• • •
Upload & Sell: Off
p.1 #10 · p.1 #10 · 5D3 AF vs Spot AF inconsistencies


schlotz wrote:
Inga,

I believe FoCal uses the phase detect AF process with multiple sampling to determine a proper AFMA setting. Someone correct me if I'm wrong. Having successfully used FoCal, I will say that after discovering the simplicity in using the Dot Tune method, I've stopped using FoCal's tethered computer based process and solely use the Dot Tune manual method. BTW: there is a caveat to the 50x recommendation: if you usually shoot at a specific distance then by all means do your AFMA at that distance.

Dot Tune > http://www.fredmiranda.com/forum/topic/1187247/0

Matt


Thanks Matt. I haven't tried the Dot Tune method, although I've seen the thread in the past. TBH the FoCal software is a pain on the 5D3 because you have to adjust the AFMA manually. On the 5D2 the software is able to do the entire process without your input until it comes to deciding if their result is better than the previous setting.

The difficulty with the target distance is that headshots need to be critically sharp at the eyes so I guess that distance should take priority. But I also shoot some basketball and some wedding work with that lens, both at much longer distances. Guess I can always test both distances and see what the difference is...

I see a lot of tripod and focus target time in my future. Sigh.

Thanks guys.





Aug 28, 2014 at 11:51 PM





FM Forums | Canon Forum | Join Upload & Sell

    
 

You are not logged in. Login or Register

Username   Password    Retrive password