Are there any which don't display the focus shift? I understood that it was a design issue due to not having a floating element and not a manufacturing error.
Nov 29, 2012 at 11:36 AM
Lars Johnsson Offline • • • • • • • Upload & Sell: Off
Lars Johnsson wrote:
Yes all the 50/1,2 lenses have focus shift. From the first to the last one. Not a big deal for me when shooting, but all have it.
+1, an optical bending light effects @ stepped down aperture like many others fast lenses under 50mm from canon, nikon, pentax, zeiss....that makes your focal plane to be different than your optical plan when using off center AF sensor(s). The focus shift was caused by your body with lesser AF but not from the lenses...
to illustrate the aperture light bending effects of your image plane to be shifted after your focal plane has been set:
1) AF sensor(s) read the default optical plane for AF. Using the above GUI to see your f1.4 optical plane at wideopen and remember your focus plane position.
2) Let say you want to shoot at f2 and use off center AF for your shot. Using the above GUI to see your f2 optical plane position. This optical plane position is your final image
3) Compare your focus point from last two optical planes (f1.4 and f2) to see why your image is in backfocus. The closer to the center the less bending effects you will get.
Who to blame now: Lens or Body...i would say the BODY. Some bodies have on-board fast lense database to offset the bending effects, but most don't and we do have a long list of canon fast lenses from 14 to 50mm.
to illustrate the aperture light bending effects of your image plane to be shifted after your focal plane has been set:
1) AF sensor(s) read the default optical plane for AF. Using the above GUI to see your f1.4 optical plane at wideopen and remember your focus plane position.
2) Let say you want to shoot at f2 and use off center AF for your shot. Using the above GUI to see your f2 optical plane position. This optical plane position is your final image
3) Compare your focus point from last two optical planes (f1.4 and f2) to see why your image is in backfocus. The closer to the center the less bending effects you will get.
Who to blame now: Lens or Body...i would say the BODY. Some bodies have on-board fast lense database to offset the bending effects, but most don't and we do have a long list of canon fast lenses from 14 to 50mm. ...Show more →
I thought traditional focus shift was an optical phenomenon whereby you focus once (you can use center point), do not refocus for any subsequent shots, but just stop down and continue to shoot, you will see depth-of-field increase, but also shift. In some instances, I think dependent on distance to subject, the shift could cause your intended target to be outside the plane of sharp focus.
dwweiche wrote:
I thought traditional focus shift was an optical phenomenon whereby you focus once (you can use center point), do not refocus for any subsequent shots, but just stop down and continue to shoot, you will see depth-of-field increase, but also shift. In some instances, I think dependent on distance to subject, the shift could cause your intended target to be outside the plane of sharp focus.
What you said ...works wells for an old film system where focus system and lens aperture can be controlled separately by the user. Our DSLR does not work that way. DLSR AF system performs it function while lens at wide open and our image sensor recording image right after the aperture stopping down. Below is a sample to illustrate the camera and lens sequence. Please, note the black square as our focus point.
Our AF focal plane stay the same where image focal plane has been flattered out by lens stepping down to f2.8. This DSLR focus shift phenomenon is a bit different than older film system
I think we are losing something in translation, and talking about two different phenomenon. The focus shift I am referring to is not related to an "old system" no more than yours. In fact the Diglloyd link has examples at center point AF with a 1Ds3 and 50L f/1.2. The pictures show the shift.
Not saying the field curvature issue you are showing isn't real. I can see how it could throw off focus using outer points, but what I am calling "focus shift" is not what you are illustrating.
Asummed our DSLR has only one AF point (no more than one AF sensor) then the focus shift phenomenon might be easy to explain in both Diglloyd case and the case that i present here.
Too bad, our current Canon DSLR has more than one AF sensor(s) and all of them reading the object focal plane then reporting to AF firmware at the same time. At this point who knows what canon AF firmware is doing and which AF main sensor has controlled the focal plane. I hate to say this... when the focal plane has shifted at an angle, all light break loose by canon AF firmware since we don't know which sensor has the best contrast to wrap on (even only one AF sensor was actived). As long we have our focal plan that in parrallel with our image sensor then we can rely on center AF point to do its jobs. Another word, Diglloyd case spells posible different phenomenon: AF sensor shifting malfunction of canon AF firmware when shooting image in an angle with a busy background. I don't trust canon AF whenever we have multiple bright contrast points or busy background suround our main object. It always spells "trouble"
To prove my point, i have never experienced 50L focus shift with any canon DSLR when using center AF sensor.
mttran wrote:
To prove my point, i have never experienced 50L focus shift with any canon DSLR when using center AF sensor.
The 50L like many other lenses have focus shift when using the center AF. And Diglloyd also show this. He have a lot of tests and articles about it. But most of them are in the paid section.
