New Lens Align Method

DChan wrote:

The DOF beyond the subject is always greater than the DOF in front of the subject. When the subject is at the hyperfocal distance or beyond, the far DOF is infinite, so the ratio is 1:∞; as the subject distance decreases, near:far DOF ratio increases, approaching unity at high magnification. For large apertures at typical portrait distances, the ratio is still close to 1:1. The oft-cited rule that 1/3 of the DOF is in front of the subject and 2/3 is beyond (a 1:2 ratio) is true only when the subject distance is 1/3 the hyperfocal distance.

From here:


http://en.wikipedia.org/wiki/Depth_of_field

Explains why my middle point was always forward of the focus.

neverspook wrote:"I'm not familiar with it but I assume the lens align method looks at a resulting test picture, in which case for that method you are bringing depth of field into the equation and as correctly surmised if you set the MFA adjustment to equal front and back focus when the dof produced is a 1/3 to 2/3's ratio then you would actually be introducing a small element of front focus to the combination, i.e. the sharpest point would be at the 1/3 point not the 1/2 point."

Wouldn't that be back focus, since you are setting MA to be at 1/2 of the way back when it should be 1/3 of the way back for exact focus at the point of focus?

Roberta Olenick
www.neverspook.com

You could well be right Roberta, I'm punch drunk trying to work this out!

I've sat here and written up about 4 different ways of explaining how I'm seeing it but I just do not seem to be able to concisely articulate it.
 
I consider front focus to be where the sharpest part of the image is in front of where I've expected or tried to get the camera to focus and back focus the sharpest part of the image is behind, perhaps I'm wrong in thinking that way.

So it could be just the way I'm looking at it - but I've drawn it several times and I still keep coming back to the same conclusion

Irrespective of this I can't help thinking that introducing a new variable - depth of field - into MFA isn't the way to do it since it is exactly that, a variable and as there is only one true plane of focus why not use that. Of course in reality that may not be practical if you are testing against a large depth of field as it may be impossible to tell the the true plane of focus by eye (in which case why would you be using MFA). I tend to think in terms whereby the process is used to get critical focus in fairly small to very shallow dof situations where the true plane of focus is more easily determined and more critical.

Kev

I think you will find that MFA will not alter the DOF characteristics of your lens. You are zeroing in on the point of sharpest focus by assuming the response of the AF sensor is symmetrical. You are looking at the range over which the AF sensor reports the image in focus, this is quite wide, When I did the testing it was 8 to 10 units either side of focus.

So this allows you to set the camera to focus at the sharpest portion of the range over which the AF sensor reports the image as in focus. So assuming you actually hit the sharpest point when you make the adjustments, the DOF around this point will vary depending on magnification as explained above. That is you'll have equal DOF in front and behind at high magnification and 1:2 if you are at the right distances to achieve that.

I found reference in a Canon article which stated that each unit of MFA adjustment moves the focal plane by about 1/8 of the total DOF at the test condition.

ChrisRoss wrote:I think you will find that MFA will not alter the DOF characteristics of your lens.  You are zeroing in on the point of sharpest focus by assuming the response of the AF sensor is symmetrical.  You are looking at the range over which the AF sensor reports the image in focus, this is quite wide, When I did the testing it was 8 to 10 units either side of focus.  

So this allows you to set the camera to focus at the sharpest portion of the range over which the AF sensor reports the image as in focus.  So assuming you actually hit the sharpest point when you make the adjustments, the DOF around this point will vary depending on magnification as explained above.  That is you'll have equal DOF in front and behind at high magnification and 1:2 if you are at the right distances to achieve that.  

I found reference in a Canon article which stated that each unit of MFA adjustment moves the focal plane by about 1/8 of the total DOF at the test condition.

Thanks for that stat, Chris. I have always wondered how much of a difference each step of MA adjustment made.

I guess, as long as you are doing your MA tests at a distance where DOF is close to even front and back (which is the case for Lens Align's recommended distances for all but shorter lenses), then the various methods would yield similar results and be based pretty much at the plane of focus.

Roberta Olenick
www.neverspook.com

Hey Kev,

I guess it depends on your starting point whether you end up with front or back focus using Lens Align's equal front/back focus in a situation where DOF is really 1/3-2/3.
Here is how I look at it.
Let's say Lens Align at MA=0 gives you equal front/back focus. But let's say you have tested a lens at a distance where DOF is really 1/3-2/3. Then LA would be telling you that you should be focussing further back than the 1/3 point that is your real sharpest plane of focus, yes? Because LA says to select an MA that gives 1/2-1/2 and 1/2 back from the front is further back than 1/3 is back from the front, if that makes sense. So if you used a different MA method that selected the MA based on the actual sharpest plane of focus (as opposed to 1/2-1/2 focus), then this non-LA method would tell you that at MA=0, you are back focussed to 1/2 of the DOF and you need to adjust the MA so you are focussing further forward than what MA=0 does. At least on my camera (Canon 1DX), a minus setting compensates for back focus and a + setting compensates for front focus. So using a method that looks for sharpest focus rather than 1/2-1/2 DOF front and back, I would need to set a minus MA to achive the 1/3-2/3 focus required for max sharpness in this hypothetical situation. Based on Chris's info that 1 MA point moves the focus by 1/8 of the DOF, a setting of slightly more than -1 would be the recommended MA for a non-1/2-1/2 method based on achieving sharpest focus.
But as stated above in various posts, since for all but the shorter lenses, DOF is close to 1/2-1/2 at the distances most people tend to test MAs, both the LA and other methods should give very close results. For example, according to the Lens Align distance tool, for a 200 f2.8 tested at a distance to the target of 50X the focal length, 49% of the DOF is forward of the plane of focus and 51% is behind. Not enough difference to even worry about.

Does that all make sense? Really, my head spins whenever I think about MAing. Gone are the good old days when cameras did not have an MA adjustment option

Roberta Olenick
www.neverspook.com

Here's a link to the article that notes how much the focal plane moves with each step for Canon cameras:

http://www.learn.usa.canon.com/resource ... icle.shtml

Thanks for that, Chris.

Not all of that article is true today on the newer bodies. It is accurate for the 1D4/7D/5D2 generation but not for the 1Dx/5D3 generation which has significant;y more micro-adjustment capability.

E.J. Peiker wrote:Not all of that article is true today on the newer bodies.  It is accurate for the 1D4/7D/5D2 generation but not for the 1Dx/5D3 generation which has significant;y more micro-adjustment capability.

Is it just that you can microadjust both the short and long end of a zoom lens on the 1DX/5D3, or there also some other new capability now?

Thanks.

Roberta Olenick
www.neverspook.com

Also the camera remembers lenses by serial number so you can have more than one lens of the same type and use different values and the camera will remember (I don't know how many lenses it can do this for).

Right, yes. And you can do more than 20 lenses now, too.

I tested two lenses, using this very simple and fast testing method. (Green Dot)
Since I tested both lenses with LensAlign and then FoCal, I practically "tested" the testing method.
Results; Canon 1D MIV + Canon 300 f 2.8 L IS. tets results lined up with that of previously established values
using FoCal and LensAlign.
Result were the same (value right on) with a Canon 400 f5.6 L using the same camera.
I couldn't test a 500 as I don't have enough room inside.
I'm impressed so far.

For bird photography, I've wondered if it makes any sense to air slightly on the side of front focus? It seems that on my 1DIV, even at say 30ft, that the main focus picks up on the outline or contrast of the bird more so than parts of the bird itself. (always exceptions). For instance if you were shooting a tennis ball, and main focus is determined by the contrasting edges,why would you want to lose half your DOF out in space.
Just a thought and I may be out in space on this but wondering what others thought.

In your example Bill, I think you want to err (not air ) on the side of back-focus precisely so that you don't lose DOF to the space in front of the bird (assuming you are focusing on the eye); however, in most cases in bird photography, unless close to the subject you tend to be more towards the lens regime where 1/3 in front and 2/3 behind the subject is in focus. If focusing on the eye, you want to leave some space for areas closer to the camera than the eye, like the body.

In your example, you could lose feather detail behind the eye if you set the camera to purposely front focus.

Best to leave it set-up to critically focus on whatever you put the AF sensor on.

Thanks EJ,
The reason I asked was because doing some very quick testing with my camera, the auto focus can and does pick up on things fairly far outside the center focus point (using center point only). Say a half of box in all directions. So even if you focus on the eye, I'm not so sure that's what is really being picked up. After using other methods and checking with this dot tune, I've been shooting slightly front focused but have been happy with the results. I'll reset and air (I mean err) toward the best results.