[Douglas_of_Sweden]

Thanks for all answers!

I was fearing that diffraction would come back and kick me in my ass.
Previous work by colleagues of mine have used video cameras, originally high resolution video tape, and more lately digital sensors. I had first assumed that the lower limit in their system of 20-30 um was due to lower pixel resolution, but it may actually have been diffraction as well. I was hoping using a DSLR would give me more resolution, but in practice the difference may not be that large. But there are more reasons why I wanted to use a DSLR...I have a couple of them, so it would be no new investment...but it also has to do with the set-up.

I understand that reversing a 50mm on a tele would give my much larger resolution, but it would also force me to work at very narrow focus distances. The main focus of these experiments is to measure the aerosol particles that form from the bursting bubbles. These are called sea spray and consist of small droplets of water, salt, and when using real seawater, or as we will do in these experiments, artificial seawater with marine bacteria mono-cultures in the water, organic surfactants. With such small focus distance I would be forced to have the camera inside the tank. This will be an air-tight steel tank, teflon coated below the water line (to minimize the amount of surfactants that stick to the tank wall), and in high grade steel above the water line to minimize particles losses to the wall. Having a camera and lenses in the air part of the tank would interfere with the aerosol. And I don't think even the Pentax seals of bodies and lenses would survive that for a long time. The particle concentrations are higher than anything you would experience in the real atmosphere on-board a ship. It would be like constantly sitting on top of a white cap with your camera. And that mixture of salt and organics do nasty things with most materials. We are looking at several hours per experiment, days in some cases.
If we would place the camera outside the tank but in the water line shooting from the side through a window would lead to other complications. Outside the inner tank there will be a second wall and between them we will be circulating water from a temperature control system. Different experiments will be made at different temperatures from maybe 30C down to close to -1.85 (the freezing point of sea water).

The reason why the initial question assumed that I would work with a 1:1 macro is that the close focus distance of a 90 or 100mm 1:1 macro roughly agree with the vertical distance we will have from the water surface to the top of the tank (which will be a horizontal steel lid on the otherwise cylindrical tank). The macro's I've got available are the DFA 100mm at f2.8 1:1, the Adaptall 2 SP 90mm f2.5 1:2 (which with the original extension tube gives me 1:1, alternatively with the 2x converter gives me 1:1 at 180mm), and the Hexanon AR 105mm belows lens which with the right extension (and my own adapter) gives me 1:1. I also have a couple of 50mm macros, but then again the close focus distance decrease. I have automatic tubes or converters that could get me a bit beyond 1:1 on the DFA 100 macro.

Another reason why I think going from the small video sensors they used before to a APS-C sensor is the upper limit of the bubble sizes. The largest bubbles are several mm large and may be very important for the droplet production, and the bigger sensor, the larger is of course the largest bubble we could size.

It is tempting to try a second set up where we just make bubbles and don't bother to collect and sample the aerosol. Then I could use the approach Falcon suggests, or use my RMS mounted Canon micro lens that takes me to somewhere around 10:1, at some ~1cm distance. Without the closed tank and not so long periods of operation, the camera gear wouldn't be that badly exposed to the salt etc. But right now I need to figure out the best solution for documenting the bubbles at the same time, in the same set up as the closed tank system where we collect the aerosol. On Friday we are meeting the workshop people who will build the steel parts of this construction.

I was aware of the contradiction between small and large apertures for max DOF and max resolution. What I hope is that it will not be that critical if we work with a thin DOF if we are exactly focused on the water surface. It won't matter if the top of the bubbles are out of focus as long as the rim of the bubbles are in focus. But that may turn out differently in the real application when the water jet that create the bubbles and the rising bubble plume will cause turbulence and mixing in the water.

Short exposure time will of course be critical. But at least we know from previous experience that the bubbles will move much slower on the water surface than they do while they rise in the water. We will have to play around with this. I intend to set up a test rig in the lab above an open water tank to just play around with light, apertures, exposure time etc.

Lot of light will be good. But I intend to use flash, not lamps. Lamps will produce heat and could interfere with the need to cool and control the temperature.
I like newarts idea to illuminate from the side below the water. That means that we will need a glass window in the steel wall (something to discuss with the workshop people on Friday), and we will need to submerge the flash inside the cooling water, but that will at least be much easier than submerging lens and camera.

Another question. Lets assume that I would buy a full frame DSLR, that would give me more resolution, right? How much more?
I might be able to fit that in the project budget. So maybe a Canon FF...(Pentax, this would be the time to announce a full frame body...I need it at the latest in September ).

[falconeye]

1. Distance: Reverse mounting a 50 leaves you with the registration distance as distance between outermost lens edge and subject. It is 46mm and not bad at all with my approach.

2. Focussing from the top to the water surface: Good luck. My intuition says it can't be done except you don't simulate the real thing and then the experiment is pointless.

3. Spray: I anticipated you put the camera into a pastic bag with front filter and bring it as close to the action as possible. You want to see the bubbles in the air, don't you? Otherwise, I don't see the point of the experiment which shall allow to describe energy exchange across the water surface. And this means above it.

4. Full frame has no advantage here because lenses are still 1:1. A smaller pixel pitch brings you below 5μm. So, a Q with DA100WR resolves a bit more. But as you say, the largest objects then will be small. Other brands have 200mm 1:1 macros. But you don't need FF for them.

5. Please, try to make some shots now in a preliminary fashion. You lack a feeling for the difficulties you are going to run into. Your concerns are rather theoretic. It would be a pitty if you need to build another tank to get new results...

[Lowell Goudge]

I have a technical application at work. We do research on how waves that break at sea form white caps, and how the bubbles in the white caps burst and eject tiny droplets into the atmosphere which we call sea spray. These droplets aka aerosol particles have for example climate implications. Try holding a hand over a glass with mineral water and you will feel similar droplets hit your skin.

What I want to do now is to photograph the bubbles at the surface of the water in a laboratory environment. We know from before that the bulk of these bubbles are 0.1 to 0.2 mm in diameter, and that the number of bubbles then gradually decrease with increasing size up to a small number of bubbles of several mm (assuming that we talk about sea water with a salinity around 3.5%, in fresh water you get much larger bubbles, and less of them). At the lower size range previous experiments mostly go down to about 30-50 micrometer (0.03-0.05 mm) because the digital video records used to count these could not resolve smaller bubbles. But some recent studies suggest the existence of even smaller bubbles. We actually don't know the lower limit in size. Now those studies used very expensive high speed high resolving cameras. I don't have room for that in my budget right now, and no use for the high speed. But I do have a couple of DSLRs, and several decent 1:1 macro lenses. And I've been thinking that I should be able to do better with modern APS-C sized sensor than they did a couple of years ago with tiny video camera sensors.

Assuming that I mount the camera looking downward some 10-50cm above the water surface with a 1:1 macro at its close focus range, have a good lightening etc and start making bubbles. What could I resolve? We will test it of course, but from a more theoretical point, could one predict how small things we could resolve?
I presume that the pixel size of the sensor matters?
And the lens resolution?
What else?
How do I estimate the size of the smallest subject I could resolve?

Many thanks in advance for any help with this.

Are you trying to prove the existence or get a good image of a small bubble ? To just prove the existence, I would assume you want a bubble to contain at least 10-20 pixels, but let's say a diameter of 5 pixels because this is an easy number to work with.

Depending on the camera, and lets say a K5 which has about 4800 pixels across 24 mm As a result you have 200 pixels per mm on sensor therefore if you want 5 pixels in diameter you have 1/40 of a millimeter or 25 microns

What this means is you need to go well beyond 1:1 macro to really prove the existence, 1:10 would let you resolve about 2.5 microns

[Lowell Goudge]

Douglas. You seem to be looking for droplets from above, what about a side shot? This might, with a reversed leans, and a flash for freezing the image be ptty good. Everything except the ones in the focal plan would be just big white globs.

[falconeye]

Douglas. You seem to be looking for droplets from above, what about a side shot? This might, with a reversed leans, and a flash for freezing the image be ptty good. Everything except the ones in the focal plan would be just big white globs.

Lowell, I understand you skipped all responses, but why then did you respond?
This is exactly what I proposed in great detail.

[Douglas_of_Sweden]

1. 46mm is not enough, trust me. We need a certain air volume inside the tank and a certain distance for the water jet that creates the bubbles. Sorry, but these constrains are solid. They are based on a large number of previous experince and several peer reviee papers.

2. I understand your concern, but it has been done before in open tanks.
We are not going to shoot the bubbles in the air, they dont get air born, but we are to size them on the water surface. Perhaps you missunderstood me: It is not the droplets in the air we are to size. These are much smaller...from a few nanometer to a few micrometer. We sample the air from the tank and size them using partly Mie-scattering from a laser beam and partly using differential mobility ( for those too small to scatter light).
What I dont understand in your comment is why it would not simulate the "real thing".

3. A lens and a camera close to the water surface would interfere with the produced aerosol. When a bubble burst it eject droplets with enough speed to hit the lens. Hence we would loose droplets in the size range above ~100 nm to what we call.interception. Smaller droplets than that is rapidly lost by Brownian diffusion to any surface. Here the surface properties are important. Steel has the smallest sicking coefficient of materisl.we can use in practise. Particles stick easilly to plastics. So a plastic bag would offer a bad aerosol sink, which would modify the aerosol spectra and make the experiment pointless.
This is not at all about energy exchange, but about mass exchange where the size spectra of the resulting particles depends on the bubblen wize, and the buble size in its turn depends on surface tension more than any other factor, which depends largely on the biogenic organic surfactants.

4. Thats a pitty...I wanted a FF.

5. As I said, we have some previous experience in an open tank and will do more tests particulary on the lightening to see how we get the highest contrast, since this is important for the sortware that will size the bubbles. And no...I'm much more of an experimentalist than a theoretical wcientist. But the tests cannot change much of the tank design. Constrains on how we produce the bubbles, sample.the.aerosol and rgulate temperature constrain that design. What we can play around with is where we place "windows" in the steel tank for camera and flash. This means of course that I later on can send the tank back to the workshop and add more windows in different places.

My appologies that this became very technical far from the optics and much more detailed than I intended to and I'm sorry if much of this has become a discussion above your head between two physicists. But since I apprechiate Falcons advice very much I felt I needed to explain it in some more details to show that I dont rule out his first suggestion for any insignificant reasons. Reversed lenses and camera inside the tank we have already ruled out beforehand.

1. Distance: Reverse mounting a 50 leaves you with the registration distance as distance between outermost lens edge and subject. It is 46mm and not bad at all with my approach.

2. Focussing from the top to the water surface: Good luck. My intuition says it can't be done except you don't simulate the real thing and then the experiment is pointless.

3. Spray: I anticipated you put the camera into a pastic bag with front filter and bring it as close to the action as possible. You want to see the bubbles in the air, don't you? Otherwise, I don't see the point of the experiment which shall allow to describe energy exchange across the water surface. And this means above it.

4. Full frame has no advantage here because lenses are still 1:1. A smaller pixel pitch brings you below 5μm. So, a Q with DA100WR resolves a bit more. But as you say, the largest objects then will be small. Other brands have 200mm 1:1 macros. But you don't need FF for them.

5. Please, try to make some shots now in a preliminary fashion. You lack a feeling for the difficulties you are going to run into. Your concerns are rather theoretic. It would be a pitty if you need to build another tank to get new results...

[falconeye]

Thanks for your reply allowing me to gain so much insight.

I now understand much better.

What I dont understand in your comment is why it would not simulate the "real thing".

That's because of an impression I gained from your OP. That it is the ocean's water/air surface you're going to study. And this I cannot imagine without waves (the real thing for a surface). And waves would make it very hard to impossible to gain focus on the surface with 1:1 macro photography. I imagine waves are a major factor in the properties of this surface, as far as weather or climate models are concerned.

[stanislav]

That's because of an impression I gained from your OP. That it is the ocean's water/air surface you're going to study. And this I cannot imagine without waves (the real thing for a surface). And waves would make it very hard to impossible to gain focus on the surface with 1:1 macro photography. I imagine waves are a major factor in the properties of this surface, as far as weather or climate models are concerned.

That is exactly my consideration (or question) - what happens with all the waves and turbulences?

There are 2 more things I am thinking about:

1. Provided that the water is smooth enough, would it be a good idea to use a "sweeping laser" for illumination? If the laser is sweeping just above the water surface you would basically get a 2D image of the water with the bubbles lit from the side. Hopefully there would not by too much difraction so the edges should be visible. If the sweep can be controlled than shutter speed covering just one sweep is more than enough to capture enough light for the "2D" shot. Now this may not be feasible, but It would be at least cool to try

2. You may consider using polarizing filters to get more contrast. It may be worth to experiment with polarizers both on light sources and the lens (use linear polarizers on light sources and circular on the camera if you use any automation regarding the shutter speed).

[newarts]

My appologies that this became very technical far from the optics and much more detailed than I intended to and I'm sorry if much of this has become a discussion above your head between two physicists. But since I apprechiate Falcons advice very much I felt I needed to explain it in some more details to show that I dont rule out his first suggestion for any insignificant reasons. Reversed lenses and camera inside the tank we have already ruled out beforehand.

Thank you for your extended discussion and details of the experiment it is very helpful to know the details of what you'd like to do and the practical obstacles.. Falk's concern about the motion of the air/water surface is valid. It seems like you are talking about duplicating the cold, raging north sea in your little tank.

Is the water jet pointing up? There must be turbulent vertical motion at the surface. I hope it isn't too overwhelming

Let's imagine a scenario where the camera is looking straight thu a port in the one layer wall at the top of the tank. Use the highest resolution lens you can at the highest magnification that will allow a sharp focus, then rely on "Catch in Focus" to trigger the camera.

If CIF isn't fast enough to catch a bubble as it flicks in and out of focus, you'll have little other choice than to take movies and wait for random coincidences.

The minimum sized feature you'll be able to resolve on the sensor is limited by diffraction disk size and size of the sensor's pixels. At a 1:1 magnification (sensor size/subject size) the smallest feature that can be seen on the sample equals the sensor pixel spacing; at 2:1 magnification, one sensor pixel spacing is twice the actual feature spacing, etc.. Therefore you want to optically magnify the image on the sensor with an f-stop that matches the diffraction limit appropriately with the sensor pixel pitch.

It is helpful for me to work backwards from the image on the sensor. Say the image is a bright pixel followed by a dark pixel followed by a bright pixel etc. A square wave. Each pixel's brightness is due to a source on the sample. The source could be infinitely small - we've no way of knowing because diffraction spreads it into a disk. Clearly it is to our advantage that the diffraction disk be equal to or smaller than the pixel.

The image was projected onto the sensor with a magnification, M, thus distances on the sample are equal to distances on the sensor divided by M. Clearly we want to have a large M to see small features. It takes 2 adjacent pixels to resolve a line pair so referring back to the sample, the resolution (lines per mm on the sample) is M*sensor.pitch/2.

The working distance limit of about 100mm seems ok for available lenses. (at least for non-IF lenses). Say you want a magnification of 4X. The working distance for a 100mm enlarger lens at 4X is wd=Focal.length(1+1/M) = 120mm. The diffraction disk diameter at 4X is nominally 4 times the size of the disk from the unmagnified image so nothing is gained. But this is not always the case in real life. Here's an example of an Enlarger lens in which the on-the-subject resolution improves with magnification: The blue line is MTF=50%, the green line is MTF=10%.

Schneider Componon-S 50mm f/2.8 Macro lens tests

notice that magnifying by 4x (extending a bellows for example) cuts the on-sensor 50%MTF resolution about in half, but the associated magnification has decreased the associated on-sample spacing by a factor of 4. This means smaller features can be resolved by increasing magnification beyond 1:1 with this lens. This setup would be about 3micron per pixel on the sample.

Try lighting from the side under the water surface. Angle the beam of light towards the water surface - the light will then skim along the underside of the water surface trapped by the total internal reflection effect except where its direction is drastically changed by something like a bubble wall. This should result in well lit bubbles on a dark background.

An alternative scheme would be to shoot from the bottom upwards. Use a microscope objective at the end of a probe a few mm under the surface of the water. Then spray and sloshing will not interfere. Such instruments are likely available in the medical field. The reason for suggesting this direction is to increase resolution. Optical resolution increases with decreasing f-number - high magnification objectives have large effective f-numbers. A standard 10X objective has a Numeric Aperture of 0.25. F-number is 1/2NA which yields a diffraction spot diameter of 2.44*2*wavelength ~ 2micrometers in the blue at a magnification of 10X. This is a large improvement over what you can do with a camera lens.

Dave

[falconeye]

Thats a pitty...I wanted a FF.

You can fake an argument ...

You can say you need the extra working distance offered by a 200mm 1:1 Macro as Nikon offers over a 100mm 1:1 Makro. But you need closer to the field of view of the 100mm on APSC. This would leave you with a fullframe Nikon and if opting for max. resolution, the D800. This should exactly match your Xmas wishlist, if that's what you're looking for

[Catalana]

Why do I keep thinking of "Alka Seltzer" in a glass/tray of water?

[newarts]

@Douglas.

Have you made progress towards your goal? What has worked for you?

Dave in Iowa