- August 9, 2011 at 12:26 am #761
I have noticed a fairly high number of mis-focused photos when using the 180mm at near the close focus distance. Up till now I have blamed environmental conditions (e.g., wind or user error (e.g, body sway) as the culprit, but it kept nagging at me because I missed some decent shots due to focus error. As a result, I performed a focus test of the 180mm as shown below.
S2, 180mm, Tripod mounted, 2 sec self timer, image cropped to see detail of focus point (at the box corner on the 18 mark)
I believe there is about 1/8″ to 3/16″ back focus shown here. What do you think? Is this a flawed test or is there back focus?
By the way, I defocused manually and then hit the autofocus (rear button). I did this three times and the results were identical. The autofocus is very consistent and accurate to where it thinks it should be.
- August 9, 2011 at 3:19 am #762
Looks like the auto focus picked the 8 as the area with the most contrast and thus looks like it is back focusing. Does this happen at close distances with a flat target? One way to do this is to get parallel with a flat target, let auto focus choose the focus point and then ever so slightly tweak the focus back and forward, each time taking a shot. If the flat target with auto focus is the sharpest then it is correctly working.
- August 9, 2011 at 9:14 am #763ka7197New MemberJoin Date: Jul 2011Posts: 13Offline
Mark Gowin;495 wrote: What do you think? Is this a flawed test or is there back focus?
It's a flawed test. There might be some back focus but you cannot be sure.
The flaw is the target being slanted. The AF sensor has some physical size so there simply is no well-defined distance to focus at when the target is not parallel to the sensor.
For the next test, set up a flat target, e. g. a cardboard box with some fine-print text on it (the finer the better; if it's some product package then use the rear side), so that it's parallel to the sensor and the AF sensor sees nothing but the box's front face. Keep in mind that the actual AF sensor usually is a bit larger than the marker in the viewfinder suggests. Arrange the slanted ruler close to the image center right above or below the box but outside the view of the AF sensor. There might be some field curvature so don't try to assess focus accuracy anywhere near the frame's edges.
Maybe it's better to use not a ruler but something else instead because the point of maximum sharpness is hard to tell when it's falling in-between two numbers, as the ruler's surface is mostly smooth and featureless. A piece of wood might be better … how about a wooden ruler? Or stack two cardboard boxes on one another, one parallel (to focus at) and one slanted (to see where the sharpness is).
- August 9, 2011 at 2:41 pm #764
Thank you for the replies. I suspected that my test was flawed and should be reperformed. I will try again with a flat target. Also, I have an idea for constructing a 3D target which would be shot head-on that I might try as well.
For what it is worth, the test subject target above was my second attempt at a focus test. The first test was done in the field where the test target was a large bolt head on a wooden bridge. I figured it would make a good test target because there was very high contrast between the shaded side wrench flats versus the brightly sunlit head face. However, the results were the same as the ruler test above. I am not convinced either test was as good as it could be and perhaps not conclusive, but the bolt head test was a real world subject and the auto focus did not pick the right spot. Either the camera needs adjustment or I need to adjust my expectations of the auto focus. Hopefully proper test will resolve where the adjustment needs to occur.
- August 9, 2011 at 8:48 pm #767jrvNew MemberJoin Date: Jul 2011Posts: 16Offline
In that picture I think the focal point was just to the right of the “8”, at about the 18 1/4 point, maybe slightly (1/32) to the right of that.
For future cameras that's something for Leica's suggestion box: provide more information at the time of shooting as to where the focus point is.
More information recorded for later inspection would be useful too. Canon records such information in files so you can see what happened after a focus disaster.
Perhaps even record the measured distance.
- August 10, 2011 at 5:38 am #770
- August 10, 2011 at 4:14 pm #771
- August 11, 2011 at 1:42 pm #777David FarkasNewbieHollywood, FLJoin Date: Aug 2014Posts: 414Currently using:
Leica M, Leica S, Leica SL, Leica CLOffline
Mark Gowin;505 wrote: I have seriously considered getting the Lens Align. It looks like a well designed product and should be easy to use.
When the Lens Align first came out, I wrote a review for Digital Photo Pro Magazine. Here's the online version:
My primary focus (no pun intended) of the article was to explain its use for Nikon and Canon users, but it can be a very valuable tool in evaluating focus accuracy as well. Might be worth getting one.
- August 11, 2011 at 10:28 pm #779
With no ability to adjust the focus point on the S2 would this expense be fruitless? With Canon and Nikon the user can adjust the focus so the tool is useful. On the S2 if the lens is back/front focusing you have to send camera and lens in? So just a flat high contrast target would be the logical choice.
- August 11, 2011 at 11:15 pm #780David FarkasNewbieHollywood, FLJoin Date: Aug 2014Posts: 414Currently using:
Leica M, Leica S, Leica SL, Leica CLOffline
Atanabe;513 wrote: With no ability to adjust the focus point on the S2 would this expense be fruitless? With Canon and Nikon the user can adjust the focus so the tool is useful. On the S2 if the lens is back/front focusing you have to send camera and lens in? So just a flat high contrast target would be the logical choice.
The benefit of the Lens Align is that it has a flat front standard with high-contrast focus target and angled ruler in plane with the standard. While you can't adjust AF like on Canon or Nikon, it does give you a a pretty definitive method of quantifying AF accuracy. We've actually used it on M cameras to see rangefinder accuracy.
- August 11, 2011 at 11:57 pm #781
Agreed, the Lens Align is useful when you can make adjustments as with the M cameras but not useful at the price, with the S2. Also, the amount of correction +/- will vary by the distance from the subject as well.
Best to spend the money on Rogaine as you pulling your hair out in frustration. 🙂
- October 28, 2011 at 3:40 pm #1316Bobby LeeNew MemberJoin Date: Sep 2011Posts: 33Offline
I know it is a bit stupid to introduce a Chinese language web blog to you guys.
However, ths is the test I have made on the 180mm.
Pls note the size of the target with reference to the SD card below.
Athe closeup shot was the average result (sharp).
I have positioned the focuse circle at various point as shown in the second photo.
You must be logged in to reply to this topic.