Godfrey - GDGPhoto Flickr Stream
To answer your question, here is a cut and paste from the firmware updated user manual:
Bracketing ([Focus BKT] added)
[Focus BKT] (focus bracketing) is added to bracketing (P.90).
Take a series of shots at different focus positions. Focus
moves successively farther from the initial focus position.
Choose the number of shots using [Set number of shots],
the change in focus distance using [Set focus differential],
and the charging time for the external flash using [#Charge
Time]. Choose smaller values for [Set focus differential] to
narrow the change in focus distance, larger values to widen
Press the shutter button all the way down and release
it immediately. Shooting will continue until the selected
number of shots is taken or until the shutter button is
pressed all the way down again.
• For flash shooting, set the shutter speed to 1/20 sec. or
• Focus bracketing is not available with lenses that have
mounts conforming to the Four-Thirds standard.
• Focus bracketing ends if zoom or focus is adjusted during
• Shooting ends when focus reaches infinity.
1 Member(s) thanked for this post
Thanks Dave but that extract is for Focus Bracketing and not for Focus Stacking.
My crude attempt at focus stacking last night obliquely shooting the spines of books on a shelf and starting with the focus in the centre of the row/frame, demonstrated that the focus moves both further from and closer to the camera, and is automatically limited to eight shots (not enough to get a perfect result in that particular situation).
It implies from the wording in your cut and paste for focus bracketing, that the focus moves successively FURTHER from the initial focus distance set; so to achieve the same result of having all the spines in focus using bracketing, the initial point of focus should be on the closest book and the focus point will move further away for each of the user chosen number of shots; I've not tried that.
But back to Stacking and my initial question which I'm having trouble putting into words:
I shot my row of book spines with a vertical 'film plane' so there was only one plane of focus to follow, but what if the camera was also pointing upwards or downwards as well as along the row, how would the camera know which of the two planes of focus it should choose (from top to bottom of the spines, or closest to furthest book)?
I also can't fathom out how the merging works (I can't fathom much out ),
looking at each of the eight individual frames that make up the stack, very few appear to have any point in focus at all!!!
My advice : focus in the middle, or a little closer, and choose the f-value and focus differential wisely ...
Olympus could have found a more intelligent way, it seems to me ! Like : choose closest focus, choose furthest focus, and let the camera calculate optimum focus differential and f-value ...
But, that's only my idea ...
E-M1/GH2/G1 Full Spectrum & lots of lenses
http://www.flickr.com/photos/f6cvalk...th/9226689839/2 Member(s) liked this post
OK, here's my stab at seeing what focus bracketing/stacking will do for me and how it works:
The revised manual includes a few pages on new stuff in release 4.0. This is only part of it. Focus bracketing is one more kind of multishot bracketing. You can set it up and only use it when you set bracketing ON. When bracketing is OFF it will not interfere with normal shooting. I only tried stacking, which creates and combines 8 jpegs. The closest one is taken at the focus position you select (I used the touch screen, and touched the nearest object in the scene). The rest are taken at distances incremented by a number from 1 through 10. There is no explanation of whether this is object distance, amount of lens movement, or what. Only that 1 is small changes and 10 is big. Working with scenes that had a range of 2-3 meters, using the 12-50 at 12 or 14 mm focal length, "1" was plenty. I had my camera set for RAW only and after shooting, I found 8 ORFs, 8 JPEGs interspersed and the 9th JPEG (incrementing the frame number one more) was the combined result. You can look at JPEG #8 and see if anything was usefully in focus. If not, reduce the focus increment.
Here's one result:
PB281593 by scott kirkpatrick, on Flickr
click thru to see it on a larger scale. I uploaded the jpeg at 100%.
Was it worth the trouble? For comparison, here is the same scene shot in a single exposure at f/8, focused on the dry flowers in the middle:
PB281630 by scott kirkpatrick, on Flickr
1 Member(s) liked this post
Rafael starts with the focus in the middle or a little closer
and Scott starts at the closest point.
My take on the manual is that for Bracketing you should start at the closest point,
but for Stacking it is best to start around the middle.
Ian.1 Member(s) thanked for this post
The manual has a page on bracketing (up to 99 shots, put them together yourself afterwards, extending out from the closest object with an interval that you set ranging from 1 to 10). Followed by a page on stacking, where it puts 8 jpegs together. It only says "automatically changing the focus point slightly for each frame and merged to create an image that is focused on broad range from near to far field." This doesn't say the ordering of the images or whether or not the scale that you set for bracketing is used to space the stacked shots. The bracketing scale does influence the range over which the camera varies focus in doing stacking -- I could see that in the experiments that I did for the previous post. But...
You have to look at the 8 shots that end up on the chip to see what it is actually doing. So this time, I shot a tape measure, extended for a meter or so, and used 80mm focal length, @f/2.8. It seems to do the following. The first shot is at the point of focus. Then two shots are taken nearer to the camera and then five more further away. When the distance increment is 1 and I focused about a meter away, the regions in focus at f/2.8 and 80 mm overlapped. When the distance increment was set to 3 they didn't quite overlap. So I conclude that for stacking you should focus about 1/3 into the region of interest, and set the distance scale to match the lens focal length and f/stop to be used, by experimenting.
2 Member(s) liked this post
Reading the earlier posts more carefully, I see that Rafael and I have come to the same conclusions about the stacking algorithm.
1 Member(s) thanked for this post
Thanks to you guys and gals for helping clarify how "Focus Bracketing" and "Focus Stacking" actually function. It will help save me from doing this !
"Reality leaves a lot to the imagination." John Lennon
Bart, you've got the right equipment for this piece of reverse engineering. I don't have a LensAlign ruler. What are the dimensions in mm of the gradations that you show?
I assume you focused on the 0 point (the bright white line) in each shot? And what is the depth of focus on the lensalign scale for a single shot? (In my 80 mm f/2.8 test, it was about 5 cm.) I wonder if the focus changes that they make are keyed to the depth of field (which would mean a dependance on focal length and on aperture) or are just fixed distances. To test this would require shooting with the 60 macro at, say 5.6, stacking with 1 unit, and then with the zoom at maybe 18mm, f/2.8. Want to try it?
EDIT: Notice when you set the increment to 5, the steps were bigger than the depth of field of a single shot, leaving blurry portions at +20, +7.5 -5 and -14. I bet if you ran 5 units at f/5.6, those blurs would sharpen right up, and the total area covered would stay the same -- meaning that the units are actually lengths. That still leaves the question of whether the lengths are scaled to the distance at which you initially focus. Oh well, experiments still needed to use this.
1 Member(s) thanked for this post1 Member(s) liked this post
the way this ruler works is that you focus on a vertical plane that's aligned to the 0 on the ruler, so the nil mark is the point of focus.
From 0 to 16 on the ruler is approx. 50,5 mm. Mind you the ruler was at a 20 degree angle for the test.
And I agree using f5.6 or perhaps even better f8.0 will cover the in-between blurry areas when using increment 5, be it at the cost of subject isolation.
But why 18mm ?
I'd like to do some more testing, but not tonight.
If the increment in the focus bracketing menu (1 to 10) is distance, then changing the f stop without changing the focal length would fill in the soft gaps when you need more depth, but the centers of sharp focus would be in the same place in the individual jpegs. If it is drawn from a table of the depth of field at the focal length and aperture set, then the jpegs would stretch further when the fstop is set higher or you choose a shorter focal length. Of course the simplest answer is just to try it out, find something that works, and stick with it.
My initial attempts focus stacking on a small nature walk with the 12-40 pro. f2.8 used as in low light and closeup at 40mm (80mm effective)
3 Member(s) liked this post