Johann Schutte Loc: NZ

My apologies of this has been addressed previously or my math is incorrect. Please comment and correct.

With all the variation possible with lens/tube/diopter/objective combos I found it increasingly onerous to try and calculate magnification and inevitably would return to the crude 'take a pic of a ruler and calculate' method.

So here's what I think might improve accuracy. I have checked it out with my MP-E65 at 3-5x and it seems to work, as well as my tube/diopter/objective as in the attached pic. Let me know what you think.

Attached file:
(Download)




(Download)

naturepics43 Loc: Hocking Co. Ohio - USA

I see at least two problems. 1. Gonna be really hard to count pixels at low magnifications. 2 If you want to improve accuracy, how is "approximately 12.5 units" improving accuracy? I don't mean to sound critical but measuring the picture of the rule in pixels & doing the math IMO is quicker & more accurate. Thanks for showing us another option.

Johann Schutte Loc: NZ

My interest is in calculating magnification of 5 and above. I would see no need to calculate magnification in this way (or with a ruler) when using an MP-E65, as the calibration (1-5) on the lens itself is good enough and altered by turning a single control on the unit.

With 5+, objectives, tubes, reversed lenses, and diopters (as tube lenses for finite objectives) come into play and the calculation is more complex. This is where this method works for me. It’s quick and offers more accurate guestimates than is possible with realsize mm.

As to comparative accuracy - If you estimate perfectly using this method (method 1) or the ruler method (method 2), your deviation from the truth is 0.

Lessay you make a Guesstimate error (GE) as follows:
(Snip attached as pic for improved readability)


M1 Error % M2 Error %

GE of 0.1 12.4px 0.08 3.8mm 0.26
GE of 0.2 12.3px 0.16 3.7mm 0.54
GE of 0.3 12.2px 0.25 3.6mm 0.83
GE of 0.4 12.1px 0.33 3.5mm 1.11

The error difference between M2 and M1 is in the order of a factor of three at each level.

As a further benefit, If you wanted to make a calculation at all (as I do) you obviate a physical ruler completely. You can use the pixels on your screen and not have to transfer a pic of a ruler for further analysis. As a matter of fact you don’t even have to take a pic with M1 at all. You can just view and count on your camera monitor, or alternatively do the count on the computer screen if you have your camera appropriately connected.

I am in process of developing a spreadsheet that describe the various configurations I use, e.g. combos of various lengths of tube, with various objectives, using various diopters as tube lenses (where the objective requires a tube lens).

Further as to accuracy. Scientific measurement is never finite. You can’t measure the true length of an inch. Our cameras reflect this reality. They’re all manufactured within certain tolerances. The idea is to be as uniform and accurate as possible though.

naturepics43 Loc: Hocking Co. Ohio - USA

Sorry. You did say "Let me know what you think". I guess the reason for wanting to know the magnification would dictate the degree of accuracy needed. But in the end, both ways are close enough since the stacking program changes the magnification of the final image. Your spreadsheet sounds like an interesting project. Let us know how it works out.

photosbytw Loc: Blue Ridge Mountains

Wow Johann.........you put a lot of time and effort into this, I'll give you that. Personally, I've never been that curious. Admittedly, I have taken pictures of a ruler and calculated the magnification but too much thinking drives me to drink.

I've been enjoying your work so this is not finding fault. I understand and appreciate those who explore possibilities and ask themselves "what if".............



tw

Mark Sturtevant Loc: Grand Blanc, MI

If I understand correctly, this is no different from using a ruler except it is a very finely divided ruler for measuring mag. at high magnifications. Also it is very clever.
An alternative would be to get a micrometer scale on a microscope slide. That would save you from the earlier step of calibrating the ruler before you calibrate the camera set up with this 'ruler'. But a micrometer would also be a bit less fun.

Johann Schutte Loc: NZ

I'm appreciative of all of your responses. Haven't thought of the micrometer idea. I will have to look into that now.

Mark hits the nail on the head - the pixels represent a finely divided ruler and that's the essential benefit.

The frustration for me originates from variables that are difficult to calculate in terms of their cumulative effects.

Examples:

• I don't know how deep into the camera the sensor is located and what effect this has on the length of applied tubing. 4/3 mirrorless differ from smaller bodied dslrs, which in turn differ from larger bodies ones, Canon is different to Nikon and the 6d differs from the 6dii, i.e. generational variances occur.

• Equivalently specced 10x objectives differ structurally in the sense that the back element is more or less deeply recessed. Similarly, equivalent diopters are thicker/thinner. Adapter rings vary in width.

Once you start adding all the components together in free experimentation with various configurations, the butterfly effect/chaos theory comes into play.

I am sure some would view all of this as indicative of anal-rententiveness. I am happy to wear that label, as I think it’s dispensed fairly ( I know it’s true) and has for example over the years precipated a good deal of eye-rolling from my wife. I always point out to her that she should be grateful that I’m not anally-expulsive in personality expression, as this would have made her life even harder.

It seems like quite a process to go through, however it isn’t – It’s literally a matter of minutes to do the calibration once you’ve got the camera set up. In my case it’s literally a question of counting the pixels on the screen (I don’t take a pic) multiplying by 0.311 and then dividing 35.8 by the result.

I ‘ll soon be moving into 20:1 at which point this is going to become even more significant to me.

PS. The more accurate your mag calculation, the more economical and efficient the stacking process. It allows increased accuracy in calculation of DOF and stepsize, which of course translates to improved outcome (e.g. less noise) in the stacked image.

photosbytw Loc: Blue Ridge Mountains

Johann Schutte Loc: NZ

Thanks for the affirmations. My wife has threatened me with the MH act before AND called the %^*#!!# family. I calculated germination percentage of carrot seed mindful of climatological and metereological conditions, grain size of sand and then computed a proportional admixture of carrot seed and sand volume, to be manually dispensed in a rehearsed motion over specific time increments with a hand held shaker into a garden bed of a specific size to yield a precise amount of crop per area to feed a family of five for a specified period. This was too much for her. and more to the point - My calculations were out.

photosbytw Loc: Blue Ridge Mountains

Are reading Terry Pratchett books? Or, should I say ook?

Mark Sturtevant Loc: Grand Blanc, MI

Getting to 20:1: You are just going to have to mount the camera onto a full-blown microscope. I am sure you will do well. Micro-focus stacking and even panning with the X, Y stage controls to stack more. Then stitching the whole thing together into a mega-mega-pixel image.
Be sure to share with us!

sippyjug104 Loc: Missouri

Interesting topic and one that is close to my heart for I enjoy shooting under various degrees of magnification. Honestly, I have no idea where I am most of the time other than I try to fill the camera frame with the area of interest of the subject, like mouth parts, eye, or a head view.

I use microscope objectives that are both finite and infinite with a Raynox DCR-150 (208mm) as the tube lens. I also stack lenses together and generally just putz around and experiment with the good and not so good results.

I've played with magnifications at 20X and even up to 40X however staging was so critical that without micrometer based staging I gave it up. At 20X and above, I couldn't even breathe and touch anything and the working distance was only a few millimeters away.

Personally, I would move into a metallurgical microscope setup with tool-maker objectives. Perhaps I'll treat myself to a generous Christmas Gift before my time passes for there is an entire new world in the realm of microscopy to be seen.

Johann Schutte Loc: NZ

Yep. The thought appeals to me. Would love to do a 100:1 Gigabyte image.

I have to get into understanding objectives better. I have only three, those being my Nikon 10x, El Cheapo Amscope 4x (which is surprisingly good), and a Lomo 8x, which has been a disappointment. My current knowledge of objectives is largely based on what other people have found to be useful configurations.

Have you come across this yet? https://www.closeuphotography.com/qioptiq-magx-5x-objective

Sells for US$16k apparently.

Only the lotto would make it possible.

Thanks for your suggestions/comments Mark. You defined succinctly that the pixel method really is just a better ruler!

Regards, Johann.

Johann Schutte Loc: NZ

Gimme a fer instance Sippy!