View Single Post
Old Jan 31, 2013, 03:48 PM
Tom Frank is offline
Find More Posts by Tom Frank
Dance the skies...
Tom Frank's Avatar
United States, MA, Walpole
Joined Dec 2003
19,008 Posts
Originally Posted by ridgewalker View Post
Used above ^ method for focusing the camera,happy with the results,thanks for sharing that method.
This method works so well that you don't even need to actually view the image while conducting the focus process!

As i said the video focus was looking pretty good from taking video ,check it ect.
So decided to do the focus method based of the photo file size blind...take a photo and only noting the file size without viewing the image.
First photo...benchmark...second photo turn lens,file size decrease....third photo (turn lens back past benchmark spot file size increase.....forth photo,turn lens further from benchmark, file size decrease...fifth photo turn lens back toward benchmark,file size increase 50 over benchmark file size and left it there.

Now time to view the images and see how this blind test worked out.
They looked good and the last image was most clear.
Step outside to record a test video,come back in and view the video.
Video looks good and image is in focus better than it had been with using the record video,check,adjust lens repeat method.

If your needing to focus your camera try the photo file size comparison method,definitely gives very good results,recommend this method for anyone who needs to focus your camera.
I just ran my own test, because I wanted to test how easy the time lapse photo method would be (I set for 5 sec. intervals). In between each photo I turned the lens very slightly, maybe 1-2 deg, holding the case in position as best I could. This gave a very quick sequence of photos with pretty much the same scene (the new flat-bottom case really helped with this). My goal was not to set the lens to it's best position, but to see how close the files size method was to what my eye considered best. I used a lens that was being considered for the D lens... similar AOV but with some out of focus at the edges (typical with these inexpensive lenses with wide AOV).

My conclusion... the file size will get you VERY close, but not necessarily the best if measured by your eye. This was an acid test because the lens focal plane is not flat like the CMOS is. So the center focus is slightly different from the edge focus, and this led to a false positive on the file size method giving the "best" focus compared to what my eye determine was better across the entire frame.

YMMV, and for many people who aren't overly concerned if the focus is dead on, the file size method will get you VERY close if you can zero in on it.

FWIW, I have attached the photo sequences from my test... one distant and one fairly close up. I would use the photo method again to zero in on the near-best lens position, but would shoot additional photos to zero in to dead on as determined by my eye, with a method to identified what rotation of the lens was for each photo (e.g. some tape on the lens housing and lens barrel).
Tom Frank is offline Find More Posts by Tom Frank
Last edited by Tom Frank; Feb 01, 2013 at 07:16 AM. Reason: spelling
Reply With Quote