An Investigation of E-M1 Focus Bracketing

Aucupium wrote:

Thanks Richard for this - quite illuminating!

What happens when we attach extension tubes to the 60mm lens? [...]

I don't know the full answer to your question, but I know that it works very well:

Pen-F, 60 mm + ET (crop, real magn ≈ 1.7x)

f/4.5, 50 steps (step 4)

This was done with a live moth, it would be unthinkable with focussing rails.

Nice! Moth species? 50 steps - quite a lot !

I can see I have a lot of trial and error to do over the winter months.

Andrew

Aucupium wrote:

Nice! Moth species? 50 steps - quite a lot !

I can see I have a lot of trial and error to do over the winter months.

Andrew

The moth is the Abraxas pantaria, a pest of the common ash tree. That was exactly where it was found, an ash tree near where I live (central Portugal).

The parameters I used were not the result of careful thought, It was just what I had found to be safe to do a very fast job at this scale. Being nocturnal, the moth was very quiet for some intervals, but I never knew when it would star twitching. I think that 30 steps would do, but the difference in total time is very small, better safe!

Good luck with your winter moths, I hope we will see some results.

Gary from Seattle wrote:

Richard Turton wrote:

Your interpretation is correct. The total size of the focus bracket is equal to the number from the table multiplied by the focus differential and number of shots settings.

The minimum focusing distance of the 60 mm macro lens is 8 inches. But that is the distance from the subject to the sensor ( manufacturers seem to agree on this definition). But the sensor is 4" behind the front of the lens, so at minimum focus the front of the lens is only 4" from the subject.

Ok, Richard. I read the thread but missed where your distances are to the sensor. This is a great table and I'll be sure to save it. Thanks.

I agree with others that Olympus would be interested and should publish this on their website or send it out by e-mail!

I second that, apart from the inches bit :/

As one retired ginger beer to another, excellent work buddy! This is exactly the sort of sleuthing we need to counter the marketing shenanigans that goes on in the camera manufacturing industry.

Having said that, I'm almost ashamed to ask you to go just a bit further and take a look at Oly's in-camera focus stacking. I do a lot of in-the-field macros of native flowers. Having the composite result right there on the camera in second is a God send for me and saves heaps of PP. However, again Oly is being mysterious. The stack does not start at the point of focus. It jumps forward two shots then backward 6 shots but I don't know if that is consistent and of course there is the issue of what is a focus differential. One would expect it to be the same as in focus bracketing but is it?

Also, on Oly's focus stacking site http://cameras.olympus.com/stack/en-au/ a landscape scene is shown.  The standard shot has only the foreground in focus.  The stacked shot has everything to infinity in focus despite using focus step 1.  This implies to me that there may be a non-linearity involved.  Of course there is next to no explanation.  The shot was taken at f3.2.  This would be a boon to landscape photographers who might otherwise have to stop down to f16 to get that DOF.

Again, brilliant work Richard. I like the log-log plots.

Hey Olympus, whatever you payed Kazuo Unno for that mediocre article on focus stacking/bracketing, you should pay Richard Turton double.

Richard, there is a wealth of data in these graphs but it's really hard to get one's brain around it.  In the field, once having composed, one then decides what continuous DOF is needed.  I use the 8 step in-camera focus stacking.  For me using the 60mm macro, knowing the focus distance, I need to determine what combination of focus differential and aperture will do the job.  At the moment it's just hit an miss and Olympus are no help at all.  Can you give some thought into what sort of field chart would do the job... then patent it.  I'll pre-order one.

BTW, I like your use of metric units.  No photographer should have difficulty with metres and mm's.

I too have switched from the 12-60 to the 12-40mm. It's way more compact and like the 12-60 it allows focusing almost to the surface of the lens. Most of all I like that they seem to have removed the annoying handlebar distortion that the 12-60 exhibited at 12mm.  The 12-40 still has some barrel distortion at 12mm but that's easier to correct in PP.

Richard Turton wrote:

I need to come up with a method of presenting the focus bracketing information that is practical to use in the field.

Hi Richard

I've been trying to do the same for in-camera focus stacking, it is a non-trivial challenge indeed (smile).  Looking forward to your results....

What you say makes a lot of sense. Actually I've more or less settled on a focus differential of 5 with an aperture of f5.6 but I'll give your 2 at f5.6 at try. Someone suggested that matching the focus differential number with the f-stop number seems to work. That may be a good starting point but I notice that Kazuo Unno's shots do not follow this pattern. If only we could get into his head.

I don't know how Oly's focus stacking compares to others but it is certainly very susceptible to motion. Hand held shots make for some fun images but for quality. a tripod and a motionless subject are essential. I no longer attempt to photograph wild flowers if there is even a breeze about. Forget about moving insects. The subject will be sharp so the software is aligning quite well but the motion artifacts around it are too distracting.

Yes, choosing the right focus point for focus stacking is a bit of an art.  The software seem to have some smarts built in to decide on how many shots it needs to take in front of the focus point.  I'm going on the assumption that it is using the 1/3 in front 2/3 behind rule but it would be nice to know.

Jeff,

Based on your use of a Panasonic lens on your Olympus body there seems to be only two rules for selecting a lens for focus bracketing:

1. M43 mount

2. AF

Is this correct?

Thanks,
Bill

Thanks a lot for this very informative post. This Chart answers a lot of questions. For what it's worth, in most macro/close up shooting situations, I find the using the #3 interval and aperture of f5.6 works very well.

Good stuff! Now to design, manufacture and market a circular slide rule of this data. One would have thought that Oly could build this into the menu. Better still, a total DOF slider in the VF. The camera knows the focus distance, the CoC, the chosen total DOF, it can then adjust the aperture and focus differential, then shutter speed.

BTW, I normally use a cable release for focus stacking. Today I tried the Wi-Fi release but got just one exposure. Couldn't find mention of this in the manual. Perhaps it's a bug.

I tried focus differential of 3 at f5.6 but couldn't quite get the depth I needed for wild flowers so I went back to focus differential 5.  I'm not seeing any OOF banding, at least not on the screen.  I could go to f8 but then the background blur will not be as good.

Great job Richard, thanks. For sharing. Your example here confirms my meager attempts at decifering the optimum step/differential  for my type of shooting, which is flowers. I have been using 5 steps, differential  3 as my default. Shot below is a 5/3 handheld.

Nice!  You must have hands of steel.

BobT3218 wrote:

Nice! You must have hands of steel.

Actually, at a burst rate of 10 fps, the 5 frames in the bracket set happens pretty quickly, and the IBIS of the E-M5 II is pretty amazing. Halicon Focus does a great job of aligning the shots, and a slight movement left or right of the initial focus point can be accommodated. But..........if you move forward or back, the size of the subject changes and your stack will turn to mush.

If I were to use to many more than 5 frames in the set, it would be more difficult to hand hold.

Richard Turton wrote:

The total depth of focus for 8 shots is shown

Hi Richard

I've measured actual depths of field in an Oly focus stack, and, except for the 0.2 m distance, the table seems a good approximation (I've only measured for distances of 200, 250, 330, 500, and 1000 mm to subject).  For some reason, the DoF in my near 1:1 focus stacks is only around one third of that predicted by the table (I measured at apertures 2.8, 4, 5.6, 8):

At greater subject distances, the actual stack DoF seems to agree within fairly broad measurement error, for example, at 1 m: