An Investigation of E-M1 Focus Bracketing

Truly a great post indeed! Thx a lot 👌.

I will most probably be using Focus Stacking w/ the 40-150 PRO (with and without MC-14) for dragon- and butterflies (when a flawless EM-1 fw v4.3 is out that is). I will reference the guidlines here for sure, although the longer FL will require some other focus values. I guess for a 1m working distance, some diff. value at 2-3 and F5.6 will be in the ballpark w/ the 40-150. I will need a total stacked DOF around 10-12 cm or so.

Exiting times ahead 😊.

Richard Turton wrote:

The previously posted Focus Bracketing table has been reworked for the Focus Stacking function of the E-M1. The differences between Focus Stacking and Focus bracketing are:

1. With Focus Stacking the final output is produced in camera as a JPEG. With Focus Bracketing the shots have to be combined later using third party software.

2. The Number of Shots is fixed at 8 and not adjustable.

3. The initial focus point is located part way through the sequence instead of at the beginning. For Focus Stacking the step sequence is 0, -2, -1, 1, 2. 3. 4. 5 where 0 is the initial focus point. For Focus Bracketing the step sequence is 0, 1, 2, 3, 4, 5, 6, 7.

Separate tables are provided for each aperture. The total depth of focus for 8 shots is shown as a function of Focus Distance (runs vertically) and Focus Differential (runs horizontally). If the Focus Differential is greater than 3 there is a risk that the depth of field may be less than the focus step. Depth of field is not considered here. The original document is in PDF, which has been converted to JPEG so that it can be embedded here. It should fit on a single letter size sheet.

I just tested these out in the range of 200mm with the 60mm lens with settings from F5.6, 8, 11 focus step 5, f8 and F11 step 4, and F8 and F11 step 3. I found results very similar to your chart results in that range, perhaps 1.5mm less total DOF on average - but that is totally dependent on the limits of my eyesight and measuring methodology. I set up at an approximate down angle of 45 degrees at minimum focus distance +- (I measured with a tape measure) and shot a ruler with a $20 bill beneath it. I focused at the 3" mark on the ruler but adjacent to the bill such that focus peaking showed equal amplification on either side of the edge of the ruler adjacent to the bill. I felt that my vision with the marks on the ruler would not have allowed accurate measurement but that the ink patterns on the bill allowed me to be accurate to around 1/32" or .8mm.

My interest at the present time is shooting mosses, not so much for photographic artistry but as to be able to firmly identify most mosses primarily from the general habit or growth form of the mosses, their habitat, and in focus stacking with the macro lens to obtain leaf detail that is on the order of 3X more magnified than a 15X loupe.

I originally started out with single, manually focused images at F11. All of the macro subjects require a tripod and I used a 12 second self timer to minimize my effect on the steadiness of the camera. Mosses are, however, very difficult subjects in that some are on the ground, some on branches, and some on various parts of tree trunks. Of necessity the tripod is often placed in difficult positions on sometimes not too stable of a substrate. Nonetheless, it was generally possible to get good images with that DOF but sometimes critically visible leaves (not very many are completely visible) were not in focus and besides the images looked better with even more DOF. Thus I tried focus stacking and read yours and Lester's threads to get a rough handle on the appropriate aperture and focus stack step given my estimate that on average I needed about 1/2 to 3/4" to be in focus. With my initial tests of focus stacking in the field I chose F11, step 5. I found, of course, that I was unable to use a self timer with focus stacking employed so in reviewing some of my initial sets of images I noticed that the first one or two images sometimes were not sharp, suggesting I moved the camera on pressing the shutter. Interestingly, some of these stacks still turned out OK though I suspect they were not up to the best quality I could obtain.

I have since discovered that I can set a 2 second delay in silent mode which is the mode of focus stacking and future results should be better. My test above and your chart indicate that my initial choice of parameters for focus stacking were close and that I should be able to get the results I would like by shooting at F8 with a focus step of 5. It may well be that with some practice I may be able to recognize situations that would benefit from somewhat different settings. As you indicated in your last posts on this thread, there is a practical limit to how far the mathematics of a particular shooting situation can be helpful. Mosses, for example, unlike my tests (and yours and Lester's) are not strictly planar but also rise in the vertical. It would take a ton of trigonometry and measurement to be able to precisely quantify settings in advance. Here is an example of one of my earlier focus stacks. The curvature and tips of the leaves Identifies the species.

Claopodium crispifolium

Anyway, thanks again for your and Lester's works on focus bracketing and focus stacking. Thanks to your enlightenment, I will probably also learn to use my 8mmFE for situations in which I want to shoot wildflowers very close with a DOF to infinity.

Thank you, Richard, for your efforts and your sharing in such a splendid way of this interesting and usable information!

Richard Turton wrote:

Contrary to my expectations, this information has not proved to very useful other than helping to understand how the focus bracketing works. It is simply not practical to determine distances and consult charts in the field. For flowers, which is my usual focus stacking subject, I almost always use 8 frames with a focus step of 2. I have found no detectable difference between a focus step of 1 and 2. In fact, image quality does not deteriorate noticeably up to a focus step of around 5, which is well beyond what the usual depth of field calculations indicate.

I agree, all the charts in the world are not worth a 10 minute exercise, varying the parameters and determining what works best. For me with images like yours ( great image, BTW) I always use 5 shots with a focus interval of 3. Works 99% of the time.

I suspect that what we are trying to do reverse engineering on is Beta-functionality for Olympus. In itself it is a major breakthrough, but Olympus is fully aware that this whole field will significantly mature and become more user-friendly in the next years. And for now they want to have their hands free to change the ground rules and algorithms underlying this technology.

Thanks all for contributing to this extremely enlightening thread. I only take one thing into the field, and that is that the DoF is roughly equivalent to a stepsize of 3. Anything more will detract me from focusing ( ) on composition etc.

Richard,
Is there a way of determining the "focus differential" used in an image from EXIF data?  Looking back on images, it would be very useful to know what one used.  After all, this is the purpose of EXIF.  I couldn't find a specific item for it but a number of related items as follows:

• Focal Length
• F Number
• Focus Distance
• Focus Step Count
• Focus Step Infinity
• Focus Step Near
• Stacked Image:

Looking through your data it seems to me there should be a relatively simple relationship for focus differential, at least how Olympus is defining it.  I'm not familiar enough with the concept of focus step count to work it out myself from your data.

I use PIE as an EXIF reader.

BobT3218 wrote:

Richard,
Is there a way of determining the "focus differential" used in an image from EXIF data? Looking back on images, it would be very useful to know what one used. After all, this is the purpose of EXIF. I couldn't find a specific item for it but a number of related items as follows:

• Focal Length
• F Number
• Focus Distance
• Focus Step Count
• Focus Step Infinity
• Focus Step Near
• Stacked Image:

Looking through your data it seems to me there should be a relatively simple relationship for focus differential, at least how Olympus is defining it. I'm not familiar enough with the concept of focus step count to work it out myself from your data.

I use PIE as an EXIF reader.

First off, I don’t think ther is any information like you seek as part of the exif data.

I assume you are looking for some formula to determine the step “interval” and than add the number of steps to get an idea of the how much could be expected to be in focus. I have been using the Olympus focus bracketing since it was introduced and have come to the conclusion that no such equations exist.

I suspect they would be useless anyway since the actual results seem to  have a lot to do with the variables of focal length of the lens, the f stop of the lens (DOF variable) and the distance from the subject. So from my experience, shooting a 5 shot set with an interval of 3 with a 25mm lens at F/5.6 from a distance of 12”” from the front element gives  me one set of images that covers a certain distance. The same step interval for the same number of steps  on a 150 mm lens gives a different result since the DOF is different, and the closest focusing distance of the lens may be different. The variables are plenty enough that I have decided the best way for me is just plain old trial and error to gain the experience to set the parameters as I go along.

gary0319 wrote:

BobT3218 wrote:

Richard,
Is there a way of determining the "focus differential" used in an image from EXIF data? Looking back on images, it would be very useful to know what one used. After all, this is the purpose of EXIF. I couldn't find a specific item for it but a number of related items as follows:

• Focal Length
• F Number
• Focus Distance
• Focus Step Count
• Focus Step Infinity
• Focus Step Near
• Stacked Image:

Looking through your data it seems to me there should be a relatively simple relationship for focus differential, at least how Olympus is defining it. I'm not familiar enough with the concept of focus step count to work it out myself from your data.

I use PIE as an EXIF reader.

First off, I don’t think ther is any information like you seek as part of the exif data.

I assume you are looking for some formula to determine the step “interval” and than add the number of steps to get an idea of the how much could be expected to be in focus. I have been using the Olympus focus bracketing since it was introduced and have come to the conclusion that no such equations exist.

I suspect they would be useless anyway since the actual results seem to have a lot to do with the variables of focal length of the lens, the f stop of the lens (DOF variable) and the distance from the subject. So from my experience, shooting a 5 shot set with an interval of 3 with a 25mm lens at F/5.6 from a distance of 12”” from the front element gives me one set of images that covers a certain distance. The same step interval for the same number of steps on a 150 mm lens gives a different result since the DOF is different, and the closest focusing distance of the lens may be different. The variables are plenty enough that I have decided the best way for me is just plain old trial and error to gain the experience to set the parameters as I go along.

Your first sentence probably says it all.  However, while I agree with the rest of what you say, my point here is, if I looked back at stacked images I did a year ago, is there a way to determine what focus differential setting I used?

It's curious that while EXIF tells one so much, even how many images were used in the stack, it doesn't appear to indicate the focus differential setting itself.  I appreciate it's a new thing and Oly probably haven't got around to setting it up in EXIF yet but I thought there may be a way to calculate it from the data available.

BobT3218 wrote:

gary0319 wrote:

BobT3218 wrote:

Richard,
Is there a way of determining the "focus differential" used in an image from EXIF data? Looking back on images, it would be very useful to know what one used. After all, this is the purpose of EXIF. I couldn't find a specific item for it but a number of related items as follows:

• Focal Length
• F Number
• Focus Distance
• Focus Step Count
• Focus Step Infinity
• Focus Step Near
• Stacked Image:

Looking through your data it seems to me there should be a relatively simple relationship for focus differential, at least how Olympus is defining it. I'm not familiar enough with the concept of focus step count to work it out myself from your data.

I use PIE as an EXIF reader.

First off, I don’t think ther is any information like you seek as part of the exif data.

I assume you are looking for some formula to determine the step “interval” and than add the number of steps to get an idea of the how much could be expected to be in focus. I have been using the Olympus focus bracketing since it was introduced and have come to the conclusion that no such equations exist.

I suspect they would be useless anyway since the actual results seem to have a lot to do with the variables of focal length of the lens, the f stop of the lens (DOF variable) and the distance from the subject. So from my experience, shooting a 5 shot set with an interval of 3 with a 25mm lens at F/5.6 from a distance of 12”” from the front element gives me one set of images that covers a certain distance. The same step interval for the same number of steps on a 150 mm lens gives a different result since the DOF is different, and the closest focusing distance of the lens may be different. The variables are plenty enough that I have decided the best way for me is just plain old trial and error to gain the experience to set the parameters as I go along.

Your first sentence probably says it all. However, while I agree with the rest of what you say, my point here is, if I looked back at stacked images I did a year ago, is there a way to determine what focus differential setting I used?

It's curious that while EXIF tells one so much, even how many images were used in the stack, it doesn't appear to indicate the focus differential setting itself. I appreciate it's a new thing and Oly probably haven't got around to setting it up in EXIF yet but I thought there may be a way to calculate it from the data available.

I suspect that it may never be in the exif, if it isn’t now. The focus bracketing feature has been with us for a few years and a number of firmware updates. That said, I haven’t been an advocate for the in-camera focus stacking, so I have not looked into what is available from that feature, maybe it’s included.

Richard, thanks so much for this article. I have been struggling with this for a while (I am writing this in March of 2019) and I recently was working on tethering my EM-1 Mk II using Olympus Capture to LR when I decided to do some more research and came across your 2016 article.

So, I tried to setup a spreadsheet using your two basic formulas - to replicate your calculations so I could then add some other columns and complete the rest of the arithmetic.

Perhaps I have missed something, but my version of your 'jpeg' tables were not even close to yours.

I have included a screenshot of my spreadsheet - with the formula showing for a cell that I thought would match your table.

Can you explain what I've missed.

Focus Step (metres) = f-Number x (Focus Distance (metres) / 21.4) ** 2

thanks, Jim

Richard Turton wrote:

eaa wrote:

Truly a great post indeed! Thx a lot 👌.

I will most probably be using Focus Stacking w/ the 40-150 PRO (with and without MC-14) for dragon- and butterflies (when a flawless EM-1 fw v4.3 is out that is). I will reference the guidlines here for sure, although the longer FL will require some other focus values. I guess for a 1m working distance, some diff. value at 2-3 and F5.6 will be in the ballpark w/ the 40-150. I will need a total stacked DOF around 10-12 cm or so.

Exiting times ahead 😊.

Unfortunately I will not be testing the 40-150 Pro because I have an FT 50-200 f/2.8-3.5, which is not as good but is already paid for. I did measure the focus bracketing performance of the 12-40 Pro (https://www.dpreview.com/forums/post/58770991).

I have had almost no success with focus stacking of insects because most of them refuse to hold still for the one second or so it takes for the 8 exposures. The individual frames are sharp but stacking them just produces a buggy blur. It can be useful to take a stacking sequence and select the one frame that has the best focus. This also applies to flowers that are moving slightly in the wind.

"flowers that are moving slightly in the wind" have been a bugbear of mine so I've invested in a Wimberley Plamp.  So far so good.  It promises to make more of my day usable for native plant photography.

Richard Turton wrote:

A simple square law is just not a very good fit to the measured data. The actual equations I used are:

Focus Count = 1546 * (D + FL) / D^2 +1228

Step Count = -2.423 * (D + FL) / D^2 +75.18

The constants correspond to a least squares fit to the measured data for the 60mm macro. Other lenses will have different constants.

Oops, I'm getting confused now.....we have two new terms "Focus Count" and "Step Count".....and 2 new variables, D and FL ???

I suspect D is a measure of Distance - from front of subject to sensor .... but FL - something to do with F-Stop 'cause that has to be in there.

Could I ask you to perhaps define these and I'll take a crack at the spreadsheet again.

thanks, jc

Thanks a lot for the splendid work you did on explaining how the stacking size works on Olympus cameras.

I considered this as being a very valuable document and wanted to reproduce the table you presented with the step sizes using your formula, but didn't succeed in it.

Focus Step (metres) = f-Number x (Focus Distance (metres) / 21.4)²

Finally, I found out that something had happened to the representation of your formula and that what you wanted to show is this formula

Focus Step (metres) = f-Number x (Focus Distance (metres) / (2^-1/4))²

or

Focus Step (metres) = f-Number x (Focus Distance (metres) / 0.841)²

With this formula I have a perfect match between my version of the table and yours.

Now that I know the theory, it's time to start practicing as so far I didn't succeed in getting acceptable results with my stacking efforts.

Thanks again for this nice work!

Kind regards,

Dirk Claesen

https://photona.org

Ifixit wrote:

Thank you for all the work you have put into this! I wish Olympus would pay you to write their manual:-)

Agreed!
We need an algorithm with user inputs of lens to subject, depth of subject, f-stop for background blur and let the camera can optimise the rest.  I can't believe Oly doesn't already have this in the pipeline but are refusing to release it for some insidious reason.  They are leaving us blundering in the dark with hit and miss of a macro feature that is so immensely powerful.